[BUILD] SEM VCF

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

Moderators: Kent, luketeaford, lisa, Joe.

Post Reply
User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Fri Aug 18, 2017 5:09 am

Hello,

Having some trouble with this build and was hoping someone might have a few pointers. It looks like I am having a similar issue to Emalot above.

Essentially, the module works but the frequency knob only seems to have an effect from fully counter clockwise to about the 10 o'clock position. Within that band the majority of effect seems to occur between 9 and 10 o'clock.

Measuring the voltage of the pot as I sweep up towards 10 o'clock it increases from 0V in a fairly linear manner up to about 10V then stays at 10V right through to fully clockwise.

Also when using a CV to control the frequency the CV attenuator has to be set to a very minimal amount otherwise the frequency is sent beyond the narrow working band and I lose sound.

It is very frustrating because it sounds awesome within that narrow band.

Really struggling with this and don't want to start desoldering stuff until I get an idea of where I should be looking.

Any thoughts?

Your help is greatly appreciated.

User avatar
paulstone
Veteran Wiggler
Posts: 536
Joined: Mon Apr 02, 2012 1:13 pm
Location: Modena IT

Post by paulstone » Fri Aug 18, 2017 5:53 am

:hmm:
again : have you installed the AttenuVert pot ?
whitout it the filter no work fine .

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Fri Aug 18, 2017 7:38 am

Hello,

Thanks for the reply.

Both CV pots fitted as per the instructions and double/triple checked the wiring.

:despair:

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Tue Sep 05, 2017 8:37 am

Hello,

So I have been staring at this thing for some time now and still haven't found a way of essentially broadening the the very narrow sweep area. Here is a picture of what it is like for Low and High Pass filters:
Image

The area between the two red lines is where the filter knob has any effect and as you can see it is very narrow. I have swapped out most of the components, tried different FETs, checked resistor values and can't find anything wrong. I have read this whole thread a few times but still can't find a way to broaden this band.

AndrewF further up suggested this:
Problems like frequency pot range should be ironed out when you have it all working and then set the trimpots, if not it means changing the resistor coming from the freq pot's slider - no big deal at all.
I did this and the effect was to move the effective band in one direction or the other depending on whether the value was higher or lower than the original resistor. It had no effect on the width of the band.

It seems like it should be a calibration issue but none of the trimmers seem have a great enough effect.

As I said previously, using a CV I can control the sweep within the band. This requires a very small amount of CV applied and if I add too much I do get some interesting/funky gating effects particularly in the low pass when the CV takes the frequency the into the silent area.

Anybody got any thoughts? It feels like it could be a wrong resistor value somewhere but I am at a loss as to where that could be. I think I will try and reseat the trimmers next but after that I just don't know.

One other thing, just from staring at the circuit diagram, I noticed that two of the transistors in the group of 4 are superfluous to the module when not building with CV resonance. They are the two closer to the power input and are both marked with a $. In the build doc and throughout this thread I noticed they are seen as required for the non CV resonance build but from what I can tell in the circuit diagram that they aren't needed. Has this been overlooked or am I just showing my ignorance?

User avatar
Mr_Mojo_Risin
Learning to Wiggle
Posts: 34
Joined: Thu Feb 26, 2015 3:37 pm
Location: East Coast - USA
Contact:

Post by Mr_Mojo_Risin » Tue Sep 05, 2017 3:02 pm

milton_segretti wrote:Hello,
So I have been staring at this thing for some time now and still haven't found a way of essentially broadening the the very narrow sweep area.
.
.
Since R23 and R26 make up part of the Frequency control section, I'll need to double-check with you that the those 150K resistors were replaced with 120K resistors? :hmm:

Re: Transistors
I have all six installed (T2-T3-T4-T5-Q1-Q2) even though two are silkscreened with a "$" symbol.
I paid more attention to the assembly manual snapshots of the PCB component placements rather than the "$" symbols on the PCB.

Image

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Wed Sep 06, 2017 2:23 am

Hello,

Thank you for your reply.

Both 120K resistors are installed.

I have done the trace cut and reversed the correct transistor.

Also just for reference the attenuverter is calibrated and seems to be working with zero in the centre as expected.

Regarding the transistors, I have all 6 installed too, it was just something I noticed while pouring over the schematic. Behaviour is the same whether those two are installed or not.

User avatar
roglok
Super Deluxe Wiggler
Posts: 2454
Joined: Fri Dec 30, 2011 3:49 pm
Location: germany

Post by roglok » Wed Sep 06, 2017 11:53 am

are you sure that the cutoff frequency pot is wired/placed correctly? this almost sounds like the ground reference for the voltage divider might be missing...

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Thu Sep 07, 2017 3:01 am

Hello,

Thank you for your reply.

One of the things I tried was replacing the pot. I only had a 100K B but that made no difference anyway. I also replaced the wiring to the board at that point too. So I would say that area is quite sound.

I'll have another poke around in the area where the wires are attached though to see if I can spot something on the board.

User avatar
roglok
Super Deluxe Wiggler
Posts: 2454
Joined: Fri Dec 30, 2011 3:49 pm
Location: germany

Post by roglok » Thu Sep 07, 2017 3:18 am

which board revision is that? can you share some pictures of the build?

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Thu Sep 07, 2017 3:56 am

I'm not sure. The revision number is under a capacitor. I'll pull it and have a look tonight.

I'll also post some pictures. Be prepared though, it is now very ugly as it has taken a beating from the amount of desoldering, reflowing etc. I have been doing.

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Fri Sep 08, 2017 2:58 am

Hello,

So here are the images. As I said its all a bit of a mess now. I lost a few pads with swapping out the 2N4302 so many times. I had real bother with them as they were old stock with a blob top format. On one of them the plastic had perished and it wasn't working. I have since got these ones in the image and I have also got some 2N5458 which all work fine.

I have swapped out all the IC chips for different ones and also the transistors. I haven't put back the other pair yet but as I said they don't seem make any difference anyway.

The lemcos are new. I originally had plain disc capacitors.

Some of the resistor soldering is a bit iffy as I pulled a few to check their values.

I had some bother with the trimpots. I think the wires were such a tight fit to the holes that it was difficult to get the capillary action and have the solder flow through properly.

Image
Image

User avatar
roglok
Super Deluxe Wiggler
Posts: 2454
Joined: Fri Dec 30, 2011 3:49 pm
Location: germany

Post by roglok » Fri Sep 08, 2017 3:17 am

one thought: this being a later revision board (green colour):

are you sure that the PNP transistor needs to be flipped? would be odd since it was a major bug in rev.1 boards and should've been fixed for the re-run...

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Fri Sep 08, 2017 3:33 am

It was a concern and I tried both directions.

I think you can see in the image though that the track still goes to the two 15K resistors. If it had been changed I think either it would be going to the other leg or the silkscreen would be facing the other way. Does that make sense?

User avatar
roglok
Super Deluxe Wiggler
Posts: 2454
Joined: Fri Dec 30, 2011 3:49 pm
Location: germany

Post by roglok » Fri Sep 08, 2017 5:08 am

milton_segretti wrote:It was a concern and I tried both directions.

I think you can see in the image though that the track still goes to the two 15K resistors. If it had been changed I think either it would be going to the other leg or the silkscreen would be facing the other way. Does that make sense?
yep makes sense - the PNP needs to be flipped on your PCB, too. it really doesn't help that the documentation for this project is so poor. it seems like nobody knows how many board revisions exist and which build docs correspond to which board...

if i were you i'd ask for an image of the PCB layout for your specific revision and trace back the schematics with a marker (that's what i did when i discovered the reversed PNP)...

User avatar
roglok
Super Deluxe Wiggler
Posts: 2454
Joined: Fri Dec 30, 2011 3:49 pm
Location: germany

Post by roglok » Fri Sep 08, 2017 11:27 am

As requested earlier, here's a video of my build and the sweep range you should expect:

[video][/video]



Edit: in case you are wondering about the MODE CV input - i've added a crossfader to my build in order to put the variable (NOTCH) output under voltage control.

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Tue Sep 12, 2017 4:48 am

Thats delightful and a gorgeous looking synth!

I had a good think and realised I couldn't be entirely sure of the provenance of the CA3080 chips I had installed. I had previously had LM3080s installed and the module still didn't work but behaved differently. This made me think that if the module is working properly everywhere else then it had to be something to do with my purchasing decisions.

I bought them all a while back, before I had much experience in buying older/obsolete chips. Well thats my excuse anyway.

So the long and the short of it is that I have bought some different CA3080s from a more reputable source and bingo! Seems to be working perfectly.

Thanks everyone for your help. I have learned a valuable lesson.

User avatar
roglok
Super Deluxe Wiggler
Posts: 2454
Joined: Fri Dec 30, 2011 3:49 pm
Location: germany

Post by roglok » Tue Sep 12, 2017 4:59 am

milton_segretti wrote: So the long and the short of it is that I have bought some different CA3080s from a more reputable source and bingo! Seems to be working perfectly.
Fantastic! This thought crossed my mind the other day and I planned to post about it here, but forgot. These chips have their price - whenever they show up at giveaway prices, I'd be extremely wary.

Do the defective chips look like they were sanded and re-lasered? Apart from reports about blacktopped CA3080 (which may be factory rejects and work OK), there's also this article: http://huebnerie.de/blog/2011/01/beware ... e-ca3080e/ unfortunately the author has only posted pics of the package but not of the die...

User avatar
milton_segretti
Learning to Wiggle
Posts: 12
Joined: Sun Apr 30, 2017 5:29 am
Location: Scotland

Post by milton_segretti » Tue Sep 12, 2017 5:20 am

The tops do look exceptionally smooth and the edges looked very square which would suggest being sanded.

Live and learn.

User avatar
Supervillain
Common Wiggler
Posts: 107
Joined: Sun Sep 25, 2016 2:37 am
Location: Lyon / France

Post by Supervillain » Tue Sep 12, 2017 10:17 pm

I built mine with part supplied in the synthcube's kit.
Must i give a last try and find some replacement CA3080s? Or do the ones from the kit 100% sure?

User avatar
synthcube
you will be assimilated
Posts: 2131
Joined: Wed Dec 21, 2011 6:42 pm
Location: Boston
Contact:

Post by synthcube » Tue Sep 12, 2017 10:23 pm

We pretty much have a perfect record with CA3080
But if you want to rule it out send us an email and we'll send you a loaner
facebook: https://www.facebook.com/2SynthCube

synthcube webstore www.synthcube.com/cart
music from outer space: www.musicfromouterspace.com
blacet research: www.blacet.com
MOTM DIY Analog Synthesizers: www.motmsynthesizers.com
music from outer space euro smt: www.mfoseuro.com

User avatar
Supervillain
Common Wiggler
Posts: 107
Joined: Sun Sep 25, 2016 2:37 am
Location: Lyon / France

Post by Supervillain » Tue Sep 12, 2017 10:30 pm

synthcube wrote:We pretty much have a perfect record with CA3080
But if you want to rule it out send us an email and we'll send you a loaner
Email sent! Thank you

User avatar
oneoff
Common Wiggler
Posts: 77
Joined: Tue Dec 25, 2012 12:18 am

Post by oneoff » Wed Oct 18, 2017 9:48 pm

hello everyone :party:

i have the PCB and all the parts sitting in a drawer for more then a year now...
i'd like to start the build but it's a bit confusing since i've read of potential errors/ trace cuts/ changes/ reversed transistors (?) on some of the PCBs versions, i wonder - is there one page with a clear information or a final list or maybe pictures of all the changes needed?
i have a green PCB ,bought from synthcube about a year a go, i want the basic filter w/o the res CV mod.

many thanks!!

User avatar
synthcube
you will be assimilated
Posts: 2131
Joined: Wed Dec 21, 2011 6:42 pm
Location: Boston
Contact:

Post by synthcube » Fri Oct 20, 2017 12:17 pm

while its not perfect, the build guide on our webstore (available free of course) is pretty well up to date.. hope this helps
facebook: https://www.facebook.com/2SynthCube

synthcube webstore www.synthcube.com/cart
music from outer space: www.musicfromouterspace.com
blacet research: www.blacet.com
MOTM DIY Analog Synthesizers: www.motmsynthesizers.com
music from outer space euro smt: www.mfoseuro.com

User avatar
tito_tunes
Wiggling with Experience
Posts: 315
Joined: Fri Feb 20, 2015 1:33 am
Location: San Franpsycho

Post by tito_tunes » Fri Oct 20, 2017 4:46 pm

So I've been using mine for a couple months now and the tone is really sweet, but it seems to be pretty noisy. I can faintly hear other oscs in my system which are not plugged in and radio interference. Are other people's builds like this? It keeps me from using it certain sensitive applications.
danielriera.bandcamp.com

User avatar
keninverse
Super Deluxe Wiggler
Posts: 1013
Joined: Mon Jul 26, 2010 4:12 pm
Location: mission control htx

Post by keninverse » Sat Oct 21, 2017 8:48 am

It could be a build problem or a power supply problem. I have no problem with any of my builds but I used screened cable on input and outputs but I also use a massive linear supply with heavy copper bus boards.

Post Reply

Return to “Music Tech DIY”