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

How will Tempcos non-availabilty affect us?
MUFF WIGGLER Forum Index -> 5U Format Modules Goto page 1, 2  Next [all]
Author How will Tempcos non-availabilty affect us?
Rex Coil 7
I've never built a VCO. Let's get that out there right away. While I've no experience with building an oscillator I fully understand the use of a heater to keep the VCO transistors at a continuous temperature to create pitch stability.

Ok, fine.

Lately, I've read that Tempcos will no longer be produced beginning January 2018. From what it seems, this is a hard line end of production announcement.

That said, how will this affect analogue synth design ... and how will it affect long term upkeep, and even warranty issues?

Let's not pretend that all of our favorite oscillator manufacturers have endless bulk stock of Tempcos that will be available to keep our beloved analogue synths in a state of operation forever.

Is there a chance the templates (or "sheets" or whatever is used to produce transistors, ICs, and resistors) may be bought up and a new manufacturer will produce them (kinda like the way Intersol sold off the CA3080 masks to Rochester Electronics to produce "reissue" ICs)?

Am I concerned over nothing?

Is it time to revisit analogue oscillator designs and start looking at the next phase .... perhaps a digital "core" with everything else being analogue (or somethin' somethin' ...)? That said, why hasn't there been an IC developed that would replace the standard triangle core design, or a digital chip designed that would replace the venerable core?

Thanks!
JohnLRice
Rex Coil 7 wrote:
Lately, I've read that Tempcos will no longer be produced beginning January 2018. From what it seems, this is a hard line end of production announcement.
Do you have a link to any articles that talk about why they are being discontinued?
Rex Coil 7
JohnLRice wrote:
Rex Coil 7 wrote:
Lately, I've read that Tempcos will no longer be produced beginning January 2018. From what it seems, this is a hard line end of production announcement.
Do you have a link to any articles that talk about why they are being discontinued?
There's a thread here in MW about it ... I'll see if I can find it. I was subscribing to that thread, but after I made a jackass out of myself asking about using digital means to control pitch stability I unsubscribed. Full disclosure.

Let me see if I can locate it.
Rex Coil 7
JohnLRice wrote:
Rex Coil 7 wrote:
Lately, I've read that Tempcos will no longer be produced beginning January 2018. From what it seems, this is a hard line end of production announcement.
Do you have a link to any articles that talk about why they are being discontinued?


Link = https://www.muffwiggler.com/forum/viewtopic.php?t=190994&highlight=
JohnLRice
Rex Coil 7 wrote:
JohnLRice wrote:
Rex Coil 7 wrote:
Lately, I've read that Tempcos will no longer be produced beginning January 2018. From what it seems, this is a hard line end of production announcement.
Do you have a link to any articles that talk about why they are being discontinued?


Link = https://www.muffwiggler.com/forum/viewtopic.php?t=190994&highlight=
Thanks for the link! thumbs up
kcd06
Quote:
I made a jackass out of myself asking about using digital means to control pitch stability


Thereby committing an act of apostasy before the Analog Purists, and driven from their numbers by the True Believers bearing artisanal sticks and pelted with rocks from carefully curated collections.

Not that I've ever done that, of course...
J3RK
I haven't heard one way or the other, but I assume one can still get the large KRL/Bantry ones.

If not, there are other ways to compensate for temperature. Check out Rene Schmitz' method using multiple 10K NTCs. NTCs are still very available. I'm not sure how well this method works, and not sure how easy it is to adapt to typical designs.

Another method is the "oven" method, which uses another transistor or two to heat the main matched pair. Many VCOs use this method, and it works well from what I've read. A lot of them used the CA3046/3086, but I think I've seen it done with THAT ICs as well. This is also how the ua726 works I believe. https://www.muffwiggler.com/forum/viewtopic.php?t=153796

Another way is to use the 2164 VCA method like David Dixon, Roman Sowa, Osamu Hoshuyama etc. You use one of the VCAs for tempco. I believe L-1 did this with a THAT VCA as well, but would need to look that up.

Some of these might be able to be worked into existing designs, some may need to be designed specifically to work with one or the other.

There are quite a few threads in the DIY section about many of these.

I've started playing with the 2164 IC finally (after building a model for it in my software). I may give the VCA method a try at some point.
Synthbuilder
The 1K +3500ppm/K temp co resistors I use in my VCOs are from KRL Bantry in New Hampshire. They're still going so I don't think you should be worried.

The +3000ppm/K temp co resistors I use for less critical use are from TE Connectivity and still seem to be available too.

Tony
J3RK
Synthbuilder wrote:
The 1K +3500ppm/K temp co resistors I use in my VCOs are from KRL Bantry in New Hampshire. They're still going so I don't think you should be worried.

The +3000ppm/K temp co resistors I use for less critical use are from TE Connectivity and still seem to be available too.

Tony


Nice! I've always liked the KRL ones best. I even use them with SMT LS318/LS358 matched pairs. It looks funny since they're so big, but they work nicely. More surface area contact that way too.
Rex Coil 7
I still think it's time to reconfigure thinking on this issue. Using heat means using a sort of thermostat, which means a constantly cycling up and down heat cycle, which tosses tracking.

There's got to be another way. Some type of digital stability system that works off of a feedback system from the VCO itself. I have no idea how to implement this notion, but it just seems like there has to be another way to address this issue. Manufacturing and digital technology have jumped hills and valleys over the past decade.

Oh well .... everwhat.

On topic, going by the replies here so far, I suppose there is still some availability of needed heating components, that said I guess things will remain stable for a while yet.

Good news, right?
J3RK
The way the oven setup works, is that the matched pair is raised to a temperature above usual ambient temps, so it stays stable.

Personally, I have zero problem with well implemented digital oscillators either.

I have a feeling though that if all current tempco makers stopped all of a sudden, we'd see someone else spring up to make them again. Something like a CoolAudio, Sound Semiconductor, whoever's doing the Curtis parts, etc.

I'm all for new methods, but I don't think there's too much to worry about either.
Rex Coil 7
J3RK wrote:
The way the oven setup works, is that the matched pair is raised to a temperature above usual ambient temps, so it stays stable.

Personally, I have zero problem with well implemented digital oscillators either.

I have a feeling though that if all current tempco makers stopped all of a sudden, we'd see someone else spring up to make them again. Something like a CoolAudio, Sound Semiconductor, whoever's doing the Curtis parts, etc.

I'm all for new methods, but I don't think there's too much to worry about either.
Righteous! Thanks.
KSS
TL:DR Worry, Yes. But not about availability or options. We can still Haz Analog VCOz'ez

What we call tempcos are used widely as PTC thermistors. These are not going away anytime soon, if ever. They may become even more expensive as digital makes further inroads into the analog arena where these are critical components in military and high-end analog applications. What *is* going away are a few SMD low priced PTC thermistor products from a couple or a few manufacturers.

IOW, this seems to be an economic move by some companies to update and trim their offerings rather than a supply problem like the one which caused the loss of Polycarbonate capacitors.

A secondary result for us may be that some designs based on the Panasonic and Akaneohm SMD "Tempcos" may become hard to buy or disappear for awhile for a board respin. It seems Vishay is still going to carry their PTC SMD line so maybe this is why the others are leaving the market to them?

As Graham hinton has pointed out repeatedly, the cheap SMD tempcos aren't especially suited for us anyways. I think Ian Fritz reported seeing some self-heating issues when the Panasonic ERAS tempcos first appeared.

David Dixon will probably be by soon to remind us that in addition to the oven there is the 2164 solution he uses on many or all of his VCO designs which include the Intellijel Rubicon and Dixie.

John Simonton of PAIA had a semi-success with an earlier version of the same idea using a 13700 instead of a2164.

Precision resistor in FL, USA will make whatever you want if your moneybag is large enough. Google PTC Thermistor to see that They and KRL and the others mentioned here are only a few of the many.

All things are new again.
The easy availability enjoyed today to an amazing degree and which we never had in the 70's is coming around again, returning us to having to really search for and pay high prices for some parts. Even wait for weeks! OMG! And for the same reasons. Digital. Welcome to the machine. You will be assimilated.
Rex Coil 7
"TL:DR" yet posts an eight paragraph reply.

lol

KSS wrote:
TL:DR Worry, Yes. But not about availability or options. We can still Haz Analog VCOz'ez

What we call tempcos are used widely as PTC thermistors. These are not going away anytime soon, if ever. They may become even more expensive as digital makes further inroads into the analog arena where these are critical components in military and high-end analog applications. What *is* going away are a few SMD low priced PTC thermistor products from a couple or a few manufacturers.

IOW, this seems to be an economic move by some companies to update and trim their offerings rather than a supply problem like the one which caused the loss of Polycarbonate capacitors.

A secondary result for us may be that some designs based on the Panasonic and Akaneohm SMD "Tempcos" may become hard to buy or disappear for awhile for a board respin. It seems Vishay is still going to carry their PTC SMD line so maybe this is why the others are leaving the market to them?

As Graham hinton has pointed out repeatedly, the cheap SMD tempcos aren't especially suited for us anyways. I think Ian Fritz reported seeing some self-heating issues when the Panasonic ERAS tempcos first appeared.

David Dixon will probably be by soon to remind us that in addition to the oven there is the 2164 solution he uses on many or all of his VCO designs which include the Intellijel Rubicon and Dixie.

John Simonton of PAIA had a semi-success with an earlier version of the same idea using a 13700 instead of a2164.

Precision resistor in FL, USA will make whatever you want if your moneybag is large enough. Google PTC Thermistor to see that They and KRL and the others mentioned here are only a few of the many.

All things are new again.
The easy availability enjoyed today to an amazing degree and which we never had in the 70's is coming around again, returning us to having to really search for and pay high prices for some parts. Even wait for weeks! OMG! And for the same reasons. Digital. Welcome to the machine. You will be assimilated.
Huba-Swift
I've heard that vacuum tubes are extremely temperature stable compared to transistors. Maybe we just need to backtrack hihi
Rex Coil 7
Huba-Swift wrote:
I've heard that vacuum tubes are extremely temperature stable compared to transistors. Maybe we just need to backtrack hihi
Ya never know! The things people are working on behind closed doors can pop up at any time ...... ~wink wink~

Guinness ftw!
KSS
TL:DR used here like the abstract summary at the head of professional papers. It's not that I TL:DR, it's that if you TL:DR, then the first paragraph is enough to get the answer without details. Was the reply useful to your question?
Graham Hinton
KSS wrote:
Precision resistor in FL, USA will make whatever you want if your moneybag is large enough.


They could not make what I wanted at any price. It was through having a detailed discussion with them that I realised exactly how inaccurate wirewound tempcos are.
Rex Coil 7
KSS wrote:
TL:DR used here like the abstract summary at the head of professional papers. It's not that I TL:DR, it's that if you TL:DR, then the first paragraph is enough to get the answer without details. Was the reply useful to your question?
"TL:DR" = Too Long, Didn't Read. And yet, we're expected to read your ~long~ post?

It comes across as rude, as if to say "I don't have time nor do I care to read what others have had to say ... but I expect others to drop what they're doing and read my eight paragraph post".

Whether it was your intention or not, it's how it comes across. Not everyone is a professional who reads papers that use abstract headings. Personally, I make it a point to presume that there are people reading what I post that are not internet savvy, do not know what heavy abbreviations stand for, or know what certain "givens" mean. I attempt to remember that when I post something in open forum, I am literally addressing the entire planet. The point of posting is to inform others, and we must remember that there are many thousands of people that may read what we post, and not all of them are hip to what's up in a given topic or subject. I can imagine that I annoy the hell out of experienced members when I take the 3.4 seconds it takes me to spell out "Synthesizers.Com" at least once in a post where that is relevant ... but I do that because I wish to show respect for the person new to the modular scene that may not know what "Dot Com" refers to. That's just one example of how I'm sure my "too long to read" posts can annoy some members.

I didn't take your use of "TL:DR" as a personal annoyance, because this isn't about me, per se. It's about everyone else who posts, everyone else who lurks, everyone else that locates these posts via search engines.

If you didn't read the thread ("TL:DR"), how can you begin to know if someone hasn't already covered the ground your treading? Or are we all to just blow off what everyone else has to say and just read your post?

See how it comes off as arrogant?

Shutting up about this in 3 ... 2.. 1.
MrNezumi
Uh, Mr. 7, you've misunderstood what he said. He is saying that his first paragraph is a summary. This way if you think the following eight paragraphs are too much to bother reading, you can just read his first paragraph and get the gist of what he says. He is trying to save you time. It is a secondary use of TL:dr, but isn't that unusual.
Rex Coil 7
MrNezumi wrote:
Uh, Mr. 7, you've misunderstood what he said. He is saying that his first paragraph is a summary. This way if you think the following eight paragraphs are too much to bother reading, you can just read his first paragraph and get the gist of what he says. He is trying to save you time. It is a secondary use of TL:dr, but isn't that unusual.
Uh, member *MrNezumi, perhaps some of us aren't aware of all of these "internet speak nuances" ... keep in mind when we post we are speaking to the entire world, not just those that have certain levels of posting eloquence.

That said, use of less nuanced prose would prevent misunderstandings just like this one, would it not? I actually did some research on the "TL:DR" prefix after I read the post in question (and before I replied to that post), and it's definition is far too ambiguous and far too context based to actually be used in a manner which could/would be understood specifically and correctly. Those of us without exposure to it's use could easily take it as arrogance, just as I did.

My suggestion is to use more plain speak. Again, we speak to those of all education and experience levels when we post. Not everyone has experience with "professional documentation" or journalistic op-ed stylized nuance.

A bit of "PR" advice, take it or not.

Brian.
cornutt
Rex Coil 7 wrote:
Some type of digital stability system that works off of a feedback system from the VCO itself.


Seems like it would be a fairly easy thing to do in a DCO design. You need a reasonably good way of sensing the transistor die temperature. The algorithm can use that to adjust the counter value.

(Then again, if you have a DCO design, you don't need an expo converter at all... just read the V/oct control voltage directly and scale from that.)
Rex Coil 7
cornutt wrote:
Rex Coil 7 wrote:
Some type of digital stability system that works off of a feedback system from the VCO itself.


Seems like it would be a fairly easy thing to do in a DCO design. You need a reasonably good way of sensing the transistor die temperature. The algorithm can use that to adjust the counter value.

(Then again, if you have a DCO design, you don't need an expo converter at all... just read the V/oct control voltage directly and scale from that.)
(I highlighted member *cornutt's statement). Yup ... there's a few places that the feedback signal could be tapped from, such as a VCO mixer, or perhaps a dedicated feedback signal output coming from the VCO itself so the digital tracking ~thingy~ could monitor the VCO's pitch vs the 1v/oct CV pitch.

That way, FM in the form of vibrato or other such things wouldn't be getting in the way.

It seems to me that this notion would be applied at the MIDI-CV unit, a sortof feed back system that would measure the VCO's pitch against the 1v/oct signal. You could also use tracking "curves" that would permit the use of several curves. Meaning, if you wanted to emulate the way that old R.A.Moog modular VCOs were in tune with one another near the tuning center pitch, then as you played 2 octaves higher or lower the VCOs would begin beating faster and faster, just like old modulars did. Another "curve" could be dead-solid tuning all the way through a ten octave spread.

It almost seems as though this digital tuning/feedback/tracking system would be something within the MIDI/CV device. Required, a special "feedback signal output" from each VCO that would not be affected by pulse width modulation or FM that the MIDI-CV unit could rely on to tell the tracking system what the core pitch is.

As far as uncontrolled oscillators go, I'm not sure exactly how to implement this idea. Y'know, like a droning VCO (especially since "drone patches" are so bloody popular these days). Even though it's not being controlled by a 1v/oct signal, it still needs to be kept stable. But I'm not an engineer, either.

I have, in fact, run this idea past one or two "in good faith" engineers - at least one of which is a Muffwiggler's member - that immediately and nearly in knee jerk fashion instantly dismissed this idea. That's usually a good sign, I've learned in my lifetime that when some new idea is met with resistance, it usually means I'm on to something. I tend to fall back on the notion that some folks will tell you that they have "twenty years of experience" when in fact they have one year of experience repeated twenty times over ... um ... so to speak.

In the last ten or fifteen years, digital technology and manufacturing technology has literally LEAPT forward by hills and valleys. There's 15 year old kids doing things with Arduino chips that weren't even conceivable 2 decades prior, and to those kids what they're doing is "just kid stuff".

All of this having been said, I have to believe that there is some way to use a digital stabilization method that could be used to replace the current methods of using controlled heat to keep the transistors at a constant. Some type of internal feedback signal sent to a digital tracking system that would re-pitch the VCO in real time.

Perhaps this means a return to the VCO Driver module? A special module used to keep all of the VCOs in tune, which would also serve as the input module for 1v/oct control signals. Truly a ~modular~ way of doing things!

Funny, one of the engineers that I hit up with this idea cited the Prophet 5, saying how it took that synth over 2 minutes to "auto tune". Well sure it did .... using 1978 digital tech. The tech in a modern wrist watch has 10X the processing power that the Apollo Saturn V had ... including all of the systems in Mission Control. So I am sure that it did, in fact, take the Prophet 5 a couple of minutes to settle the tuning to track 1v/oct (or however it tracked pitched intervals back then). Geez.

At the end of the day there's got to be another way. Digital stabilizing is one method, but I think going one tiny step forward may actually be more practical.....

.... Make the entire core digital, supported by analogue ~everything else~. One little digital ~chip~ that contains the entire core, with the entire remainder of the VCO being analogue. It's still a Voltage Controlled Oscillator in that sense. I know that the DCO exists, but doesn't that only digitally control 1v/oct? Or is that actually controlling drift and basic tuning?

Clearly, I'm not the one to create such things.

But as I keep saying, in regards to keeping VCO transistors stable ... there has to be another way.
J3RK
I don't believe I've seen this done before, and I can't remember if this was discussed in some of the other tempco threads, but you can use a transistor to detect temperature, so maybe a compensation circuit could be built around that. Of course that too would be temperature dependent, so not sure.

(for some reason I think I've asked about this before in another thread)
Synthbuilder
Rex Coil 7 wrote:
It seems to me that this notion would be applied at the MIDI-CV unit, a sort of feed back system that would measure the VCO's pitch against the 1v/oct signal.


Check out the Expert Sleepers interfaces. This is exactly what they do.

http://www.expert-sleepers.co.uk/

It does require manual intervention though. You have to tell it to calibrate - which in a modular is possibly the best way of doing things since the module itself can't know when the best time to calibrate should be.

I believe the new Korg MS-20s does something similar - in between notes played they 'housekeep' the VCOs and work out scaling and absolute tuning compensation curves.

Tony
MUFF WIGGLER Forum Index -> 5U Format Modules Goto page 1, 2  Next [all]
Page 1 of 2
Powered by phpBB © phpBB Group