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

!!The Pseudo 0.5 258J Build/Mod Thread!!
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2, 3 ... 65, 66, 67  Next [all]
Author !!The Pseudo 0.5 258J Build/Mod Thread!!
J3RK
This is the last proto. The only differences on the final are a slightly larger mod area on the lower right by the output, (which means the components there were moved very slightly,) and the board will be blue.

http://home.comcast.net/~r3cogniz3r/258ish.jpg

Here's a quick list of notes. I'll arrange it better soon, and maybe get a build-thread going as well:

This is on the schematic, but I'll list them here as they appear on this board:

Saw and Square wave selection is done with a group of 5 resistors. I've labeled them "Saw/Sqr" 0 through 4 on the board.

For Saw:

0=150K
1=Inf/Open
2=Inf/Open
3=820K
4=470R

For Square:

0=Inf/Open
1=330K
2=680K
3=Inf/Open
4=2.2K

For the exponential converter, you'll see that it is a DIP14 layout. You can use a THAT340, SSM2220, MAT03, or pretty much any matched pair of PNPs, with a 2K Tempco sitting on it/them.

It is important to note that Pin 1 of the THAT 340 needs to face down (you can tell by the silk screen.) An SSM 2220 would face up, and occupy the top set of 8 pins (also indicated by the silk screen.)

I've labeled the pins as Collector, Base, and Emitter, so if you use a MAT03 or discrete transistors, you'll know how to orient them based on their data-sheets. It's really quite simple once you look at it. I'm using THAT340s on mine. There is another advantage to this. The THAT340 includes and NPN pair as well on the other end of the chip. This takes care of the NPN pair (Q7 and Q8 on the Verbos schematic.) If you use another PNP pair for expo, you'll need to place the two NPNs in the other half of the DIP14 layout. (I hope that makes sense.)

For the ICs. This seems to be fairly flexible for the most part. I would recommend using the CA3080 for the OTA. It works great with a bit of trimming and well chosen components. For the op amp labelled CA3140, you can obviously use one of these (which is what I did on this last test board,) or you can use a TL071, or equivalent single op amp. I've used both, and both work well. I've read some things that mention a nice voltage swing on the 3140, so thought I'd pop it in there just to see if it helps. I can't really tell, but this board works great. On this board, the one labelled TL071 was originally an LF356. I have some of these, but I wanted to keep most of these consistent, so chose the 071 instead. No problems here. The dual op amps are both TL072s. (though I'm sure any similar FET input duals would work fine)

Now this is where the fun part starts:

The transistors and FETs really seem to play a huge part in waveform symmetry, initial amplitude, etc. I've mentioned the expo pair and NPN pair above, so I'll mostly cover the others here. The NPN pair if not using a THAT can be anything 2N3904, the BC550 (though make sure your pins are oriented correctly,) etc.

The lone PNP (Q9 on the schematic,) can be anything really, I've used a 2N3906, and got decent results. However, I've used 2N3638 (NTE replacement,) metal cans this time, and this board functions amazingly well, and makes waveforms nearly identical if not identical to the originals that I've seen in some scope images. It seems to help the wave shape, and FM, so I've decided to keep using the more expensive and harder to find 3638s.

The FET used in the wave transitioning (Q11 on the schematic) is fairly important, but I've tried a few things with decent results. I've used a 2N3819, and while that works fine, even when trimmed, it seems to be a little bit lop-sided on the fade. Using an MPF-102 solved this, and I get a really even fade both in sound, and on the scope.

This one is VERY important if you want a nice symmetrical and sine-sounding sine wave. (Q10 on the schematic) in the sine shaper circuit. If you want the best symmetry, and a nice sine wave, I recommend a J201 FET. It really did wonders for this function. Beautiful sine wave, really symmetrical and pure. That said, I used a 2N3819 here on my first one, and it sounds just a touch more meaty (in an interesting sort of way,) just not as pure. Either sounds good, but I personally lean toward the J201. From what I understand, this is the plastic case version of the originally used 4338/4339 metal can. Easier to get, basically the same, and pretty cheap. After seeing it work, I didn't even bother trying to find the original. Note: I tried an MPF102 for this FET, and it didn't really produce a waveform that I liked. Interesting shapes on the scope, but didn't sound as good, and probably not going to give you what you expect.

Onto Caps... As you can tell from my previous image a few pages back, other than the few electrolytics, I used all styrene caps. This isn't necessary, but I happened to have a ton of each of the needed value, so I used them. For the most accurate timing, you'll want to use a styrene cap for the .0047. That said, I've used a polyester (greeny(ish)) style cap here, and I think it was partially responsible for the slightly more vintage sound that I got from my initial trial before most of this project started. I really like the sound from both, and will have a pair using each type in my new synth. This is really down to taste. You could probably use a C0G ceramic, but I haven't tried it, and since it's not really an enhancement over styrene, and not comparable to the vintage, I don't see why.

Quick explanation about "The Muck"

The muck is where you can build your mods. The main Muck is large enough to do some heavy modding in. You could try using an LM13700 OTA here maybe, or build a discrete one. (I still haven't tried this, but Jurgen Haible's discrete OTA used in his 440 filter seems a prime candidate.) After the performance of this last test version, I can't see any reason this would be an improvement though, other than maybe if all the CA3080s ran out. The uMuck section is just big enough to put a non-inverting amp (op amp and two resistors) to boost the level of the output to the usual 10V P2P range. (two 10Ks would do it, but maybe someone has a more optimum value?) The other Muck sections are really small, and may be good for Peake's expo-FM mod, or just jumpering things. I'll probably use these to hack in my Phase Lock circuit when I'm done with it. (unless I build the whole thing in the main Muck section.)

More on symmetry. This may have also played a part in the great symmetry I'm seeing on this version. I matched the diodes, and paired resistor values in the sine shaper section. Simple enough, toss them on a meter, and get them as close as possible. There are two diodes, and two or three pairs of resistors over there to check. Very little effort for something that may (I can't be 100% sure, but maybe 90%) have a good impact on things. I matched the 680Rs, the 330Ks, and the diodes.

For fun, play around with the symmetry and wave shape trims a bit, both sonically, and on a scope if you have one. You can get some pretty crazy shapes out of this thing, that you may even prefer. Or if you're building a few of them, might lend some variation. I hooked a regular pot to them, and even though it's a little hard to trim up, it worked fine. (point being, you could bring these out to the panel if you want)

I guess this ended up being pretty thorough, so maybe I'll just paste it into a doc.

Here is a link to the Electro-Music thread on this VCO:

http://electro-music.com/forum/viewtopic.php?t=16502

Links to the original layout and schematic:

http://www.simple-answer.com/DIY.html

I promised Mark that he wouldn't be getting questions from people about this, so please don't send him anything. (part of the deal for me to do this) Please post all questions here, or in the build thread that I will start up soon, and I'll be happy to answer anything I can, and there are a lot of helpful people already posting here.

I'll try to get an official parts list done soon, but really the important ones have been mentioned, and the board is completely screened with values. Not sure how much another list will help, but I'll try to get it together soon.

Once again, thanks for everyone's encouragement, purchases, and ideas. Looking forward to more projects!

Edit: (I'll add a couple things as I think of them)

Note the two 0.5W resistors in the upper right on the power rails. I don't know how critical it is that they be 0.5W. I've used 0.5 and 1W here, with zero issues. You could probably use 1/4, but I haven't tried it, and can't recommend it.

Measurements:

Board dimensions are 5.2" Height, 3.2" Width.
Mounting hole distance: 4.75" Height, 2.75" Width (center of holes)
The holes are M3 size.

Misc.

I'm looking into the possibility of a surface mount tempco and a triangle wave pad now. No promises on the tempco, but I'm fairly certain the tri-pad will be trivial. I just don't want to make too many changes before sending it out. If something went wrong with the production run... To quote Egon, "it would be bad..." Dead Banana

Updates

I've created pads for both the Triangle wave, and the Pulse wave off the point between Saw/Sqr-1 (R40 on the schematic,) and the 15K. I assume this will allow someone to build the Saw version, and pull a pulse and triangle off of it as well. One thing to consider, if you're planning to boost the signal with the non-inverting amplifier for all of these waveforms, you're going to need a quad op amp, and that won't fit in the mod area by the main output. You'd have to use the main mod-area for that. Anyway, those waveforms are available now. I'm looking into the tempco now, but still not sure about that one.

Note:
I've confirmed that the resistor from the scaling trim to ground should be 390R like the schematic, not 100K like the original press 'n' peel layout. Until I swapped this, the scaling trim did nothing. I updated the silk screen just before sending in the layout, so it will be correct on the final boards. Now with a couple of twists I've got 3 solid octaves (by ear.) I'm going to toss it on the scope in the next day or two, and see if I can get it trimmed up to get a bit more. I would be extremely happy with 4 octaves on a PNP converter design.

Another Important Note
Please be careful when orienting the PNP (2n3638/NTE129.) I based the pinout on the board on a datasheet I found, and the NTE129 ended up being different. I had to bend the base pin through the other pins to the other side. (just a simple reversal) Be sure to match what you have to the pin labels on the board, and it will work fine. The labels are correct, functionally, but the locations are different (at least with the NTE parts I've used.)
active
soooo excited!!! any builders located in NYC? we could meet up and have a drink/solder party when the boards arrive!
synthi
Taking position at the first row... applause
chriscoady
hey i live in NYC and will be building these.

i dont really know too many people that do this so i'd be game for a soldering party.

get in contact and i'll meet up-

cc
giorgio
this looks great and may just be the kick to get me to build the DIY banana synth I've been dreaming up...
boothnavy
DO IT PTVN.

I am...
photon
giorgio wrote:
... build the DIY banana synth I've been dreaming up...


Me too. My studio rig is too cumbersome for gigging. I've been planning a portable for a while.
patchdub
perhaps there is enough interest for someone to design a 5u motm panel from somewhere like front panel designs that us 5u folks could go in on and bring down the panel costs?

i know i'd be in. i don't know how to use the software yet though and don't understand all the options yet to have any ideas.

i emailed scott at bridechamber to asking if he was thinking about making one but haven't heard anything back yet.
J3RK
I do 5U designs, but I don't have anything that generates decent quality HPGL files, so I end up paying quite a bit for each engraving object. The way I usually do it is find a panel I like, and adapt it.
giorgio
photon wrote:
giorgio wrote:
... build the DIY banana synth I've been dreaming up...


Me too. My studio rig is too cumbersome for gigging. I've been planning a portable for a while.


i'm thinking 2 VCOs, probably these, 2 serge/cgs VC slopes, and 2 LPGS, gotta have a midi to cv in there for me, and maybe a utility or 2 but I'm not sure I'll even need anything else for this. all on 1 panel, all banana, all awesome.
J3RK
I've just started playing with the scaling/tracking. With only a few minutes of tweaking I've got 3 solid octaves. It sounds like there could be a few cents this way or that, but it sounds really good. From what I understand, this is more or less inherent in the OTA/PNP Pair exponential converter design. I'm quite sure more can be squeezed out with a proper calibration. I'm just doing it by ear right now. I imagine trimming up a few values slightly could help too.

The linear FM sounds incredible, and is tracking the pitch wonderfully. I'm really happy with the sound of this. I'll be working on trimming up the scaling over the next few days, and see what I can get out of it. I haven't properly coupled the tempco to the THAT chip, it's more or less just laying on the top right now. I've been reading about these converters, (particularly looking at some of Thomas Henry's designs) to see if there is anything that may help the scaling a touch. This is more octaves than I got out of my original M15s, though I'm sure they could have been fixed.

I can't wait to get more of these built. Right now I'm using my old Verbos DIY layout version and this new one together, and they're a great pair.

I'm going to add some more notes to the main post here.
J3RK
A little more tweaking, and I've got around 3.5 octaves now. I'll keep at it for the next few days. It's peanut butter jelly time!
brother303
Hi,

@ J3RK:

A (stupid?) question on the frontpanel-controls.

There are four pots on your pcb :

- Waveshape > o.k.
- FM level > o.k.

- Initial freq : Is this an attenuverter for the voltages present at the
"control in"-pad?

- Freq : This is the master-tune,right? But why 2 V/Oct. inputs?

hmmm.....

The frontpanel layouts flying around here always show 2 attenuverters labelled "CV". Where ist the second cv input? Or is the second one just an attenuated cv input wired to one of the V/Oct-pads?

seriously, i just don't get it


Thanks lot and best regards
Greg
djs
Can we get a list of what pots are being used (i.e. 100k lin, etc.) for each section?
J3RK
I am using all 100K Linear pots. The EM thread has some other values, which I can grab, and put in the notes. I've tried all 100K, all 50K, and the mixed values from EM, and all work fine.

I'm actually not sure why the V/Oct input is split up that way. That's not new in my layout, and was there on Mark's as well. Freq is the master tuning pot. I haven't completely decided yet, but I may use 10 turns on mine.

I pulled the Initial Frequency name from the schematic. It's just a bipolar FM amount control.
active
Wheee! Placing a mouser order this week. I know fingertappin and I plan on doing a night of soldeing and drinking in brooklyn, so please join us chris and anyone else!!!
brother303
Hi J3RK,

J3RK wrote:
I'm actually not sure why the V/Oct input is split up that way. That's not new in my layout, and was there on Mark's as well. Freq is the master tuning pot. I haven't completely decided yet, but I may use 10 turns on mine.

I pulled the Initial Frequency name from the schematic. It's just a bipolar FM amount control.


Thank you for the information on that!

thumbs up
J3RK
I've give this to Modularland, so there will be a proper BOM soon courtesy of him! Thanks again for this! Here's the list though, in case anyone wants to take a look before that. If you happen to notice anything odd, or have any questions, please let me know.


ICs:

2 TL072
1 CA3080
1 TL071
1 *CA3140 or TL071
1 *THAT340, SSM2220 (for exponential converter or discrete, see Transistors *)

Transistors:

2 *2N3906 or other MATCHED PNPs (for exponential converter, see ICs *)
2 *2N3904 or other MATCHED NPNs (for matched NPN pair if not using THAT340, see ICs *)
1 2N3638 or other PNP (2N3638 or NTE 129 hard to find/expensive, but worth it, otherwise other PNP)
1 MPF102 (or other similar FET, MPF102 recommended for wave form fade circuit)
1 J201 (or original 2N4339, or 2N3819 or other FET, J201/2N4339 HIGHLY recommended)

Caps:

1 .0047uF (Recommend Styrene for accuracy, Mylar for "classic", (or C0G also for accuracy, untested))
1 220pF (I used Styrene, but pretty much anything is fine.)
1 100pF (Once again, I used Styrene because it was on-hand, but anything will work here)
1 15uF Electrolytic
2 100uF Electrolytic

Trims:

1 100K Trim
1 100R Trim
1 1M Trim
1 20K Trim

Resistors:

6 100K 1%
4 10K 1%
4 15K 1%
2 68K 1%
3 330K 1% (I matched the two in the sine shaper)
2 680K 1%
4 680R 1% (I matched the two in the sine shaper)
1 820K 1%
2 3.3K 1%
1 33K 1%
2 6.8K 1%
1 6.81K 1%
1 2.49K 1%
1 4.99K 1%
1 1.5M 1%
1 1.8K 1%
1 180K 1%
1 390K 1%
1 390R 1%
1 2K PTC Tempco

1 10R (Half-Watt / One-Watt)
1 2.2R (Half-Watt / One-Watt)

Shaper Resistors:

#0 Saw-150K Sqr-None
#1 Saw-None Sqr-330K
#2 Saw-None Sqr-680K
#3 Saw-820K Sqr-None
#4 Saw-470R Sqr-2.2K

Diodes:

2 1N4148 (Matched)

Power Connector

1 MTA-156-4 (MOTM Style)
Luka
thats all i wanted - thanks

i have most stuff in stock - just need to work out if i need to order anything
iopop
PN3638 would probably work for 2N3638, right? Have not checked specs but was suggested as an alternative when searching.
Jarno
I actually searched for 2n3638 and got pn3638, specs look the same, but you will need the "non -A" version because that one has a much higher Hfe.
Could only find the pn3638a on eBay, anybody got a source for these?
Peake
J3RK wrote:


Transistors:

1 J201 (or original 2N4339, or 2N3819 or other FET, J201/2N4339 HIGHLY recommended)

Caps:

1 .0047uF (Recommend Styrene for accuracy, Mylar for classic, (or C0G also for accuracy, untested))


The J201 and 2N4339 indeed have similar but not exact idss ranges, which is the important bit for the sine shaper. I bought 35 J201 off eBay and got two with an idss in the .8x range, four in the .7x, and the rest were well below that, in the middle of its indicated range. Buchla specified the sine shaper as requiring an idss from between .7 to 1.2, "preferably" .8 to 1.1.

The 2N3819 is way higher than this, even for its minimum; you may find interesting distortion waveforms from this...

Test rig for determining idss:

http://www.forsselltech.com/media/attachments/JFET_Jig1.pdf

Ignore everything below grounding S. In the diagram, "ground" is the negative battery terminal. One milliamp meter, one 9V battery, three alligator clips, and you're in business. G and S go to the negative terminal; a single clip to both leads is fine.

Or more simply:

-Alligator clip the + VOM lead (set it to DC current) to a 9V battery's + terminal
-Clip the - VOM lead to the FET's Drain pin
-Clip the FET's G and S pins together and on the other end, to the battery's - terminal.

Note that regarding caps, Buchla used .0047uf Polycarbonate in every one of his oscillators from the 158a through the 258c, upon which this clone is based. There is no "mylar for classic" in this instance. In the Moog 901s and EMS Synthi oscillators, yes.

I've started a parts/build blog on this oscillator, for those who wish to use some vintage parts or not. Will be updating as more is learned:

http://ebolatone.blogspot.com/2011/03/vintage-parts-in-j3rks-258ish-cl one.html

Thanks J3RK!
J3RK
By mylar for classic I meant it contributes to a slightly thicker? Sound. There are a lot of variables at work here, but it did seem to make a slightly audible difference. Polycarbonate should be listed as well, but I haven't personally tried them. I'll definitely get some to try out now though w00t nice blog too! I'm planning a pair of accurate/modern ones, and a pair of vintaj/mostly vintage ones.
Rod Serling Fan Club
Glad you are doing that Peak. I am still working out my approach but I am going to do 2 double oscillators and I think I might build one of the “258s” vintage per your info and maybe go for cleaner modern (or just different) approach with the other. I like variety.
Peake
Interesting that, about the mylar. I'll try it too. RSFC, yes, variety is fun. I'm glad in this project that we get lots of choices.
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page 1, 2, 3 ... 65, 66, 67  Next [all]
Page 1 of 67
Powered by phpBB © phpBB Group