[IN STOCK] Eurorack: Discrete State-Variable VCF v1.4

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

Moderators: Kent, Joe., analogdigital, infradead, lisa, parasitk, plord, sduck

Post Reply
User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Mon Dec 07, 2015 6:48 pm

negativspace wrote:
sromanel wrote:I don't think it'll be worth the shipment price, we're talking about almost a hundred euros send+return. I've had it checked by a synth repairman, there was a faulty trimmer (6V IIRC) but that didn't change much. HP works just fine, the only problem is with notch out. He even changed all SMD components, but with no luck. Is there any measurements I should do in order to understand a little bit of what's wrong?
It's ~$15 to ship one of these things from here to Italy, I've done it more than once. Surely Italian airmail isn't $85 for a 12oz package.

Anyway... if you *do* have a HP out then there is only one possibility for the fault and that is somewhere in the SMT/notch balance pot/jack "subsystem." Try probing the CW and CCW pins of the balance pot with a patch cable just to see if there's a signal on either, and we can work backwards from there. A photo might also help, otherwise I'm just taking blind guesses. :hihi:

6V trimmer will only have an effect on the mixer section, that rail is not used on the filter side.
I know shipment itself doesn't cost that much, but believe me, customs clearance does (there's a 100% package stop rate from/to USA & Canada here, and they make you pay for the package inspection - seriously, fuck Italy). Back to technical stuff - if I got that right, I hadto feed a signal into the filter, attach one clamp of the oscilloscope to left & right pins of notch balance pot. Here's what I measured:

Sine in, fully CCW balance
Image

Sine in, fully CW balance
Image

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Mon Dec 07, 2015 7:19 pm

Here are pics of the boards

Image


Image

The board seems sticky & greasy because of isopropyl alcohol[/img]

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Mon Dec 07, 2015 7:47 pm

I guess I'm spoiled with the USPS... I think in my entire life I've had to pay import duty one time. :doh:

I presume you have an 0805 1k resistor installed on the backside of the main board, since the through-hole alternative is missing from the front. (Wouldn't explain the issue you're having in any event.)

On the scope shots... fully CW is HP side, looks like that's the one that works correctly? If that's true then it's your LP side that's broken. Is that accurate? If so, start by checking the soldering on the 47k 0805 input resistor. Possibly a cold joint there in your pic.

On the sticky stuff... I had to wash my boards about 3 times with alcohol when I was using rosin-core solder. It's always a gummy mess after the first pass. :hihi:

Jop
Wiggling with Experience
Posts: 292
Joined: Fri Dec 28, 2012 1:18 pm
Location: The Netherlands

Post by Jop » Tue Dec 08, 2015 4:40 am

With my build I experienced some slight HF rolloff (posted orginally on page 44 with some sound clips). Jason confimed this was also present on his builds and suggested to lower the three 220pF caps on the output buffers to 22pF.

That did the trick, all frequencies are there with the cutoff fully open!
Thanks Jason :tu:

So if you experience this same behavior you know what to do. 8-)

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Tue Dec 08, 2015 6:55 am

negativspace wrote:I guess I'm spoiled with the USPS... I think in my entire life I've had to pay import duty one time. :doh:

I presume you have an 0805 1k resistor installed on the backside of the main board, since the through-hole alternative is missing from the front. (Wouldn't explain the issue you're having in any event.)

On the scope shots... fully CW is HP side, looks like that's the one that works correctly? If that's true then it's your LP side that's broken. Is that accurate? If so, start by checking the soldering on the 47k 0805 input resistor. Possibly a cold joint there in your pic.

On the sticky stuff... I had to wash my boards about 3 times with alcohol when I was using rosin-core solder. It's always a gummy mess after the first pass. :hihi:
I guess I screwed something up on the last measurements. I checked again, now this is what i have both CW and CCW

Image

while this is what input signal looks like

Image

Regarding 0805 1k resistor, you're right, this is SMT section on the main board

Image

I tried resoldering all of the SMT section, but nothing changed

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Wed Dec 09, 2015 1:14 pm

Just did full continuity test on the control board, everything seems just fine :despair: And listening more carefully, it is clear to me that notch works fine, it just comes out at half volume

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Wed Dec 09, 2015 6:23 pm

Yeah, I was sitting on it overnight giving it some thought but it's clear from your last photo that both sides of the notch output are working...

There are 4 caps on the SMT section, two on the power rails and one in the feedback loop of each buffer. It's really hard to tell in the photo, but are you certain that the 0.1uF are in the correct spot as are the 22pF (0.1 on the rails, 22pF in the buffers.)

If you put a .1uF in the feedback loop of the buffer, you'll get a muted and filtered output. (From experience.)

Looking very closely at your photo it looks like you have physically thinner caps on the power rails and physically thicker caps in the feedback loop - that could be the problem. (For me, anyway, the 22pF I use are thinner than the 0.1uF. Mine are also different colors which helps a lot.)

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Wed Dec 09, 2015 7:05 pm

negativspace wrote:Yeah, I was sitting on it overnight giving it some thought but it's clear from your last photo that both sides of the notch output are working...

There are 4 caps on the SMT section, two on the power rails and one in the feedback loop of each buffer. It's really hard to tell in the photo, but are you certain that the 0.1uF are in the correct spot as are the 22pF (0.1 on the rails, 22pF in the buffers.)

If you put a .1uF in the feedback loop of the buffer, you'll get a muted and filtered output. (From experience.)

Looking very closely at your photo it looks like you have physically thinner caps on the power rails and physically thicker caps in the feedback loop - that could be the problem. (For me, anyway, the 22pF I use are thinner than the 0.1uF. Mine are also different colors which helps a lot.)
You know what? I'll drop by the electronics parts shop where I usually supply and see if they have those components in stock, and try switching one at the time until (hopefully) the problem will be gone. Do you think I should try replacing the SMT TL072 as well?

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Wed Dec 09, 2015 7:08 pm

If it's passing a clean sine like it should on both sides, no. Seems to be working fine.

You won't have done any damage to the caps if they're in the wrong spots, swapping them around should work. (If indeed that's the issue.)

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Wed Dec 09, 2015 7:44 pm

negativspace wrote:If it's passing a clean sine like it should on both sides, no. Seems to be working fine.

You won't have done any damage to the caps if they're in the wrong spots, swapping them around should work. (If indeed that's the issue.)
Yeah, but SMT components are quite delicate, and I'm rather clumsy at unsoldering :hihi: I'll see what I can do

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Wed Dec 09, 2015 9:11 pm

True... in my experience the ceramic caps can take some rework. Hope yours survive. :hihi:

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Thu Dec 10, 2015 2:02 pm

negativspace wrote:True... in my experience the ceramic caps can take some rework. Hope yours survive. :hihi:
After double checking, I can quite surely say that caps are correct (the ones "above" and "below" TL072 are 0.1 uF, while the ones left and right of it (IE the ones nest to "4") are 22 pF. I was wondering - gain is set by 47k resistors, right? Could they be the guilty ones? And what if I replace them with, say, 100k s? Should this solve the problem?

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Thu Dec 10, 2015 2:05 pm

Hmm... can you confirm that you used a 10k pot for the notch balance? I've had super-quiet notch issues twice since the SMT revision. Once was when I put 0.1uF caps in the feedback loops, once was when I used a 1M pot from the wrong pile.

A slight drop in the middle is normal but it should be full strength at both ends. (Full HP or full LP.)

Gain is set by the ratio between input and feedback resistors - all should be 47k, which means gain = 1, so it should all be operating at unity. You can change the feedback resistors to fix the levels but that'll be a band-aid masking the real issue. (Which is really starting to have me curious.)

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Thu Dec 10, 2015 2:45 pm

negativspace wrote:Hmm... can you confirm that you used a 10k pot for the notch balance? I've had super-quiet notch issues twice since the SMT revision. Once was when I put 0.1uF caps in the feedback loops, once was when I used a 1M pot from the wrong pile.

A slight drop in the middle is normal but it should be full strength at both ends. (Full HP or full LP.)

Gain is set by the ratio between input and feedback resistors - all should be 47k, which means gain = 1, so it should all be operating at unity. You can change the feedback resistors to fix the levels but that'll be a band-aid masking the real issue. (Which is really starting to have me curious.)
I'm quite sure everything's fine, knobwise

Image

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Fri Dec 11, 2015 10:59 am

Yep, perfect as can be. Hmm.

Back to thinking while I calibrate this stack of 16 CP3s...

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Sat Dec 12, 2015 6:01 pm

negativspace wrote:Yep, perfect as can be. Hmm.

Back to thinking while I calibrate this stack of 16 CP3s...
I'm wondering - what are the odds that the TL072 might be involved in the issue?

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Sun Dec 13, 2015 1:41 pm

Update:

After replacing all the 47k resistors (again), I found out one of them (parallel to 22 pF capacitor, pic related) keeps reading 29 kohms, no matter how many times I replace it, while every other resistor reads ~47 kohms. It may be a reading problem, but I suppose it's worth mentioning. Another thing: I decided to test the notch balance knob, and while one of the test points (left one, towards lp side) correctly sweeps from 0 to ~10 kohms, the other one (right, towards hp) has a rather strange behaviour. It starts at about 3kohms fully CCW, it keeps rising to about 6 kohms about halfwat, and then falls to 4.5/5 kohms fully CW. Furthermore, it reads various values as I turn the knob more or less quickly. Note that all the readings were done with the module unplugged

Image

User avatar
FingerTappin
Common Wiggler
Posts: 171
Joined: Fri Mar 11, 2011 6:03 pm
Location: New York

Post by FingerTappin » Sun Dec 13, 2015 9:05 pm

Hey negativspace,

So, I made an attempt at the retrofit circuit. I have one issue and that is my low pass output stopped working. I have checked over my stripboard layout ( seems good) and did some continuity testing (seems good).

I had not soldered in the notch pot before attempting the mod. So instead of cutting the traces, I tapped from the pads for the pot and then soldered to a solder lug pot. Could this be my problem?


Been sitting on the bench for a few months and with the holidays coming up, figured I would get it going. Thanks in advance.

User avatar
sromanel
Veteran Wiggler
Posts: 598
Joined: Wed Apr 01, 2015 7:53 am
Location: Italy

Post by sromanel » Wed Dec 16, 2015 6:33 am

Swapped the non parallel 47k to 22k, now signal reaches 8 Vpp for an incoming 10 Vpp signal. The resistor I talked of before still reads 26k though :hmm:

User avatar
medbot
Veteran Wiggler
Posts: 669
Joined: Mon Oct 28, 2013 2:44 am
Location: Denton, TX

Post by medbot » Wed Dec 16, 2015 2:00 pm

Does anyone have any experience using silver mica caps instead of the styrenes? I just ordered some polypropylene caps to try out and saw that Mouser had 1nf mica caps as well, but I didn't want to fork out the $6 to try them. :hihi: They were recommended in Moogah's SEM filter thread, that's what got me thinking. Going the other way, has anyone stuck some cheapo ceramics in there? I've got styrenes in mine now but am going to have it on the bench soon for the resister swaps, maybe I'll socket the caps and test some different things.

Jop
Wiggling with Experience
Posts: 292
Joined: Fri Dec 28, 2012 1:18 pm
Location: The Netherlands

Post by Jop » Tue Dec 29, 2015 6:01 am

Just finished my second unit and unfortunately there are some issue's :bang:

Build:

* Rev 1.4 PCB
* Wima polypropylene timing caps
* Experiment with different transistors, SA733P & SC945P correctly bend to allow the different pin layout (ECB) and matched within 2mV.
* Carefully build the unit and washed the PCB after soldering.

Calibration:

* -6V TP was easily set with the trimmer
* between + 5.2 and 5.6V on the +5 TP, reading change a bit depending on the frequency pot position.
* +TP a reading between + 3.2V and +5.5V depending on the null trimmer.
* -TP a reading between -141 mV and 19mV depending on the null trimmer.

Issues:

* Only the HP out carriers a signal, all other outputs are silent. The mix out jacks do work, although the plus jacks is much lower in volume, probably related to the calibration issue from above.

Frequency and resonance pot don't do anything.

In1/2/3 pot and jacks work properly.

What I already did is:

* Triple checked all transistors if the correct one's (PNP / NPN) where installed and matching the silkscreen.
* Replaced the IC's.
* Checked for LP signal at pin 6 of U2 without result.

I have the feeling my issues could be related to the transistors used (not looking forward replacing them all...) but I'm missing the technical background. Could someone please direct my where to look at first?

Attached some pictures.

Many thanks!

Image
Image
Image
Image
Image

Jop
Wiggling with Experience
Posts: 292
Joined: Fri Dec 28, 2012 1:18 pm
Location: The Netherlands

Post by Jop » Tue Dec 29, 2015 1:26 pm

In the end I replaced all the transistors for 3904 & 3906 and everything is good now!

The circuit came alive with the 733p's & 945p's but it was far from stable, some trannies need bending to one end to become active and all kind of artifacts could be heard. Strange things :hmm: :despair: , I din't bother further and just replaced all of them.

It could have been that the leg bending caused some shorts, I did see that on the DMM when testing them in circuit (C& B connected)...

In the end all good :sb:

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Wed Dec 30, 2015 12:09 am

Glad to hear! I like it when I sit down to read a troubleshooting question and it ends with "nevermind, I fixed it." :hihi:

Grumskiz
Wiggling with Experience
Posts: 341
Joined: Thu Apr 03, 2014 12:51 pm
Location: Germany

Post by Grumskiz » Wed Dec 30, 2015 2:43 pm

Hey, just wondering is the PCB set currently available? (Don't need a panel)

User avatar
negativspace
Manhattan Analog
Posts: 6571
Joined: Tue Apr 13, 2010 2:50 am
Location: Manhattan, KS, USA
Contact:

Post by negativspace » Wed Dec 30, 2015 11:32 pm

Yep, I recently restocked and I have a ton of these on hand again. :tu:

*goes back to building more*

Post Reply

Return to “Music Tech DIY”