Vref on tube booster (pedal i designed) is ~3.5v instead of 4.5v

[snow123]

hey guys, i recently built a new revision pedal i designed called the tube booster (this time its on a PCB that i designed), and its basically a tube screamer that i turned into a clean boost and gave proper EQ controls.
Most of it works, but for some reason the Vref is around 3.5v instead of 4.5v, it used to be (when i first built it and put everything together) around 2.5v, but reflowing some of the solder joints seems to have increased Vref to 3.5v.

Schematic:

(for now, I've omitted R100 'cause it was decreasing the voltage a bit too much)

Board layout:


board pics:












(yes, i know that i used a bit too much solder in some places)

[Vivek]

Try adding a cap in series with R8

Maybe also a cap in series with R11

Reason: Your left Opamp has a DC gain due to R8 directly going to VR

Then the left Opamp will will multiply the Vref on the + pin with the gain, and will rail saturate its output to reach quite close to Vcc

Then R10 and R11 will feed this DC to the Vref, altering it.

[antonis]

Good old snow..

What's your DMM's impedance..??

P.S.
Make R4 100k (and C2 100nF) 'cause it loads Q1 buffer..
(I know you've copied from original schematic but it's never too late for improvements..)

[anotherjim]

R100 at 100R, the volt drop shouldn't be excessive. The 555/relay should be powered from after D100 and before R100, not after it. The LED should be powered after D100 too.

If VR has changed, maybe because VA has changed too? VR always wants to be half of VA.

I see no excessive load on VR. IC1a is biased by VR so its gain control connections don't load VR - they want to be the same voltage. If they did drop VR, then the +input reference will drop too, so all will be equal except closer to clipping on one side.

[antonis]

@snow: What's the purpose of a Bass cut HPF filter with corner frequencies between 0.4 and 0.006 Hz..??
(IMHO, it shouldn't work even for dogs..)

You already have two HPFs of 617 Hz (R6+R7/C5) and 867 Hz (R5/C4) respectively..

P.S.1
In case you retain BASSCUT + R8 "filter", place a series cap between R8 and VR, as vivek pointed out above..
(or you'll have DC gain in the range of X1.1 to x6.5, resulting into 5V to 29V(saturation) on IC1A hence misbias IC1B pin 5..)

P.S.2
I'd connect R8, R7 and C4 VRs to GND and place a cap between IC1A pin 1 and R10 for DC isolation or just delete R11..   

[idy]

Antonis asked the pointed question, with hidden import.

What's your DMM's impedance..??

The real question is how are you measuring your Vref. If you measure at the opamp + input you will always get a low value. But if you check the opamp output you will be reading a "buffered" voltage that won't be loaded by the meter.

One of those things that until you are aware can cause headaches.

One fun trick is to use two meters. Leave one hooked up to the output, when you touch the input with the other, they both go down.

[snow123]

Antonis asked the pointed question, with hidden import.

What's your DMM's impedance..??

The real question is how are you measuring your Vref. If you measure at the opamp + input you will always get a low value. But if you check the opamp output you will be reading a "buffered" voltage that won't be loaded by the meter.

One of those things that until you are aware can cause headaches.

One fun trick is to use two meters. Leave one hooked up to the output, when you touch the input with the other, they both go down.

im using my multimeter set at 20 DC, and im measuring vref by connecting the black lead to ground, and connecting the red lead to anything on the board that is connected to vref.

[snow123]

@snow: What's the purpose of a Bass cut HPF filter with corner frequencies between 0.4 and 0.006 Hz..??
(IMHO, it shouldn't work even for dogs..)

You already have two HPFs of 617 Hz (R6+R7/C5) and 867 Hz (R5/C4) respectively..

P.S.1
In case you retain BASSCUT + R8 "filter", place a series cap between R8 and VR, as vivek pointed out above..
(or you'll have DC gain in the range of X1.1 to x6.5, resulting into 5V to 29V(saturation) on IC1A hence misbias IC1B pin 5..)

P.S.2
I'd connect R8, R7 and C4 VRs to GND and place a cap between IC1A pin 1 and R10 for DC isolation or just delete R11..

honestly i was just throwing shit at the wall to see what worked for the bass cut control, and ended up with that.
and i have considered connecting R8, R7 and C4 to ground, but i connected them to VR to make connecting everything else to VR a bit easier.

[snow123]

R100 at 100R, the volt drop shouldn't be excessive. The 555/relay should be powered from after D100 and before R100, not after it. The LED should be powered after D100 too.

well FWIW, VA was at around 7v with R100.

[snow123]

Try adding a cap in series with R8

Maybe also a cap in series with R11

Reason: Your left Opamp has a DC gain due to R8 directly going to VR

Then the left Opamp will will multiply the Vref on the + pin with the gain, and will rail saturate its output to reach quite close to Vcc

Then R10 and R11 will feed this DC to the Vref, altering it.

what capacitor values do you suggest that i use?

[Vivek]

Go here

http://www.learningaboutelectronics.com/Articles/Low-pass-filter-calculator.php

put in value of R8
Then desired frequency of 15hz
Calculate C



Is there a good reason to have R11 ?

[m4268588]

If voltage value of Vref due to load of the DMM is doubts.

This easy way solves repeated query.

Try adding a cap in series with R8

Maybe also a cap in series with R11

Reason: Your left Opamp has a DC gain due to R8 directly going to VR

Then the left Opamp will will multiply the Vref on the + pin with the gain, and will rail saturate its output to reach quite close to Vcc

Then R10 and R11 will feed this DC to the Vref, altering it.

Do you need a ridiculous formula to estimate the "Vref offset"?
Two DC paths, R8+R6 and R11, are bit more complicated. If you can use a simulator, such a calculator is useless.

[merlinb]

R100 at 100R, the volt drop shouldn't be excessive. The 555/relay should be powered from after D100 and before R100, not after it. The LED should be powered after D100 too.

well FWIW, VA was at around 7v with R100.

Well there's your problem then. Q2, the relay, the LED... you are draining to much current through R100 which is pulling your supply down to 7V, and Vref is half that or 3.5V.

[antonis]

@snow: Could you plz elaborate the role of Q2 (and respective circuitry)..??
(if you just have copied form somewhere/sometime, ignore my question..)

P.S.
I solely ask @snow123..!!

[duck_arse]

PS antonis, you spoil our fun.


on topic, not neccessarilly this topic - I think C3 will pop when switched one end to ground one end to VR, better to connect the switch NC to VR as well. I think.

[antonis]

I think C3 will pop when switched one end to ground one end to VR, better to connect the switch NC to VR as well. I think.

Stephen, I presume you're talking about C2 (not C3)..

[merlinb]

PS antonis, you spoil our fun.
on topic, not neccessarilly this topic - I think C2 will pop when switched one end to ground one end to VR, better to connect the switch NC to VR as well. I think.

Better still, move C50 so it comes immediately after Q1, before the relay. And add a pull-down resistor for C2.

[snow123]

@snow: Could you plz elaborate the role of Q2 (and respective circuitry)..??
(if you just have copied form somewhere/sometime, ignore my question..)

P.S.
I solely ask @snow123..!!

its just a gyrator circuit for the mids control.

[snow123]

R100 at 100R, the volt drop shouldn't be excessive. The 555/relay should be powered from after D100 and before R100, not after it. The LED should be powered after D100 too.

well FWIW, VA was at around 7v with R100.

Well there's your problem then. Q2, the relay, the LED... you are draining to much current through R100 which is pulling your supply down to 7V, and Vref is half that or 3.5V.

but with R100, VA was 7v and VREF was at around 2v. without R100, VA is 9v, and VREF is around 3.5v.

[ElectricDruid]

but with R100, VA was 7v and VREF was at around 2v. without R100, VA is 9v, and VREF is around 3.5v.

That suggests to me that whatever the problem with Vref is, it's nothing to do with R100. While the circuit will work without it, it's there for a good reason, so I'd put it back.