MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Panel mounted controls vs PCB mounted controls, and more!
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Author Panel mounted controls vs PCB mounted controls, and more!
Flux Jetson
042312 - EDIT: At the insistance of a few members I have changed the name of this thread. I was using a technically incorrect phrase that seemed to matter a lot more to some than the actual content and well meaning context of the post. At their behest I also tried to correct any usage of the phrase throughout the entire posting. I've attempted to change every instance of this error of mine just to make the thread more technically proper. If I miss one or three instances, you're just gonna have to replace "SURFACE MOUNT" with "PCB MOUNTED CONTROL COMPONENTS" in your head.

Just to offer some defense of my ignorance .. in the second paragraph I CLEARLY state how I am using the phrase "Surface Mount". Right, wrong, or indifferent ... there it is. I'll not change that part of my original post.

Note that I make no claims of being a qualified bonafide "TECHNICIAN" .. in fact, I barely graduated from a rural high school with a D- average in 1978. I'm just a guy that has taken an interest in what manufacturers are trying to sell to me. So if I come across as "thinking I know everything" please know that all I know is what I made the effort of teaching myself. The info I found and have come to embrace is stuff that I thought other consumers may find interesting and useful. I'm all about "arming" myself with knowledge, and if I can help spread some knowledge and/or findings, it just makes things better for a few more folks that don't have disposable cash to throw away on mistakes made that could have been avoided with a tiny bit of information.

So PLEASE take this post as it was intended .... and that is to inform other serfs like myself about a few ways to tell a good piece of equipment from a piece of equipment that simply LOOKS good, and markets well. Always remember ... just because something is popular does not always mean it's "good".

(END EDIT).


"Surface mount" (EDIT: PCB mounted control components. End edit.) vs Full floating components:

I've been throwing that phrase around a LOT since I joined up. A few people have asked me wtf I'm yakking about. So I figured I'd set up a post discussing this exact construction method and the goods/bads about it's use.

In the context that I use the phrase, "Surface Mounting" describes how the various controls on a given module or device are mounted to the device. "Surface" meaning the surface of the printed circuit board inside of the device, "Mount" refers to the fact that the controls are mounted directly to the circuit board. It essentially means that a given part, most commonly the pots/knobs, I/O jacks, and any switches .. are mounted directly to the PCB rather than being mounted away from the PCB and having individual wires connecting the pot/jack/switch to the circuit board.

Here's a 3u made by Doepfer .. I believe it's a ring modulator (out of my own previous Doepfer rig that I sold about a year ago). Notice that this is average Surface Mount (EDIT: "pcb mounted control components" end edit) tech here. All that holds the PCB in the unit is the (plastic bodied) I/O jacks themselves. The jacks are soldered directly to the pcb, and the jacks are also mounted to the panel with the chrome nuts that you can see when you look at the "outside" of the panel. This isn't too bad .. the patchcords that most 3u systems use are pretty soft and compliant and don't exert much sideways/up-down forces on those jacks, so I don't imagine that there's too much stress placed on the solder joints of this ring mod's pcb.



Now, here's a shot of a Doepfer Midi Interface. Again we have the jacks supporting the pcb, and the big rotary switch isn't mounted directly to the PCB, it's actually mounted to the panel via it's own barrel and secured with nuts. You can see the wires running from the rotary switch to the pcb, indicating it's not soldered directly to the pcb, so any rotational force that is generated by using the rotary switch is NOT transferred to the solder joints. This is good. What I'm not too crazy about is the *Hot Glue* used to mount the Midi connector to the control panel. C'mon .. hot glue? Really?





Ok .. moving up the quality ladder here. Here's some Well Done surface mount (EDIT: PCB mounted control components. end edit). This is a Livewire product. You can see the toggle switch (the red thing) has a backing nut that allows the front panel nut to tighten against so when the panel nut is torqued down all of the tightening forces are placed on the panel between the backing nut and the panel nut. Little or no forces are transferred to the pcb when this switch is used by the operator of the module. Also take note of the high-quality switch used here. And as a side note, also notice the quality of the solder joints themselves. Danged good work here! Take notice of how thick the PCB is as well. Nice!



Here's another shot of another Livewire product. You can see we're still dealing with Surface Mount tech (EDIT: PCB mounted control components. end edit.) here, however the pots and switches are all very securely mounted to the control panel with both backing nuts AND panel nuts which ~sandwich~ the control panel between them. The PCB itself is actually supported by the surface (EDIT: PCB. end edit.) mounted components however the pcb is seeing very very little operator stresses caused by the user ~tweaking on~ the controls. The controls are so well mounted to the control panel that the PCB is protected from us dumasses as we tweak the controls, or something accidently knocks into the knobs on the control panel in the studio. Also take notice of how thick the PCB itself is. This is very well done design engineering as well as well executed construction. Still not "the best way" to construct a control module though ....



.... (below) this is. Here we have Full Floating Printed Circuit Board tech. This is about the best way to build electronic gizmos. The downside is the extra labor and skill required to assemble the unit .. which can translate into costing us more money.....

The pcb is completely isolated from any of the panel mounted controls, and completely isolated from any user-generated stresses placed on those controls. The PCB is mounted on it's own legs .. or "standoffs" and is wired completely separated from the controls.



(below) .. this is a VCO and a VCO mixer that I wired together myself, easily done and the work is easy to make neat and tidy due to the extra room and construction design. Notice how any forces or stresses placed on the pots and jacks never even comes close to transfering to the pcb. The pcb is mounted on it's own and the controls are mounted on their own with interconnecting wires that make it all work.


(NOTE: a few responders realy disliked that I added this side note ~below~ on resistor quality and typification and one or two folks really wanted me to remove it. I will not remove it. I feel this issue is something that is relevant to the overall marketing attitude of the manufacturer and is something that us average consumers need to be made aware of. For those that don't like it, please skip this section. End note.)

Also take notice of the "blue" resistors. Blue resistors are rated at 1% of their rated resistance. That is to say that a 100k resistor will have a tolerance of 1k .. it may be as low as 99k or as high as 101k. Now look at the other modules pictured above in this post ... they have TAN resistors. Tan resistors are rated at 5% of their rated resistance. This means that a 100k resistor may be actually 95k or up to 105k ... so they actually have much ~looser~ tolerances than the blue ones do. Why do some manufacturers use Blue 1% and others use Tan 5%? To save about a half-cent per resistor. I shyte you not! A half cent. Yes, I realize it adds up a bit when a given module may have 50 resistors .. but even at 50 of them we're still talking about saving 25 bloody cents. I prefer a bit of acuracy myself. I'll pay the extra dollar or two that the 25 cents would translate too. But that's just me.



You can see the wiring is totally free to flex and move a bit, and the extra length of wiring also allows for totally WAY easier repairs and/or mods to be performed to the pcb when built like this. High quality all-metal pots are used in this case (these are Synthesizer Dot Com modules jsyk).



In the case of the Dot Com modules, plastic cased jacks are used, but if you look closely the plastic case has ZERO to do with the actual construction of the jack. The barrel is all metal and it is a direct mechanical connection to the tip connector as well as the solder terminals. These jacks are used due to their UBER- Buttery-Smooth insert/extract feeling. Patchcords just slide right in and pull right out with little cussing. Also notice the "broke edge" on the panel ... that raised lip the comes up the side .. it's there to stiffen the panel so it's not just a ~flat plate~ that will flex when patchcords are inserted/removed. Another +1 for this design.



Here's the two systems the modules came out of that these pictures were taken from. My old (and sold) Doepfer rig, and my current in-progress Dot Com 5u rig.





So there we have it folks. Surface mount (EDIT: PCB mounted control components. End edit.) ... some not-so-bad ... some really well done. And a bit of Full Floating mount as well.

What's best? You decide. I am partial to full-floating setups my self, but if the sound is really good I'm willing to overlook a lot of surface mount negatives. The Doepfer stuff is about the least well built type of surface mount (EDIT: pcb mounted control components. End edit.) tech I'd even consider though.

The type of surface mount (EDIT: PCB mounted control components. End edit.) stuff that uses NO panel nuts gets put on the bottom of my want lists. Unfortunately, there are a LOT of really nice little audio toys made with no panel nuts ....



(below) this one just really gets me .... it sounds so good .... but that construction protocol just eats at me .... especially for the price it sells for. All of those unsupported I/O patchpoints. Oh well.



So there it is. Thanks for suffering through yet another one of my enormous posts.

L8r Sk8rs!!

(EDIT: I'm not really sure what to call what I'm refering too. The protocols I'm refiring too when I used the phrase "Surface mounted" are a combination of hard mounted pots, jacks and switches soldered directly to the pcb AND used as a mechanical pcb mount to mount the pcb in the device. So, whatever ~that manufacturing ethos~ would be refered to is what I'm speaking of. Thing is, I'd bet most folks that read this post didn't give two shytes about the phrase used to describe it for purposes of this thread, but were just danged glad to be given the information! Same with the differences in the resistor types .. which was very relevant to the overall theme of the posting. Thanks!). smile
Knights Who Say Neve
Great post, thanks!
odecahedron
stern but fair
albiedamned
Wow talk about timing. I was just futzing with my a190-2 right before I read this post. I'm pretty sure that's the Doepfer MIDI interface you pictured. It definitely has the glue holding the MIDI jack in place. The reason I was futzing with it was because it wasn't working. And what I eventually figured out was that it would work as long as I didn't fully tighten the screws holding it into the rails. If I tighten them all the way, the unit doesn't work. I'm pretty sure this means that something on the PCB is ready to go and it only takes a little stress from tightening the screws to push it over the brink.

It takes a lot of force to put the MIDI cable into the jack, and the glue is not holding very well anymore, so I've taken to sticking my finger behind that jack to hold it in place when I put the cable in (luckily I have an empty spot next to the module so I can reach behind it). But I think from now on I'm just going to leave the cable in permanently so as not to risk putting any more pressure on the unit.

I'm not exactly inspired with confidence in the unit - I think I may need to replace it ASAP before it fails completely. To paraphrase a HAL 9000, "I've just picked up a fault in the a190-2 unit. It's going to go 100% failure in 72 hours."
Dave Peck
If there have been questions and confusions when you have mentioned "surface mount" construction, that would be because the term "Surface mount" is a standard term used in the electronics manufacturing industry and it means something very different from the way you are using the term.

There are two basic types of circuit board construction:

1. Through-Hole, with lots of holes all over the board and parts that have small wires ("solder leads") sticking out of them which fit though the board holes and are soldered to the board at these holes, and

2. Surface Mount, in which the circuit board has no holes and the parts do not have solder leads. Instead the parts solder directly to the flat surface of the circuit board. Boards that use Surface Mount technology (SMT) are intended to be assembled using automated soldering equipment and usually require less board space because the parts can be a lot smaller than their through-hole equivalent.

All of the boards in your photos are "through-hole", none are "surface mount". The more accurate terms for the design elements you are discussing would be "board mounted components" and "panel mounted components".

But this is all just a question of terminology and my reply is just meant to clear up any confusion about the phrase "surface mount". It is not a criticism of your observations about different construction techniques. In fact, the design elements you are noticing are pretty important. When designing electronic products, care must be taken to assure that there is no stress placed on solder joints, regardless if the construction type is through-hole or surface mount.

In general (but not always), it's a good idea to mount a component either on the PCB or on the panel but NOT to mechanically secure it to both. This is because if you tighten a panel mounting nut on a part that is already soldered to the board, this can induce stress on the solder joints, especially if the board is already secured to the panel in other places. But that's a general guideline and there are lots of exceptions.

BTW - so why are some board through-hole and some surface mount?
The set-up cost of a SMT job is higher than the set up of a through hole production run, but the price per board and the speed of assembly can be much less, so SMT is a better choice IF you are building enough of the same board to justify the set up cost.

Quality wise, either can be done well or done poorly and either can use good quality parts or bad quality parts.
EMwhite
Funny, when I read your other posts referring to "Surface Mount" I thought the same as Dave Peck, that you were talking about the diff between thru-hole design; wider board traces, sparsely populated (but bigger and often stacked boards) vs. SMT and tiny tiny resistors, etc, extremely small traces and all that goes with it.

But aside from that, an interesting pictorial of some of the construction methodologies used by the builders. I'd like to see some of the harder to come by modules (at least in the states) like MOS-LAB, COTK, etc. I'll plan on getting a few of them myself eventually. If anybody has any photos of some close ups of these, with some observation, that would be cool to see.
ringstone
Dave Peck wrote:
BTW - so why are some board through-hole and some surface mount? The set-up cost of a SMT job is higher than the set up of a through hole production run, but the price per board and the speed of assembly can be much less, so SMT is a better choice IF you are building enough of the same board to justify the set up cost.


Also SMT can save board space since it's much smaller. As an example, SMT allows some very complex modules to be created in Euro format that still end up being a reasonable depth and size.

I agree with Dave, I think referring to this type of module construction as "surface mount" means that a lot of people will be confused as to what you're actually referring to. Perhaps "panel hardware directly mounted to PCB" is a more accurate term (albeit not as succinct).

Cheers
Blair
ndkent

surface mount

As for the actual subject. To me panel area compared to PCB area is a big factor. Quite obviously 5U formats leave much room to mount PCBs in more secure and isolated ways. I agree that anything you can pop out or stress and potentially break isn't ideal but it gets to be more of a debate when a module might be potentially need a larger panel vs say a 4HP attached by jacks
bouzoukijoe1
oh my god. thank you for this post! excellent overview of the different pot- and port-mounting techniques. I've been waiting for someone to call out modules with such cheap construction.

that third pic from the top with the glue is the worst - a complete joke. that's how my Doepfer MS-404 basically is, but without the crazy glue even. I bought it used and you should see what jingle-jangle state the MIDI port is in. a fucking joke. that thing's got more wobble than lamest dubstep track you've ever heard. I can hardly even plug a cable in properly. same with the knobs. anything that is pcb-mounted is bound to loosen after a couple of years.

and definitely props to Dotcom and Livewire which I also own - a couple of THE BEST constructed modules around. Pittsburgh and Bubblesound too.
Joe.
Panel mount dude.
causticlogic
Dave Peck wrote:
If there have been questions and confusions when you have mentioned "surface mount" construction, that would be because the term "Surface mount" is a standard term used in the electronics manufacturing industry and it means something very different from the way you are using the term.

There are two basic types of circuit board construction:

1. Through-Hole, with lots of holes all over the board and parts that have small wires ("solder leads") sticking out of them which fit though the board holes and are soldered to the board at these holes, and

2. Surface Mount, in which the circuit board has no holes and the parts do not have solder leads. Instead the parts solder directly to the flat surface of the circuit board. Boards that use Surface Mount technology (SMT) are intended to be assembled using automated soldering equipment and usually require less board space because the parts can be a lot smaller than their through-hole equivalent.

All of the boards in your photos are "through-hole", none are "surface mount". The more accurate terms for the design elements you are discussing would be "board mounted components" and "panel mounted components".

But this is all just a question of terminology and my reply is just meant to clear up any confusion about the phrase "surface mount". It is not a criticism of your observations about different construction techniques. In fact, the design elements you are noticing are pretty important. When designing electronic products, care must be taken to assure that there is no stress placed on solder joints, regardless if the construction type is through-hole or surface mount.

In general (but not always), it's a good idea to mount a component either on the PCB or on the panel but NOT to mechanically secure it to both. This is because if you tighten a panel mounting nut on a part that is already soldered to the board, this can induce stress on the solder joints, especially if the board is already secured to the panel in other places. But that's a general guideline and there are lots of exceptions.

BTW - so why are some board through-hole and some surface mount?
The set-up cost of a SMT job is higher than the set up of a through hole production run, but the price per board and the speed of assembly can be much less, so SMT is a better choice IF you are building enough of the same board to justify the set up cost.

Quality wise, either can be done well or done poorly and either can use good quality parts or bad quality parts.


One way to get by the mechanically securing a component problem you present is to mechanically secure it prior to soldering, then there is no danger of solder fractures, etc.
I know you said there was exceptions but it might help some DIYers to point out those exceptions.

That said, the OP does bring up some valid points aside from the whole SMT reference. Some of the soldering I've seen is poor even at Class 1 standards.
Plus, anything without a plated through hole just kind of makes me sigh.
Of course we've had aircraft manufacturers ask us to do some pretty kludgey things on boards too. If the buyer still pays for the item, there is little reason to improve the product's quality unless there is an opportunity for increased sales.
I remember opening a TV decade ago and seeing tons of rosin flux inside of it. I don't remember seeing that in older ones.
bouzoukijoe1
hell yeah. it's all about the pots with the hex nuts.



joke:


not a joke:
sduck
Aside from the incorrect terminology (not just the surface mount thing), you have some valid points. This is something I've had to consider a bunch as I've built all my modules, and a few of them are adaptions or repanelings of existing euro and other stuff. My goal in my builds is: will the module survive a fall off a loading dock, if it's mounted in a standard cabinet that's inside a travel case? Hitting the ground at whatever angle. I build in MOTM format, and typically use sheet metal brackets that are held in place using the pots or jacks or a combination of those. So, if it hits the ground at certain angles, that bracket and pcb are going to flex - not much, but enough to cause havoc with pcb mount pots, so those are out of the question. So everything is flying wired between my panels and pcbs. I also having nagging cough causing qualms about insufficient or intermittent connections when using mta connections or the like - so I hard wire everything. Overkill? Most likely. But the units work!
ringstone
Oh yes, the OP has definitely brought up some good points. Some designs that utilise SMT jacks etc without any form of bracing are a complete disaster waiting to happen. If a through hold component is liberated from its PCB through use (and believe me, I've fixed a lot of issues like this!) it's generally fairly easy to fix, on the other hand in a lot of cases SMD issues like this can't be or aren't worth repairing.

Cheers
Blair
cbm
I would edit the original post to get rid of the term "surface mount." it will only cause confusion.
Monobass
Yeah definitely drop the term 'surface mount' in this context, I thought you were having a 'through hole' vs SMT rant before....
Joe.
ringstone wrote:
Some designs that utilise SMT jacks etc without any form of bracing are a complete disaster waiting to happen


This is a sentence that I can't understand, in a subject I'm very interested in, as someone in the process of designing modules.

Because of the poor choice of terminology by the op I have no idea what is actually being discussed by ringstone seriously, i just don't get it
ringstone
LoFi Junglist wrote:
ringstone wrote:
Some designs that utilise SMT jacks etc without any form of bracing are a complete disaster waiting to happen


This is a sentence that I can't understand, in a subject I'm very interested in, as someone in the process of designing modules.

Because of the poor choice of terminology by the op I have no idea what is actually being discussed by ringstone seriously, i just don't get it


I'll give an example:

My daughter bought one of those "pocket size" hard drives in order to back up all her artwork (which I had been nagging her to do). Within a couple of months I was pulling the hard drive out of the cheap plastic casing as the connector had worked off the PCB due to the cable being plugged/unplugged. I bought a "high quality" aluminium case (not cheap at around $40) and placed the hard drive in it. Another 3 months went by, and once again, the hard drive (well the casing that is) was dead again due to the stress of the mini-usb cable pulling the SMT socket right off the PCB. In neither case was there any mechanical bracing, the only thing holding the socket to the PCB was solder paste.

Cheers
Blair

EDIT: I realise I probably confused matters even more by mistakenly saying "jacks" rather than sockets there...
Joe.
Flux Jetson wrote:
Why do some manufacturers use Blue 1% and others use Tan 5%? To save about a half-cent per resistor. I shyte you not! A half cent. Yes, I realize it adds up a bit when a given module may have 50 resistors .. but even at 50 of them we're still talking about saving 25 bloody cents. I prefer a bit of acuracy myself. I'll pay the extra dollar or two that the 25 cents would translate too. But that's just me.


I completely disagree with the sentiment of this post.

You know what the tolerance of most inductors are right? or what the tolerance of a capacitor is? up to 50%.
Manufacturers continue to use them because there are very few applications where a precise value is critical.

Most resistors aren't used in a situation where their value is critical at all; A pullup or pulldown resistor, or current limiting resistor doesn't need to be exactly 4k7, and even a 20% resistor in those applications is going to allow the circuit to perform exactly as designed.

I find the tone of your paragraph insulting, as it has nothing to do with your "surface mount theory's" (or whatever bastardisation of terminology you want to use), and is a blatant stab at all producers, effectively demanding over-engineered products and increase in costs, with no actual benefit to the end user.
Monobass
LoFi Junglist wrote:
Most resistors aren't used in a situation where their value is critical at all; A pullup or pulldown resistor, or current limiting resistor doesn't need to be exactly 4k7, and even a 20% resistor in those applications is going to allow the circuit to perform exactly as designed.


I tried to express this but gave up, thanks smile
ultrashock
as we can see from the first pic - Doepfer indeed has 1% resistors in critical points in all his modules. Don't like hot glue? (sometimes called Chinese snivel) - just drop some superglue instead smile
NV
Definitely an interesting discourse here. A few thoughts:

  • Surface Mount refers to a very particular approach to circuit assembly, as others have pointed out. I believe the more appropriate and ubiquitous term in this context would be "panel mount."

  • 1% tolerance resistors may make something of a difference in particular areas of some designs, such as those requiring stable voltage references, but in designs which do not present practical differences in performance for a variation on the scale of millivolts you shouldn't notice any difference between 5% and 1% resistors. Doepfer and other manufacturers that utilize 5% resistors are generally cognizant of these sorts of differences and account for them, using 1% in areas of the circuit where voltage accuracy and/or stability may affect ideal performance. That being said I agree that the cost difference is minimal and it may be easier in the long run to just buy 1,000 of a 1% resistor rather than 800 of a 5% and 200 of a 1% and distributing them appropriately, however for some manufacturers the admittedly minor difference in price may become substantial enough to warrant the hassle.

  • The metal-body pots you reference in the Synthesizers.com modules are common Alpha pots - or typical "cheap" carbon element pots. These can generally be purchased for close to $1 each in small bulk, meaning about 100 pieces. A "high quality" pot in the modular world would be something along the lines of a sealed Bourns or Spectrol conductive plastic or cermet pot, which can run in the area of $10 - $20 per individual pot even in bulk pricing. The practical differences between these pots are always up for debate, but there is most definitely a difference in construction. Paul Schreiber of MOTM might be along soon to explain the extremes of pot variation and how a single pot can run into the hundreds of dollars for true precision applications.

  • The Synthesizers.com panels utilize a raised edge for reinforcement due to their very thin panels, at just 0.062" over an 8.75" length. Without that lip the panels would flex in the middle with rather minimal pressure. Likewise these panels utilize 1/4" jacks/plugs which will naturally carry more pressure on insertion than a 3.5mm plug. Contrast that with the 2mm (about 0.079") standards of euro panel thickness stretched between only 5.25" while utilizing plugs at half the size and a raised lip on a euro panel should make no practical difference in rigidity unless you are purposely attempting to stress the panels. I have some 14HP STG euro modules which do utilize a raised edge and compared to my other euro modules of 14HP or greater width I cannot detect any difference even when they are not secured to the mounting rails.


Overall I understand the spirit of your post, although in practice I've never personally encountered an issue of construction between panel components soldered directly to a PCB vs. wiring between the two. The Synthesizers.com method of wiring panel components to MTA connectors definitely does allow for ease of maintenance, although I think that approach utilizes the luxury of space intrinsic to 5U. Synthesizers.com also began with some considerable advantages, such as starting out with a large line of modules, prior manufacturing experience, and stream-lined assembly practices. Compare that to a one-man upstart Euro manufacturer who only has the means to get one design into production before considering further steps and the differences in assembly become a greater issue. Soldering panel components to a PCB is faster and reduces many of the common errors during manufacture, which for a very small-time operation can be a dramatic difference. A number of longtime manufacturers have gone through the trouble of redesigning their line for panel-mounted PCBs for these and other reasons, and some people have argued that the additional labor behind hand wiring and through-hole assembly has contributed to the difficulties and even downfall of a few rather renowned manufacturers.

Additionally some designs require panel-mounted PCBs. An example of this would be a module which utilizes a display or readout/graph, such as the WMD Gamma Wave Source or the ModCan Touch Sequencer. In a situation such as this it can become a complicating factor to separate displays from jacks and pots, so it makes sense to include it all into one panel-sized circuit board. A benefit of this approach is that the space between components can be used for the remainder of the circuit, ultimately placing the entire circuit into one neatly contained board - the WMD GWS is another nice example of this. This can lead to an extremely compact and shallow module which can be placed in a very ergonomic enclosure.

Just as an aside I'll include a photo comparing a few different modules utilizing different methods of construction - specifically the STG Sea Devils Filter, the Synthesizers.com Q125, and the WMD Gamma Wave Source. Not meant to be a commentary on "right and wrong" and definitely not intended to look exact or professional, just a visual comparison of the various approaches.
John Noble
It's interesting to see a consumer's view of engineering.

The OP seems to be confusing rigidity with durability, and the 1% resistor rant is totally misplaced.

Engineering is all about tradeoffs. Let's see that Dotcom module (or the STG module further down the thread, for that matter) fit in a skiff, for instance.

(And yeah, lose the "surface mount" terminology.)
bouzoukijoe1
NV wrote:

Just as an aside I'll include a photo comparing a few different modules utilizing different methods of construction - specifically the STG Sea Devils Filter, the Synthesizers.com Q125, and the WMD Gamma Wave Source. Not meant to be a commentary on "right and wrong" and definitely not intended to look exact or professional, just a visual comparison of the various approaches.


wow, props also to STG Sea Devils. I didn't know that's how it was constructed. A+ for the panel mounted pot.

interesting use of the 5u bent panel technique, but it doesn't seem right to extend beyond the panel with that screw, in case there is an adjacent panel component that juts out too. I actually hate it when some modules have the power ribbon cable extending to the side and it starts encroaching on the adjacent module's space. you have to force it to butt up against the other module. I wish there were some basic euro design specs that people would follow. this is one of them (also panel depth, panel screw holes, and other tolerances).

not too crazy about the depth either, but whatever. I don't have a skiff. smile

good points by other users about the "surface mount" terminology as that confused me at first too, but as I read and looked at the pics I kind of understood what the OP meant (at least as far as the pot/jack issues). probably better to switch to panel-mounted, pcb-mounted.

I think the component specification as far as the electronics engineering aspect (resistors, etc) should be a different thread actually. has nothing to do with usability/physical durability/wear and tear. that was a bit out of place. would be great if this thread was relegated to just usability/durability and "non-electronics" issues per se.
daverj
Flux Jetson wrote:
I've been throwing that phrase around a LOT since I joined up. A few people have asked me wtf I'm yakking about. So I figured I'd set up a post discussing this exact construction method and the goods/bads about it's use.


I suspect that the "wtf I'm yakking about" confusion in many cases comes from the fact that you are using a term with a well established meaning in a way that does not apply to the problems you are trying to explain.

If you went into an automotive forum and complained that you didn't like the steering wheels on your new car, and wished to replace them, there would be a lot of confusion, since even though the front left and right wheels on the outside of the car are used for steering the car, they are not what people in that forum call steering wheels.

Likewise here "surface mount" has a well established meaning. Pots and jacks that are mounted directly to a circuit board are known as "board mount" or "PCB mount" parts. Even the term "panel mount" can be confusing since it relates to whether or not they attach to a panel or poke through it without attaching. You can have a pot that is both PCB mounted and panel mounted.

None of the pictures above in this thread show surface mounted pots or jacks, which are board mounted jacks whose pins do not go through holes in the board. Their pins rest on the surface.



One point that I didn't see mentioned is that surface mount components (the small ones, like chips and resistors) are much smaller and lighter than through hole equivalents. This means that boards using them, which also use PCB mounted jacks and pots, are much smaller and lighter, so do not see stresses on the boards the same as heavier boards with PCB mounted parts do when they are shaken or bumped.

And when those boards are parallel to the panel so they are supported by a grid of components, not just a single line of components, those also are much stronger and put up with much higher stresses with no issues.
dslocum
Also, regarding 1% vs 5% part tolerances...

When I started out in my little enterprise, 1% resistors didn't make sense for the first couple "non-critical" MU modules, but as my production increased and new, more demanding modules required a few 1% values, I decided that any resistors I'd buy would be 1% in order to avoid "double" stocking the same value in two different tolerances. hmmm.....

Another thing that I did not see mentioned, is that the 1% resistors are typically "metal-film" type, and the 5% are "carbon-film". Besides the tolerance differences, there's the benefit of temperature stability. I think this may have been above infered a couple times, but never really stated. Also, there are tiny advantages with thermal noise with the metal film variety.

Try building a VCO with 5% carbon-film resistors, and measure the stability! screaming goo yo
John Noble
dslocum wrote:
I decided that any resistors I'd buy would be 1% in order to avoid "double" stocking the same value in two different tolerances. hmmm.....


I don't keep anything but 1% on hand for the same reasons. The extra cost is trivial these days, especially in the context of a synth module.

Quote:
Try building a VCO with 5% carbon-film resistors, and measure the stability! screaming goo yo


That's part of the reason why Moog 901s destroy everything in their path: they live and breathe!

One of the things that reputedly keeps them from drifting faster than they do, however, is the use of half watt resistors. They have more thermal mass and so don't react as fast to temperature changes.
dslocum
John Noble wrote:
One of the things that reputedly keeps them from drifting faster than they do, however, is the use of half watt resistors. They have more thermal mass and so don't react as fast to temperature changes.


Pot the whole bloody thing in epoxy !!! lol
7hz
IIRC Macbeth uses 5% carbon resistors in his modules because they sound better. IMHO using 1% metal film everywhere is one of the reasons modern analogs sound different from vintage ones.

The usual 5% carbon resistors are not 'carbon film', but in fact pellets of carbon in a binder, the leads end in cups that attach to the ends of the pellet, and the whole thing is coated in the tan stuff.
John Noble
7hz wrote:
IIRC Macbeth uses 5% carbon resistors in his modules because they sound better. IMHO using 1% metal film everywhere is one of the reasons modern analogs sound different from vintage ones.


Except for minuscule differences in noise characteristics noted above, I think the onus is on Macbeth to present evidence on this one.

Quote:
The usual 5% carbon resistors are not 'carbon film', but in fact pellets of carbon in a binder, the leads end in cups that attach to the ends of the pellet, and the whole thing is coated in the tan stuff.


AKA "carbon composition."

I don't think you can even order those from places like Digi-Key anymore. It's all carbon film in the catalog AFAICT.
7hz
Sorry yes you are right, carbon film is standard now. I had taken apart a carbon resistor just last week to see how it was made, and it seemed to be just a slug of carbon composition, but I have just cleaned off all the tan stuff, and indeed in the middle I can see the spirals with the ceramic underneath.

RE Macbeth and sound, the proof of the pudding is in the eating, right?
John Noble
7hz wrote:
RE Macbeth and sound, the proof of the pudding is in the eating, right?


Sure, but I'm questioning causation. The reason my stereo sounds good is probably unrelated to the fact that it is perfectly aligned along a North-South axis.

Asserting that something like a resistor has a particular "sound" without presenting evidence gets the alarm bells ringing.
kdjupdal
Flux Jetson wrote:
In the case of the Dot Com modules, plastic cased jacks are used, but if you look closely the plastic case has ZERO to do with the actual construction of the jack. The barrel is all metal and it is a direct mechanical connection to the tip connector as well as thge solder terminals. These jacks are used due to their UBER- Buttery-Smooth insert/extract feeling. Patchcords just slide right in and pull right out with little cussing.


Have they started to use these smooth jacks on all dotcom modules now? I´ve only got two sequencer modules that use these jacks. All the other modules have another kind of jacks that are really tight and hard to insert and take out a patch cord from. Its the only thing that annoys me with the dotcom module construction.
dslocum
John Noble wrote:
7hz wrote:
RE Macbeth and sound, the proof of the pudding is in the eating, right?


Sure, but I'm questioning causation. The reason my stereo sounds good is probably unrelated to the fact that it is perfectly aligned along a North-South axis.

Asserting that something like a resistor has a particular "sound" without presenting evidence gets the alarm bells ringing.


This is strictly conjecture, as in "brain fart / no proof", but it would be interesting to compare the impedance characteristics of a 1/2w carbon comp with both a 1/4w carbon and metal film. That MIGHT be where the potential for diffrences lies.
7hz
dslocum wrote:
John Noble wrote:
7hz wrote:
RE Macbeth and sound, the proof of the pudding is in the eating, right?


Sure, but I'm questioning causation. The reason my stereo sounds good is probably unrelated to the fact that it is perfectly aligned along a North-South axis.

Asserting that something like a resistor has a particular "sound" without presenting evidence gets the alarm bells ringing.


This is strictly conjecture, as in "brain fart / no proof", but it would be interesting to compare the impedance characteristics of a 1/2w carbon comp with both a 1/4w carbon and metal film. That MIGHT be where the potential for diffrences lies.


Going by his latest post on FB, it is carbon composition he uses.
Chopper28
I agree with most of the OP and would point out that by these criteria, the first modules to go in the skip would probably be those awful original Moogs. You can put them in my skip if you want hihi
Flux Jetson
Ok ... thanks for the replies. I think after being told about .. what .. roughly ten times that I used the wrong terminalogy I FRIGGIN GET IT ALREADY!!!! Haahaahaa!!! I think I got it after the first mention to be honest.

Note to those that made ~that~ point .. y'best tell the marketing guys at Moog that they too are using the wrong terms. I had a lengthy telecon with one of thier "sales techs" about the LP and he used ~those terms~ all throughout the telecon in reference to panel mounted vs PCB mounted control elements and I/O components.

That phrase isn't entirely accurate either ... in effect ... EVERYTHING is at some point "panel mounted" .. is it not? I'm speaking of the controls now.

RE: the "1% vs 5%" resistors ... I can't disagree with the detractors of my opinons more. If you take a particular subcircuit that uses a bunch of 5% Rs in it, and let's say all of them happen to be 5% in the negative ... that adds up to a subcircuit that is far from design. This is a matter of opinion folks, and mine just happens to differ from those that say it's not relevant. In overall quality concerns, use of this "ethos" only betrays the manufacturers overall outlook of design vs cost consciousness.

The idea that using all 1% Rs is "ripping off the consumer" is kinda silly. The added ... what? ... .30 cents is ripping of the consumer? C'mon now .. really?

And about how some parts have no choice but to be PCB mounted ... yea, that's an over obvious given. I mean, you gotta figure that the full floating method can't ALWAYS be used (displays, etc..). I didn't mean to imply that ~everything MUST be full-floating if the module is to be considered ~good~. But thanks for pointing that out. I suppose others most lilkely needed that pointed out, so good on ya! thumbs up

I buy modular partly because of the sense of "a better grade of quality" and the general ethos of "best there is going". When I see little cost saving efforts like saving .30 cents on cheaper resistors, plastic barreled jacks that are used as mounting points to mount the PCB to the module's chassis, all I see is a company that is more concerned about my money then it is about putting out the best product they can come up with.

And comparing Alpha pots to Bourns pots ... well .. perhaps Bourns ARE better than Alphas, but the overall design concept of the "panel mounted" pots that use metal barrels" was my point. Bourns pots are a LOT more expensive, and cannot be cleaned. This is copied directly from the Dot Com website ...

"We use commercial grade jacks which work great. All modules are COMPLETELY calibrated, tested, and burned in before shipping. We use open frame pots because we do not believe that using industrial grade sealed pots can be justified by the cost. We have 30-year old equipment with open frame pots that are not scratchy, and we have 30 year old equipment with sealed pots that do. Open frame pots can be cleaned if needed, sealed can not. If you have a scratchy pot in 10 years, send it to me and I'll replace it for FREE. "

I complied to the many demands that I rename the thread and correct/edit each use of the phrase "surface mount".

Now then, on the the actual subject. The use of 5% resistors is a tell tale of a few things. One of which is that the manufacturer is willing to spend a few cents on the off chance that what they're doing makes *just that much better* a product. And, in the end, that thought process actually proves to be LESS EXPENSIVE in the manner of not having to double-stock inventory (as was nicely pointed out by a few folks in the thread). If I can be made to see that a manufacturer carries nothing but 1% Rs it shows me that they ~think~ a little on their own, and don't follow "popular standards" of manufacturing. To me there is a difference between putting 50 cents worth of tighter tolerance parts in a module and still keeping the retail price within reason, instead of using a sealed pot that is going to make (let's say) a VCO cost 40% more! The return on the investment on my end seems higher in the first case. (Is paying a hundred more bux for a VCO because it has Bourns pots as good an investment as spending 50 cents more for tighter tolerance Rs?).

My reasons for posting the things I posted had NOTHING to do with "the absolute BEST QUALITY POSSIBLE. It was a comparison of manufacturing ethos and mindset. If you want to get into the argument of "best there is" then any of the stuff I offered for visual reference would not comply. But "the best there is" wasn't my intention .. my intentions were to offer a small bit of information about subtle differences between design and engineering differences that most folks have no ideas about.

To those that got something positive out of this thread, thanks a lot for acknowleging my effort and I hope I was able to pass along some useful info. Please excuse my dumassedness about the incorrect manufacturing term. In any case, I hope I was of some help.

So then, thread name changed and I tried to clear up the phrase misusages. Just call me punkin head!! smile HAAHAAHAA!!! thumbs up

Sorry about the length of this reply, however I'm attending to a number of posts with it, so please consider that.
Flux Jetson
albiedamned wrote:
Wow talk about timing. I was just futzing with my a190-2 right before I read this post. I'm pretty sure that's the Doepfer MIDI interface you pictured. It definitely has the glue holding the MIDI jack in place. The reason I was futzing with it was because it wasn't working. And what I eventually figured out was that it would work as long as I didn't fully tighten the screws holding it into the rails. If I tighten them all the way, the unit doesn't work. I'm pretty sure this means that something on the PCB is ready to go and it only takes a little stress from tightening the screws to push it over the brink.

It takes a lot of force to put the MIDI cable into the jack, and the glue is not holding very well anymore, so I've taken to sticking my finger behind that jack to hold it in place when I put the cable in (luckily I have an empty spot next to the module so I can reach behind it). But I think from now on I'm just going to leave the cable in permanently so as not to risk putting any more pressure on the unit.

I'm not exactly inspired with confidence in the unit - I think I may need to replace it ASAP before it fails completely. To paraphrase a HAL 9000, "I've just picked up a fault in the a190-2 unit. It's going to go 100% failure in 72 hours."


There's a more than fair chance that PCB mounted control components combined with lesser grade components are ~flexing~ when the module is torqued into position in the synth chassis. This is but one of the negatives of this entire design philosphy. Solder it all to the PCB, and use those parts to also secure the PCB to the panel. Apply a bit of torque when tightening the module into position and any "cold joints", or any cracks in plastic components quickly show themselves in the form of intermittent failures.

Full floating PCBs and remotely mounted controls and I/O prevent a ton of issues like this. Yes .. I ~get~ that miniaturization of circuitry and efficient use of space allows for teensy tiny little modules that are so popular is necessary to create said teensy tiny modules that are so popular. I suppose popular miniaturized modularity comes at a cost though! Compromises everywhere.
Flux Jetson
I think it would be best if I "check out" of this thread before my attitude gets worse than it already is. I can tell that I'm getting personally involved, and when that happens things can turn to shit. So rather than allowing my emotions and ego to become involved, I think I need to walk away. I really don't have anything else to add to the thread's purpose anyhow.

Not "stomping off and slamming the door behind me" so to speak. Not even a little bit. I just know when it's time to split! I honestly do wish to sincerely thank the technically educated for correcting my phrasing error. Any time a person assumes they "know it all" ya better just go right back to bed!! If it's one thing I know for sure .. it's that I don't know shyte!

So thanks again for educating me ... after all, wasn't education the entire point of this thread?

thumbs up
cbm
Flux Jetson wrote:
Note to those that made ~that~ point .. y'best tell the marketing guys at Moog that they too are using the wrong terms. I had a lengthy telecon with one of thier "sales techs" about the LP and he used ~those terms~ all throughout the telecon in reference to panel mounted vs PCB mounted control elements and I/O components.

Note that you shouldn't take the word of marketing guys when talking about technical subjects.


Quote:
RE: the "1% vs 5%" resistors ... I can't disagree with the detractors of my opinons more. If you take a particular subcircuit that uses a bunch of 5% Rs in it, and let's say all of them happen to be 5% in the negative ... that adds up to a subcircuit that is far from design. This is a matter of opinion folks, and mine just happens to differ from those that say it's not relevant. In overall quality concerns, use of this "ethos" only betrays the manufacturers overall outlook of design vs cost consciousness.

Design is all about trade-offs. One of the first things that engineers learn to do is understand the ramifications of different tolerances. If one is designing a circuit that uses caps with 20% tolerance, there may not be a need to put in 1% resistors. Dealing with component tolerances of a circuit is just part of the design process. Pullups, current limiting resistors, etc. really don't care about the exact value, within a certain range.

By your logic, .1% resistors are better still, and hey! there are .01% resistors available, too. There are just so many places in most designs where using 1% resistors just doesn't matter.

There are resistors with different thermal coefficients, too. Should you always use resistors with the smallest thermal drift? Again, many times it doesn't matter.

I don't like this "cargo cult" approach to engineering analysis. I use a mixture of 1% and 5% resistors, some 1/4 watt, some 1/8 watt, as well as various surface mount resistors. I also use resistor packs that have a 2% tolerance (although the matching within one resistor pack is usually fairly tight.) Decisions about what to use are made based on the circuit at hand. This is all part of basic engineering discipline.


Quote:
The use of 5% resistors is a tell tale of a few things. One of which is that the manufacturer is willing to spend a few cents on the off chance that what they're doing makes *just that much better* a product.

But if the engineer has done their job right, they will have put 1% (or tighter) where it matters, and 5% where it doesn't. To insist that 1% should be used everywhere shows a lack of understanding of circuit design.
paults
Quote:
To insist that 1% should be used everywhere shows a lack of understanding of circuit design


It has nothing whatsoever to do with "circuit design". It has to do with stocking/ordering/inventory.

In the past (say 25 years ago) there was significant cost difference between 1% and 5% resistors, even in huge (say 100,000pcs) quantity. I know of this first-hand, having designed precision analog filters for telephony and for DVM front ends. Therefore, cost was the driving. factor.

But the resistor manufacturers grew tired of having 2 seperate process lines (carbon and metal film). The one really good thing carbon comp resistors were good for (lightning strikes and current surges) were replaced by the IRC/Corning metal oxide resistor (around 1981).

Hewlett-Packard started the "1% everywhere" trend, and today everyone does it. If there is any cost difference, per part it's pretty tiny. The advantage is if today you can use 5% but next week, that same value needs to be 1% (like a 1K), it's a LOT easier to have 1 part number for either spot.

In rare cases, you need 0.1% (I use them in a few MOTM modules). These are still "expensive", like 89 cents each. But you pay for the low drift more than you pay for the absolute accuracy. The low drift is what I am aiming for. BTW: 20 years ago, these exact same RN55E resistors were about $3/ea.

There are other parts that are worth "paying for" based on the circuit application:

- low-drift, low DC offset op amp in a VCO CV summer
- low-drift voltage reference on a A/D converter
- matching LED brightness at a cetain Vf
- long-life toggle switches

Everyone gets to 'vote with your pocketbook'.
dslocum
Overall, I think this has been a good, instructive, if not constructive thread. I hope no one took anything I (or anyone else for that matter) out of a purely educational context. Well, that's how I've taken it anyway.

A little story for those interested, and one that sort of helped me decide on my "1% only" stocking methods...

A couple years ago I ordered 200 100k 5% resistors from a small distributor thinking I'd get the usual brown color. When they arrived, they were blue and marked as if they were 1%. I immediately smelled something fowl and measured a few in the batch. They were out of tolerance 1% "seconds" - obvious "floor sweepings". The values were all over the place.

After calling the company and reading them the riot act, they simply stated that the color didn't really matter, just that they met the 5% spec.

Now if you were me, would YOU have some 5% resistors that would likely get mixed with the 1%er's. I didn't think so. It wasn't worth my time or energy to pursue the problem, so I ended up disposing of them.

True story, unfortunately. seriously, i just don't get it hmmm..... very frustrating
cbm
paults wrote:
Quote:
To insist that 1% should be used everywhere shows a lack of understanding of circuit design


It has nothing whatsoever to do with "circuit design". It has to do with stocking/ordering/inventory.

I agree that there may be logistical reasons to go all 1%, but the initial claim was that a circuit is "better" because it uses all 1% resistors, which is not really true.
7hz
Flux Jetson wrote:
RE: the "1% vs 5%" resistors ... I can't disagree with the detractors of my opinons more. If you take a particular subcircuit that uses a bunch of 5% Rs in it, and let's say all of them happen to be 5% in the negative ... that adds up to a subcircuit that is far from design. This is a matter of opinion folks, and mine just happens to differ from those that say it's not relevant. In overall quality concerns, use of this "ethos" only betrays the manufacturers overall outlook of design vs cost consciousness.


So you didn't read the whole thread.

My point was the manufacturer or some of the more expensive and well made modules *chooses* to use old school 5% carbon composition resistors because he prefers the sound. It has got zip all to do with cost cutting.

AFAIAA Serge also uses carbon resistors. As do all vintage synths.

Your point about panel mount is fair enough regarding the cheep end of the market with boards hanging off off their pots / sockets only, and that midi port glued in is quite bad yes! But you seem to be overthinking some of this stuff IMHO. Most supported board will survive, and this isn't a nuclear reactor we are building.
John Noble
paults wrote:
Quote:
To insist that 1% should be used everywhere shows a lack of understanding of circuit design


It has nothing whatsoever to do with "circuit design". It has to do with stocking/ordering/inventory.


We're told that no less a personage than Ken Macbeth disagrees with you.

I partially agree, but in the context of through hole parts, even Yageo charges twice as much for 1% metal film as 5% carbon film. You might not care when you have SMT digital-based modules with the smallest number of part values you can manage--and the modules retail for $300 and up. But I bet Doepfer cares a lot when they have a ton of different values going into modules that sell for $50 (based on Euro conversion prices). What retail price multiplier do you use for BOM items? You don't have to tell us, but the number is much greater than one if you want to put food on the table.

Quote:
In the past (say 25 years ago) there was significant cost difference between 1% and 5% resistors, even in huge (say 100,000pcs) quantity. I know of this first-hand, having designed precision analog filters for telephony and for DVM front ends. Therefore, cost was the driving. factor.


"Double" is what I would call "significant." hihi

Quote:
Hewlett-Packard started the "1% everywhere" trend,


HP isn't a model I would choose to follow. Maybe Agilent, but they don't make consumer products.

Quote:
and today everyone does it.


This is simply not true. I have lots of stuff with a mix of parts, and it's great gear.

Quote:
Everyone gets to 'vote with your pocketbook'.


In the context of a modular synth--a musical instrument, mind you--cost probably has some bearing on the durability of certain parts (but see Roger Arrick's editorial about pots above). Otherwise, the utility of a given module for a given artist probably has very little to do with cost above whatever minimum is required to achieve the desired function.

Let's not lose sight of the fact that a lot of the best electronic music ever produced was made with unspeakably primitive designs using low-spec parts and shoddy build quality.
NV
Your reaction is a little aggressive here, but I suppose you recognized that. I feel the majority of people posting in response to your points have been accepting of your views and just trying to provide the same "arming of knowledge" that you state is important to you yet appear to be rather critical of. Even a personal opinion is subject to discussion and information, particularly when presented on a discussion forum.

Flux Jetson wrote:
RE: the "1% vs 5%" resistors ... I can't disagree with the detractors of my opinons more. If you take a particular subcircuit that uses a bunch of 5% Rs in it, and let's say all of them happen to be 5% in the negative ... that adds up to a subcircuit that is far from design. This is a matter of opinion folks, and mine just happens to differ from those that say it's not relevant. In overall quality concerns, use of this "ethos" only betrays the manufacturers overall outlook of design vs cost consciousness.


You're not disagreeing with detractors of your opinion, you're disagreeing with the way circuits function. Some areas of design simply will not exhibit practical differences in performance with tolerances of 1%, 0.1%, or so on, so to dismiss a manufacturer as not thinking on their own and just following "popular" standards because of that decision is somewhat drastic. Personally I think the price difference between 1% and 5% has become so minimal as to make purchasing both of them to be more of a hassle than a money-saving measure, but it wasn't always this way and it's not out of the question that some manufacturers are simply holding onto their old BOMs and stocking methods since it would be an unnecessary hassle to change them. In most cases it really shouldn't sway a person's decision towards a manufacturer as if done properly it shouldn't affect a circuit's practical function.

Quote:
(Is paying a hundred more bux for a VCO because it has Bourns pots as good an investment as spending 50 cents more for tighter tolerance Rs?)


That really would depend on who you ask. Paul Schreiber who just posted above would argue that more expensive pots are a sound investment since he utilizes them, whereas Roger Arrick would disagree. Synthesizers.com actually started around a similar time frame as MOTM and was in part inspired by Paul's efforts. Taken from Synthesizers.com's "About" page:

Paul decided to do his own thing and we went in sightly different directions (we're talking power supply connectors and details like that). Paul chose the route of analog perfection, industrial grade pots, and gradual product introduction. I choose a more vintage design style, solid wood cabinets, and focused on making the products very cost-effective so many people could afford them.

This leads somewhat into the next point:

Quote:
I buy modular partly because of the sense of "a better grade of quality" and the general ethos of "best there is going". When I see little cost saving efforts like saving .30 cents on cheaper resistors, plastic barreled jacks that are used as mounting points to mount the PCB to the module's chassis, all I see is a company that is more concerned about my money then it is about putting out the best product they can come up with.


Wouldn't you say that Roger chose cost saving efforts in his designs in order to make them more accessible? Roger himself states just that in the quote I posted above, saying he "focused on making the products very cost-effective." A particularly stringent engineer could point out countless cost-saving measures even in the most expensive designs. Synthesizers.com modules commonly utilize TL07x chips in the heart of the audio path - Roger could have used an LME or OP chip for increased bandwidth or even a THAT chip in some designs, but he chose the TL07x as a "good enough" cost saving measure. The same could be said about using 0.1% resistors over 1% - using all 0.1% resistors would make for an extremely precise circuit but designers make the decision that 1% is good enough and go with it, much like they do with 5% resistors. If 0.1% resistors cost exactly the same as 1% they would probably just go with 0.1% and be done with it, but the expense keeps them from doing so - another cost saving measure from manufacturers more concerned about money than the "best" designs possible. For the record, Synthesizers.com only recently began using the more expensive plastic-barrel jacks you referenced earlier. As you can see in the photo in my posting they used "cheap" jacks for the majority of their existence. The module pictured there was built in 2009 I believe, with Synthesizers.com starting in the late 90s.

It's all trade offs, and stepping into a debate over them leads to some slippery slopes. Where's the line? Cwejman might tell you that a VCA using a TL07x over a THAT is a hefty cut corner and others might argue that using any IC over a discrete design is a compromise in quality. An amp builder might tell you that including a PCB at all rather than point-to-point wiring would negatively impact the resulting signal. My point is that it's important to judge a circuit off how it works within your system for your purposes and not off hypothetical what-ifs and stringent measurements. If a module with 5% resistors does exactly what you expect of it, then shouldn't that be enough? If a module with panel components soldered directly to the PCB has never broken a joint, then is it a problem?

Anyhow, just my attempt at keeping things constructive.
Why Adapter
This is the most interesting and informative threads I have eve read on an internet forum.

Regarding the pic of those SMT jacks: "Oh, Trouble! Glad you could make it! C'MON IN!!!"
CJ Miller
What is difficult about surface mount components is that when the temperature dips into the 180-200 degree range is that my parts are likely to slide right off the PCB like hot cookies. When there is the additional support of wires there is more of a chance that the parts will still be in place when the circuit cools. (hides)
steffie268
Thank you for such an educational post...
I have next to no knowledge whatsoever of electrics or electronics and I was wondering whether or not it is necessary for potentiometers to be mounted directly on the pcb which apparently it isn't...
So, theoretically, I could learn to "solve" all the issues I have with eurorack and do "my own thing" - I mean, how does one engage in some frantic knob-fiddling when everything is covered in patch cables? And why all these flimsy little control knobs that are always too close to one another?
My "solution" would be to dismember the modules I need/want, remove all the potentiometers from the pcb's, make a new panel with personalised layout, mount the pots on the panel, hard wire to the pcb's...and rather than having patch cables ruin all the fun, why not have some ins and outs hard wired into each other and use switches or a (hard wired) matrix mixer to allow for the necessary routing options?
sduck
winterhater
some modules like mi braids have very solid feeling knobs whereas other modules have knobs with some wobble. i dont know if its really a quality issue or just a perception. it just feels more reassuring like the more rigid knobs are going to last longer.
ersatzplanet
there is another difference between 5% and 1% resistors that nobody seems to mention here. There are some 5% values that are not available as 1% resistors unless you are depending on them matching somewhere in the range of the 5% tolerance. For example You typically can't get a 1.2K resistor in 1%, you get a 1.21K instead. In fact you typically can't find any values with 120 in them in 1% - 1.2, 12, 120, 1.2k, 12k, 120k etc. Typically this also happens with 160, 180, 220, 240, 270, and many more. If exact values are what you are after and the circuit calls for one of the values not available in 1% you are not going to be as spot-on as you think anyway.

Here is a link to the common resistor values in .1%, 1%, 2% etc. - http://www.eeweb.com/toolbox/resistor-tables
cretaceousear
@steffie268 : what you want is 5U (or 4U) high modules
See "5U Format" sub forum for enlightenment
- you have a lot of Euro already I suppose.. so learn a little DIY and have at it !
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Page 1 of 3
Powered by phpBB © phpBB Group