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

OBX voice card (with non-custom CEM3310 equivalent circuits)
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page Previous  1, 2, 3 ... 67, 68, 69 [all]
Author OBX voice card (with non-custom CEM3310 equivalent circuits)
jvq
Some updates about how my crowbx is behaving.

VCO2 keeps going out of tune on two out of four voice cards. It takes several days/weeks for the tuning to go bust. I've just replaced the tune trimmers (maybe I turned them too much?) on one voice's VCO2, waiting to see if the tuning will stick now. So far so good.

Something else: sometimes the voice on/off and square/saw switches stop working for all voices. When this happens I can turn the on/off switches to 'off' and the voices all keep making sound. When I set VCO2 to 'off' it goes to 'half' instead. And I think it locks on 'saw' for both VCO's. Other switches keep working. Unfortunately it needs some warm-up time to reach this state (and turning off/on fixes it) so it will be hard to diagnose. I wonder if it's a problem with the 4011N quad nand IC on the host board.
oldcrow
Heya folks, a quick note. I am working on getting 8-voice stuff finished among my other projects. Since 3310 EGs are going to be made again I am going to try a 3310 edition of the crOwBX voice card. It will not replace my non-custom EG version, but it does simplify the build somewhat:



More later,

Crow
/**/
oldcrow
The 4011 is either defective or the +5V going to it has an issue. VCO tuning issues are usually due to a bad timing cap or bad trimmer. --Crow

jvq wrote:
Some updates about how my crowbx is behaving.

VCO2 keeps going out of tune on two out of four voice cards. It takes several days/weeks for the tuning to go bust. I've just replaced the tune trimmers (maybe I turned them too much?) on one voice's VCO2, waiting to see if the tuning will stick now. So far so good.

Something else: sometimes the voice on/off and square/saw switches stop working for all voices. When this happens I can turn the on/off switches to 'off' and the voices all keep making sound. When I set VCO2 to 'off' it goes to 'half' instead. And I think it locks on 'saw' for both VCO's. Other switches keep working. Unfortunately it needs some warm-up time to reach this state (and turning off/on fixes it) so it will be hard to diagnose. I wonder if it's a problem with the 4011N quad nand IC on the host board.
jvq
Thanks oldcrow!

I can rule out "bad trimmer" now. Replacing the trimmers fixed one of the bad VCO's, the other one is still misbehaving. It keeps drifting. Slow enough that I can sometimes get it in tune, but then within 5 minutes it is out of tune again.

I put thermal goop between all the LS318's and their tempco resistors but I wonder if the thermal bond got damaged on this VCO somehow.

I'll try replacing the 4011 on the host board and the timing cap on the remaining bad VCO2.
PWM
jvq, did you use styroflex timing caps? These are quite easily overheated.
jvq
PWM: I used WIMA FKS2 http://nl.mouser.com/search/ProductDetail.aspx?r=505-FKS2D011001AKC00

I don't know if they're quite the right type, I just decided to use the WIMA brand and picked parts by price and trying to match specs from the BOM.
PWM
I really can't tell you if these are adequate for this application. I tend to use 1% or 0.5% tolerance types. These Wima's are 5% and up.

I don't think this cap is the problem though, you would've encountered it in all VCOs otherwise.
Also, I don't think the tempco is the problem. If I understand correctly, the VCO drops in frequency and stays there? Does it still track, after dropping? If the tempco or it's mounting is faulty I would suspect fluctuations in the frequency, not a drop.

You could build a connector with wires. This way you can troubleshoot the Voice PCB on your workbench and measure values in a working voice.
jvq
PWM: so I had two bad VCO's. One of them was dropping but then staying where it was. I replaced the trimmers on that one and now it's steady.

The other bad VCO is drifting. I replaced its trimmers too but no change. When I play that voice it sounds like it has a different tuning problem from the first one. The fact that changing the trimmers didn't help backs that up.

It's almost as if the tuning of the bad VCO is "floating". It stays in one place for a while, and then it just drifts somewhere else. In the course of seconds / minutes.
PWM
Ah, ok. I understand now. Had the two VCOs mixed up. My apologies.
Drifting could very well be a tempco problem.
ultravox
jvq, I had a VCO stability/tuning issue on a different synth (TTSH) that turned out to be a grounding issue. Two of the three VCO's would drift and the 1V/Oct scaling would not align. I corrected the mainboard grounding and everything aligned and tracked perfectly.

This may not be your issue but something to keep in mind.
jvq
Thanks ultravox!

My best guess is that I did a bad job soldering the tempco underneath the matched transistor pair and that flexing the voice board repeatedly has damaged one of the tempco solder joints. Or that I didn't use enough heat transfer paste.

I'm not an experienced solderer, especially not with SMT. So I certainly won't rule out a bad solder joint.

I've been getting a little frustrated with this tuning problem so I'm going to take it a little slow and focus on another project for a moment.

I've also built and installed the Hexinverter MIDI-CV converter into my CrowBX. Sadly, it's a source of further frustration. It's linearity is poor compared to my clunky outboard MIDI-CV converter which makes it harder to tune all the voices. A so-so CV source is not the sort of obstacle I want when trying to fix a tuning problem. smile I think I will just build my own converter. It was encouraging to see Oberling's project at https://github.com/oberling/AVRMIDI-CV and I've been getting some MCU programming experience of my own recently.

All in due time. smile
jimfowler
I strongly recommend Peter's Midi2CV and MidiGal boards. Easy to build, easy to use, more flexible than Hexinverter. https://midisizer.com/
jimfowler
oldcrow wrote:
...but you can wire 100K pots that have CW lug to +15V, CCW lug to -15V and the wiper to the leftmost pin of each FT header to provide fine tune options.

--Crow
/**/


This nets you a 2 octave "fine tune" per VCO. So far I kinda like it as you can do some neat stuff in unison mode and still have the ability to slightly detune each VCO...but getting a 5-10 cent variation is kinda fiddly. Somebody who knows more about this than me can tell you where to put a resistor to relegate the adjustment to a proper fine tune range.
jvq
Thanks jimfowler, the midisizer.com boards look really nice!

They use the same DAC family as the Hexinverter MIDI-CV though: MCP4922 / MCP4822 12-bit. I'd like to build something that uses a 16-bit DAC such as the AD5676R.

Detuning is nice/awesome, but given the choice I'd rather have a more accurate MIDI-CV DAC that I can detune via software (e.g. a MIDI CC) when needed than a less accurate DAC that makes my job harder while tuning the CrowBX expo converters. smile

Plus, building this myself is fun.
livewire
FWIW, I am interested in obtaining four more of the bare "old style rev B" Crowbx boards. Have not seen any of these for sale in a long time. If I have to buy the "kit" with all the other support boards, so be it. I am restoring an old OB-X and still need boards.
randel
It only took me a few years, but I managed to completely finish 4 voice cards!

Imagine my delight when I plopped the first one in my OB-X, and it worked immediately. I almost fell out of my chair when the next 3 worked as well!

Three cheers to Old Crow for making this project a reality, and for faithfully recreating the Obie goodness of the original voice cards in a tasteful blue package! hyper

I have a few parts on order to complete the upper tray board, and soon my project will be complete.

A few notes for those expanding a vintage OB-X:

You may wish to use a top-adjusting trimmer for T7 and T8.

The upper tray board requires longer male Molex connectors if you wish to use the factory 1/4" standoffs. I realized this after soldering a bajillion connectors, and am hoping that 3/16" standoffs will work.

I have another 8 voice boards 95% complete, and will delve into the big carrier board project once the snow begins to fall...



[/img]
randel
Last night I finally got to hear my OB-X in all of it's 8 voice glory - 4 original cards and 4 crOwBX cards + upper tray carrier.

This represents lots of hours on the bench, but my gosh, it is worth it.

And hey, after crimping 90+ Molex pins for the connectors I have mastery of a new skill!

I'll post some pics and audio clips in the very near future, but in my opinion, the crOwBX cards - and Old Crow - absolutely rock.

I do have a few questions:

1. The Portamento times of the new cards are markedly longer than those of the original cards. While the difference in portamento rates is part of the Obie's charm, what I'm hearing with some of the new cards is not, I suspect, what Tom intended. I recall seeing some discussion about portamento and resistor matching earlier in the thread. Has anyone made progress with this, or added a trimmer pot?

2. The voice card calibration procedure for the EGs calls for adjusting pots to achieve a 61Kohm reading in several spots in the circuit. On each of my boards, the adjustment range is around 30K to 37K. Has anyone else seen this?

Thanks
jvq
A quick update. I fixed the second problematic VCO2. I bit the bullet and removed (destroyed) the LS318 IC to examine the solder joints on the tempco underneath; they looked not-great. Reflowed the tempco solder joints and installed a new LS318. Bingo.

Lesson: order some extra LS318's when building that version of the voice board (the new one announced by Crow doesn't use them?).
The Real MC
randel wrote:
1. The Portamento times of the new cards are markedly longer than those of the original cards. While the difference in portamento rates is part of the Obie's charm, what I'm hearing with some of the new cards is not, I suspect, what Tom intended. I recall seeing some discussion about portamento and resistor matching earlier in the thread. Has anyone made progress with this, or added a trimmer pot?


The reason the portamento times don't match is the OTA on each portamento circuit in the OBX not the voicecards. OTAs aren't matched in early OBX so the rates are spread out. Can't be fixed with resistor matching or trimpot - can only be improved with matching OTAs. Mine is an early one, I basically swapped out OTAs between portamento circuits and VCA on voicecards (NOT the VCF OTAs, leave them intact!). Still not matched portamento rates but close enough, and actually a cool effect.

Later OBX used matched OTAs for the portamento circuits. Still wasn't a perfect solution.

When Oberheim introduced the pinstripe OBSX, they implemented digital portamento... complete with simulation of mismatched rates. I actually like the effect.
randel
Thanks for the info on the portamento - I will investigate your suggestion.

Here are a few photos of the project near completion - note the non-conductive washers used for the standoff nuts close to the Molex pins:





oldcrow
@randel Very very very nice job sir! I will have to get the rackmount stuff finished. Guinness ftw! Guinness ftw! Dead Banana
oberling
oldcrow wrote:
Heya folks, a quick note. I am working on getting 8-voice stuff finished among my other projects. Since 3310 EGs are going to be made again I am going to try a 3310 edition of the crOwBX voice card. It will not replace my non-custom EG version, but it does simplify the build somewhat:



More later,

Crow
/**/


wow oldcrow - nice! Now I get the whole picture. THAT voice card has it all we're not worthy no more slot adapters (I see the panning and portamento circuit) and it has the filter mod already included...
Also you got rid of the old LS318 and the thermistor it seems... CA3086? SOIC?
Did you sort the portamento issue with that zener diode there?

Shame on me for prematurely ordering another 4 voice set + 8 voice backplane to expand my 4 voice system meh
The only question remains: what to do with all that left PCB space from the EG-circuitry!? Mr. Green Additional EGs? some LFO per Voice?
srfnsmy
Having problems finding the 10 and 3 pin KK-156 Molex connectors. Does anyone have a link where I could find them?

hmmm.....
MUFF WIGGLER Forum Index -> Music Tech DIY Goto page Previous  1, 2, 3 ... 67, 68, 69 [all]
Page 69 of 69
Powered by phpBB © phpBB Group