State variable filter part of circuit just won't sweep

[patrick398]

I've been banging my head against the wall all night with this one.
Just had some PCBs back, built the circuit, and everything works except the damn state variable filter won't sweep. The schematic is a bit messy because i thought i'd be the only one having to look at it. I have the circuit on the breadboard at the moment too and everything is working perfectly. I've been through the breadboard twice comparing it to the PCB and can't see any difference. Voltages on the TL074 all check out, all sitting at half supply except power pins.

It's a weird circuit but lots of fun. Input goes into first op amp which applies some gain and splits the signal. One path goes into a 567 ring mod type thing. Out of that into the state variable filter. A STOMPLFO shifts the frequency of the ring mod via a transistor, and also (should) sweep the frequency of the filter about.

I've been over the filter part of the circuit a hundred times and can't see anything wrong in the schematic or the layout. It's basically a Sabro SVF with some value adjustments and a voice mod.
I've audio probed it and signal goes in at pin 3 of TL074, out at 1, then disappears at 6 and 7. That's got me stumped.

I'm just not sure where to look next. Any suggestions?

Full schem:


Filter section:

[patrick398]

Just built another one to see if it was a build error but has exactly the same issue, so must be a schematic/layout error. I can't for the life of me see it!

[Digital Larry]

What's the resistance measured across the LDR elements?  Any possibility the LDR's LED polarity is backwards?  What voltage are you getting across the LDR's LED?  Can you confirm that current is going through it?

[patrick398]

Both LEDs are firing, i can see them lighting (they are heatshrinked though) up according to the LFOs output.
The resistance across the LDRs is also shifting but way less than i'd expect. With the LFO on squarewave the resistance goes from tens of ohms to hundreds of ohms accordingly. What i'd expect is for it to range from hundreds of ohms up to around 20K as dictated by the 220R and 22K resistors in series and parallel with it.
Interestingly the one on the breadboard is measuring from 2k to 200k...that's even more confusing. I thought those two resistors essentially set minimum and maximum spread.

Also, the 5v reg and stomplfo are getting hot...not burning hot, but hotter than i'd like them to be.
I've taken all other chips off one of the pcbs so all that's on there now is the state variable filter and LFO section. Guitar is going straight into pin 3 of tl074 and coming out of 8. It's behaving pretty much the same way, so the issue is hopefully specific to this section and not a layout problem concerning the rest of the circuit

[Digital Larry]

Well, I think you have a place to look.  If they don't ever get over a couple hundred ohms then I wouldn't expect much action.

Try removing one of the LDR leads from the circuit to measure the resistance.  The other things (for example, the output of the previous op amp and the input of the one that the LDR forms part of the input circuit) are held at Vb, so even if the LDR is at high impedance you are measuring the series 220R.

At least, that's what I think, my analog skillz are a bit rusty.

[duck_arse]

I look on your schematic and see you have signal into the inverting input of U3B. when I compare to, say, the naughtyfish, cause I have v. little idea of svf's, I see signal to non-inverting. is yours as you thought it should be?

[patrick398]

Well, I think you have a place to look.  If they don't ever get over a couple hundred ohms then I wouldn't expect much action.

Try removing one of the LDR leads from the circuit to measure the resistance.  The other things (for example, the output of the previous op amp and the input of the one that the LDR forms part of the input circuit) are held at Vb, so even if the LDR is at high impedance you are measuring the series 220R.

At least, that's what I think, my analog skillz are a bit rusty.

I removed both optos, and replaced them, this time only soldering in the LED sides. Measuring the LDRs now i still only get between 100-700 ohms which is so bizarre. These LDRs normally range from 100-10M or something like that. The LEDs are plenty bright enough too

I look on your schematic and see you have signal into the inverting input of U3B. when I compare to, say, the naughtyfish, cause I have v. little idea of svf's, I see signal to non-inverting. is yours as you thought it should be?

Yeah i'm going off the Sabrotone schematic for the mutron, topology is the same just with some value differences

[Digital Larry]

I removed both optos, and replaced them, this time only soldering in the LED sides. Measuring the LDRs now i still only get between 100-700 ohms which is so bizarre. These LDRs normally range from 100-10M or something like that. The LEDs are plenty bright enough too

Well, resistance goes down with light.  Are you sure the LEDs are going "off" enough?  I couldn't really figure out what is driving them.

DL

[patrick398]

So i put in a couple of 50k trimmers instead of the LDRs, nothing. But when i removed one of them, i could sweep the frequency manually. I can do this with either of the trimmers, but not when they're both wired in...i'm going to do some sleuthing with the pcb layout

The LEDs are driven by the STOMPLFO output, they look plenty bright. The stomp output is driving three leds and also going into a transistor for the ring mod section, i feel like i might be pushing it a bit, might explain why the 5v reg is getting hot

EDIT: Scratch that, both trimmers are now working. Progress...but very confusing progress

[duck_arse]

you do have a seperate CLR for each of those leds, and they just don't appear on the circuit diagram. that's right, isn't it.

[patrick398]

you do have a seperate CLR for each of those leds, and they just don't appear on the circuit diagram. that's right, isn't it.

You're right. I'll have to check how it is on the breadboard, maybe i found that CLRs weren't needed as the stomplfo doesn't put out enough current but i've just looked at other schematics where i use the stomplfo and i did use a CLR. That might explain why the chip is getting hot but i can't see how that would stop the LDRs resistance sweeping enough.

Just before i left today i hooked up the opto from the breadboard to the PCB, so the LED is being driven by the breadboard circuit but the LDR is on the PCB and that seemed to work fine. I'm going to have to think about that for a while, doesn't seem to make much sense.

Thanks for the CLR catch Duck! I'll check the breadboard tomorrow

[Kipper4]

Hey Pat.
So now the filter is sweeping albeit manually. Great news.

As for the vactrol. It is odd that you are getting such a small variation in resistance.
How is the Stomplfo depth tied ?   To ground or +5V or adjustable. Might explain this.

Could you share the Sabro schematic you were working from. I'm think I know it just not sure.
Is the blend set for a filter dominant output?

I'd expect a fair bit of heat from a regulator but the Stomplfo shouldn't get too hot.
After all we're expecting the regulator to dissipate half of the +V. Some of it is bound to generate heat.

Hope you get it sorted, seems like you've already gone to a lot of trouble if you have had pcbs made.

Rich

I hope you get it sorted.

[Kipper4]

Oops didn't see the full schematic before I made the last post.
I see now how the depth is tied.

I note your using 1k series to Stomplfo. Don't know if it makes a difference but I normally use 10k for those duties.


Found this Sabrotone SVF.
http://www.sabrotone.com/wp-content/uploads/2012/12/EnvelopeSVF.png

Here lol

https://www.diystompboxes.com/smfforum/index.php?topic=118545.msg1104002#msg1104002

[patrick398]

Oops didn't see the full schematic before I made the last post.
I see now how the depth is tied.

I note your using 1k series to Stomplfo. Don't know if it makes a difference but I normally use 10k for those duties.


Found this Sabrotone SVF.
http://www.sabrotone.com/wp-content/uploads/2012/12/EnvelopeSVF.png

Here lol

https://www.diystompboxes.com/smfforum/index.php?topic=118545.msg1104002#msg1104002

Hey Rich, good to hear from you!
Yeah you're right about the 1k to 5v, it's a 10k pot ordinarily so i don't know why i used a 1k there, it could be an error when i was drawing up the schematic, meant to type 10k but missed the 0. I'll change that, might not make a difference but good practice anyway.

hahaha that old thread! I think that was not long after i joined the forum! I actually had a look at the thread today to see if i'd made a similar mistake, that would have been embarrassing all these years on ha

[patrick398]

So the regulator being in backwards probably wasn't helping things. It always seem to be editing that footprint in Kicad and never seem to get it right. Also changed the 1k depth resistor to 10k, CLRs on the LEDs and there was a separate issue in the first gain stage where i forgot to include a cap to ground.
All working fine now, thanks as always guys

[ElectricDruid]

Good work, glad you got it sorted.

For the record, the StompLFO draws about 5mA without the output heavily loaded. If you start piling LEDs on the output, that'd go up obviously, but the chip is supposed to be ok up to 20mA (per pin - with a something-or-other max for the whole device). That's a reasonable number of modern LEDs...or one red one from the 1980s!
Either way, no way it should be getting even noticeably warm. That's a big red flag.

T.