Befaco Even VCO

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Joe., lisa, luketeaford, Kent

Post Reply
User avatar
JP8
Learning to Wiggle
Posts: 16
Joined: Thu Mar 01, 2018 7:41 am
Location: Amsterdam - Netherlands

Post by JP8 » Mon Jan 14, 2019 9:28 am

Hi guys,

OK, so this is a bit strange... I finished the marvelous Even VCO kit I got from Thonk. Put it together very carefully, double checking values on all parts and minding the orientation where it matters. First power up makes noise on all outputs and the PWM is working perfectly. The octave switch and tune pots work great. Everything fine so far.

When following the calibration flow in the back of the build manual it says that I need to connect a multimeter to TP1 and ground to measure 1.500 volts. When off, adjust the ref_adj trimpot.

I can get it as low as 2.1 volts and up. No matter how many turns I give the trimmer I cannot seem to get it below a measurement of 2.1 volts.

Any ideas on what went wrong here?

Cheers,
Ricardo

User avatar
degeneratedsines
Common Wiggler
Posts: 157
Joined: Thu Dec 28, 2017 1:03 pm

Post by degeneratedsines » Mon Jan 14, 2019 5:26 pm

Maybe triple check the LM336 is a 2.5V, not a 5V. It is easy to order the wrong one since the 5V pops up earlier in the search results on mouser (I made the mistake myself).
Check again R5 is 1K 1%, the trim pot ref_adj is 10K and D1 D2 D3 D4 are 1N4148 .

User avatar
JP8
Learning to Wiggle
Posts: 16
Joined: Thu Mar 01, 2018 7:41 am
Location: Amsterdam - Netherlands

Post by JP8 » Fri Jan 18, 2019 8:51 am

degeneratedsines wrote:Maybe triple check the LM336 is a 2.5V, not a 5V. It is easy to order the wrong one since the 5V pops up earlier in the search results on mouser (I made the mistake myself).
Check again R5 is 1K 1%, the trim pot ref_adj is 10K and D1 D2 D3 D4 are 1N4148 .
Thanks for your suggestions. I checked all parts and everything checks out. Every component is at the right place and soldered properly as the attached pictures show.

I can't get through the calibration procedure like this; over 1 octave the tuning is halve a note off (C to C#) thus rendering the VCO useless... So I really want to get this solved. Any help would be great.

Cheers,
Ricardo

[/img]Image
ImageImage

User avatar
degeneratedsines
Common Wiggler
Posts: 157
Joined: Thu Dec 28, 2017 1:03 pm

Post by degeneratedsines » Fri Jan 18, 2019 4:27 pm

degeneratedsines wrote:Maybe triple check the LM336 is a 2.5V, not a 5V.
Oops I actually meant the opposite, it should be 5V. Yours is good.
What does your voltmeter read on the middle pin of the LM336? It should be 5V (V_REF on schematics). If not, what are the min & max values you can get messing with the ref_adj trimmer?
Check that you also get 5V here
Image

User avatar
JP8
Learning to Wiggle
Posts: 16
Joined: Thu Mar 01, 2018 7:41 am
Location: Amsterdam - Netherlands

Post by JP8 » Sat Jan 19, 2019 5:29 am

9
degeneratedsines wrote:
degeneratedsines wrote:Maybe triple check the LM336 is a 2.5V, not a 5V.
Oops I actually meant the opposite, it should be 5V. Yours is good.
What does your voltmeter read on the middle pin of the LM336? It should be 5V (V_REF on schematics). If not, what are the min & max values you can get messing with the ref_adj trimmer?
Check that you also get 5V here
Hi,

Thank you so much for your suggestions! I really hope we'll be able to solve this issue and get this VCO to work properly.

I checked the middel pin on the transistor (IC1) and it gives an output of 9.6 volts instead of 5. At the other two suggested points on the control board I also measure 9.6 volts. So I think we're getting closer to the problem here. Could the transistor be faulty?

Also, when messing with the REF-ADJ trimmer I get a minimum of 2.1 and a maximum of 2.9 volts. So it's range seems to be working just fine but the input voltage must be too high.

What do I do next?

Cheers,
Ricardo

Edit: I got the latest schematics from Manu (Befaco) and if I'm reading it correctly there should be 2.451 volts instead of 5. Could it be they changed the specs and enclosed the wrong transistor in the kit?
Image

User avatar
degeneratedsines
Common Wiggler
Posts: 157
Joined: Thu Dec 28, 2017 1:03 pm

Post by degeneratedsines » Sat Jan 19, 2019 10:22 am

Also, when messing with the REF-ADJ trimmer I get a minimum of 2.1 and a maximum of 2.9 volts. So it's range seems to be working just fine but the input voltage must be too high.
Yes. I suppose those measures are from T_P_1. Which means in your case:
T_P_1 = 2.1v -> V_REF = ~7V
T_P_1 = 2.9V -> V_REF = ~9.6V
if I'm reading it correctly there should be 2.451 volts instead of 5.
In fact, no. If you look in the schematics, pin 1 is minus (-) connected to ground, pin 2 is plus (+) connected to V_REF, pin 3 is ADJ connected to the trimmer ~2.451V. So according to LM336 datasheet the pinout is like this:

Image

Now, I did measure everything in the V_REF section on mine. Can you please do the same measures and report here?

R5
Left leg: 11.92V
Right leg: 5.00V

D3
Anode: 5.00V
Cathode: 4.45V

D4
Anode: 4.45V
Cathode: 3.91V

ref_adj
Pin 1: 3.91V
Pin 2: 2.33V
Pin 3: 1.10V

D1
Anode: 1.10V
Cathode: 0.54V

D2
Anode: 0.54V
Cathode: 0.00V

LM336
Pin 1: 0.00V
Pin 2: 2.33V
Pin 3: 5.00V

Also, good to hear that Manu from Befaco is helping you ;)

User avatar
JP8
Learning to Wiggle
Posts: 16
Joined: Thu Mar 01, 2018 7:41 am
Location: Amsterdam - Netherlands

Post by JP8 » Sun Jan 20, 2019 7:25 am

Thank you so much for helping me out. I hope we’re getting there.

For reference: I measured 2.3V at TP1 when I started measuring the points below. My multimeter does only 1 decimal above 2V which will reflect in my measurements. I followed your instructions as described and this is what I came up with:

R5
Left leg: 24.8V
Right leg: 9.5V

D3
Anode: 9.6V
Cathode: 8.4V

D4
Anode: 8.3V
Cathode: 7.2V

ref_adj
Pin 1: 7.2V
Pin 2: 5.1V
Pin 3: 1.6V

D1
Anode: 1.6V
Cathode: 0.5V

D2
Anode: 0.5V
Cathode: 0.0V

LM336
Pin 1: 0.0V
Pin 2: 9.5V
Pin 3: 5.1V

I hope this clarifies some issues. It looks as if the current coming in to the circuit is already too high (almost twice the value). Could that mean there’s a resistor not working properly somewhere else on the board?

Cheers,
Ricardo

User avatar
degeneratedsines
Common Wiggler
Posts: 157
Joined: Thu Dec 28, 2017 1:03 pm

Post by degeneratedsines » Sun Jan 20, 2019 7:52 am

I am not 100% sure but I think you have a short somewhere. Can you do a continuity test and make sure -12V rail and ground are not shorted on your module?
If so, it could be a solder bridge somewhere, a faulty component or even the pcb traces, hard to say.
At least the components you measured seem ok.
What I try to do is testing for shorts early during the build and regularly as the build progress. Because I don't know a good technique to locate where the short is when the module is fully built.

User avatar
JP8
Learning to Wiggle
Posts: 16
Joined: Thu Mar 01, 2018 7:41 am
Location: Amsterdam - Netherlands

Post by JP8 » Sun Jan 20, 2019 9:04 am

Hi,

I always check for shorts on my build before connecting it to power. I did check the VCO and (since you suggested it) checked it again just now. There are no shorts on the power header.

I checked for solder blobs and other traces of bad soldering but I found none (I took some close-up photo's of my work). If one thing I'm a pretty good solder-er... :hihi: With regards to the solder job I found this to be a fairly easy build since the pads are quite large and generously spaced. I also followed the instructions very carefully. The only thing I did different is the placing of the ferrite beads as they did not fit on the component side of the board. I only used the preselected components that came with the kit.

It really puzzles me how the expected voltage is almost twice the value on the measured points. Wouldn't a short make it to stop working at all?

Also, how could I find a component that's (gone) bad?

Image Image

User avatar
degeneratedsines
Common Wiggler
Posts: 157
Joined: Thu Dec 28, 2017 1:03 pm

Post by degeneratedsines » Sun Jan 20, 2019 10:08 am

JP8 wrote:Wouldn't a short make it to stop working at all?
I think so. It would probably blow a fuse on the PSU and/or some components on the module itself. But it never happened to me.
JP8 wrote:Also, how could I find a component that's (gone) bad?
I don't have a technique for that, except using a component tester out of circuit. If it was a short, it could have been a leaking capacitor for instance.
Did you check the voltages on the busboard connector? Is your PSU delivering +12V, -12V and 0V?
Maybe try using a different power cable?
Also check voltage around D6, D7 and ferrites, to see if they have fancy values or not.
I am not sure what can cause the 24V.

It is also probably better to make a dedicated thread in the DIY section. Some experienced electronic people won't even look at the general forum.

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Sun Jan 20, 2019 8:34 pm

I am not hugely knowledgeable about electronics, and I don't have this module, yet (it's in the mail). But are you sure about that 24.8V on R5? Because I don't see how that makes any sense. From the schematics R5 should connect directly to the +12V bus. If you were measuring from there to the -12V bus you'd see 24V but the voltage to ground should be +12 as degeneratedsines found. If you have 24V to ground anywhere in your system then... well, that's just wrong.

User avatar
JP8
Learning to Wiggle
Posts: 16
Joined: Thu Mar 01, 2018 7:41 am
Location: Amsterdam - Netherlands

Post by JP8 » Tue Jan 29, 2019 11:58 am

Just when I was about to give up and buy the 'happy ending kit' from Befaco (turn in your project and get a guaranteed working module in return) it occurred to me that there was one thing I had't tried yet: changing the power cable. After all what could be wrong with it...?

Well that's just stupid but it did the trick! I now have a fully functional module and I was able to calibrate it perfectly over 7 octaves.

I really have no idea what could have gone wrong with the power cable and the bus power. In my book it would not have been possible to measure the previous readings I got. But then again, I'm an electronics nitwit who was just following the instruction manual...

Thanks for all your efforts to help me out. In the end the solutions seemed to be a simple one check the power cable. From now on that will be the first step I take into troubleshooting anything.

Cheers,
Ricardo

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Tue Jan 29, 2019 12:52 pm

Excellent news! In my day job I do experimental physics, which involves electronics modules shoved into racks and connected to one another with cables (sound familiar?), and one thing I've learned over the years is if something looks wrong, check your cables and connections.

For instance if your neutrinos are going faster than light: https://www.livescience.com/18603-error ... rinos.html

Makes sense the same principle would apply in modular synthesis.

User avatar
degeneratedsines
Common Wiggler
Posts: 157
Joined: Thu Dec 28, 2017 1:03 pm

Post by degeneratedsines » Wed Jan 30, 2019 8:43 am

I am glad you where able to find the issue :tu:

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Sun Feb 10, 2019 5:39 pm

I'm now up to the last steps of building this VCO and I've run up against a problem. See photo. The octave select switch is decidedly taller than the jacks. With the front panel flush against the switch the jacks are well short of flush. Am I misunderstanding something? The instructions emphasize the importance of getting everything flush to the panel. Should I just "float" the jacks above the circuit board with the pins barely reaching the other side and solder them that way?

Image

User avatar
mutedial
Learning to Wiggle
Posts: 48
Joined: Sat Sep 03, 2016 4:16 am
Location: Tokyo

Post by mutedial » Sun Feb 10, 2019 7:14 pm

rsholmes wrote: Should I just "float" the jacks above the circuit board with the pins barely reaching the other side and solder them that way?
Yes that’s the way. I would hold the board and pcb firmly and flip it over to let the jacks be flush with the panel, then put nuts on the jacks and solder. IIRC the manual has a suggested order for doing the nuts (pots then jacks)

User avatar
sduck
experimental use of gravity
Posts: 13614
Joined: Wed Feb 18, 2009 6:03 pm
Location: Vortepexaion, TN, USA

Post by sduck » Sun Feb 10, 2019 10:14 pm

Moved to the DIY forum as this is a build thread.
flickr cloud of sound touyube NOT A MODERATOR ANYMORE

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Mon Feb 11, 2019 8:35 pm

Thanks, got an email from Befaco confirming that.

I finished soldering, plugged it in, seemed to work! I calibrated it following the flowchart with no difficulty.

Then I put it in my rack and... the V/Oct isn't tracking.

The octave knob still tracks, 16' to 1/8' (32' and 1/16' seem to be beyond my tuner's ability to cope) within a few cents. But at the V/Oct inputs (either one), going 0 to +1 V, it's 20 cents short of an octave.

[Edited to add:] I've verified 1.50V at TP1, and that successive contacts of the octave switch differ by 0.50V.

Any ideas?

(Everything else seems fine!)

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Tue Feb 12, 2019 8:55 pm

rsholmes wrote:the V/Oct isn't tracking.

The octave knob still tracks, 16' to 1/8' (32' and 1/16' seem to be beyond my tuner's ability to cope) within a few cents. But at the V/Oct inputs (either one), going 0 to +1 V, it's 20 cents short of an octave.

[Edited to add:] I've verified 1.50V at TP1, and that successive contacts of the octave switch differ by 0.50V.

Any ideas?

(Everything else seems fine!)
Anyone know what the purpose of R115 is? It's a 1K in series with R121, 50K. I don't see why it wouldn't just be 50K between the octave selector and V_OCT_IN, to convert the switch's 0.5V steps into the same current as 1V at the V/Oct jacks with their 100K resistors.

Haven't found the tracking problem's cause yet.

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Wed Feb 13, 2019 3:17 am

so, heres my theory: R115 is there to compensate for the usual 1k that some CV sources have. but, its the wrong size for this (should be 500ohms), and your CV source might not have that 1k in it. you could try bypassing R115 and see if your CV tracks better.
openmusiclabs.com

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Wed Feb 13, 2019 8:16 am

Thanks, my CV source is a Moog Mother-32; would that have the 1k? Anyone know?

Edit to add:

And wouldn't this go in the wrong direction anyhow? If R115 is too large the current from the octave switch will be too small, and calibrating on the octave switch will result in larger response, so that you'd have less than 1V/Oct at the CV inputs, correct? But I have >1V/Oct.

So maybe I should verify R115 is not too small for some reason. Or shorted somehow. I'll do that later.

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Wed Feb 13, 2019 12:28 pm

no, if youve calibrated with an extra 1k in there, then a signal without an extra 1k will produce a larger current, and therefore a larger voltage at the ouput, and a few more octaves per volt than expected. you can just jumper the 1k and see if it fixes it.
openmusiclabs.com

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Wed Feb 13, 2019 12:36 pm

But that isn't the problem, I get less than one octave per volt.

User avatar
guest
Super Deluxe Wiggler
Posts: 4186
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Wed Feb 13, 2019 3:16 pm

well, that is odd. here is a way to test it:

1. put your multimeter at the output IC100A (pin1), and make sure the octave switch still reads 0.5V increments.

2. put your multimeter on the CV input jack, and measure the input voltage for various octaves, and make sure its exactly 1V/octave. does it drift by the same amount across the range?

a drift of 20cents is an error of 1/60, which is pretty close to 1k:50k, maybe just a coincidence?
openmusiclabs.com

User avatar
rsholmes
Learning to Wiggle
Posts: 38
Joined: Sat Dec 01, 2018 5:08 pm
Location: Syracuse, NY
Contact:

Post by rsholmes » Fri Feb 15, 2019 8:38 pm

Here's the good news: I may have built the VCO to function perfectly on the first attempt with no debugging necessary!

The bad news: There is a tracking problem, but not (apparently) with the VCO.

Oh, and before going on, I should confess my remarks above about R115 were based on looking at the schematic. Had I looked at the board I would have remembered R115 is no longer used but is replaced with a jumper! So R115 is definitely not the problem.

What does seem to be the problem is that I'm using a Moog Mother-32 to supply the CV, and it does a lousy job of it. Per my voltmeter the keyboard CV it puts out is -2.99 V at C1 and +3.98 at C8 when not plugged into the VCO, but when I plug into the VCO, it goes from -2.96 to +3.95. That's just about exactly enough to account for the mistracking. Then I went looking in the Moog forum and found several threads on the subject of which this was one:

https://forum.moogmusic.com/viewtopic.php?f=19&t=25783

tl;dr: I'm not the only one. This seems to be a misfeature of the Mother-32. Suggested solution there is to run it through a buffered mult, of which I have none, so that's probably my next module.

Post Reply

Return to “Music Tech DIY”