DIY Wampler Faux Analog Echo opamp bias issues

[Domin]

Hi, i made this PCB and i have problems, after few seconds sounds cut off and i found that voltage on voltage divider going into 220K and 1K slowly rises, pedal works when its bellow 4V but after bias voltage gets to 4,3V, sound is completely cut...any suggestions?

[PRR]

There may be a 100k+100k divider near a 220uFd cap. The grounded end of one of those resistors is not really grounded. Good bright light and magnifier may find it.

https://www.madbeanpedals.com/forum/index.php?msg=107685

[bean]

The 220k and 1k to vRef is borrowed form the Rebote delay design. I suspect they were used to reduce the current consumption of the TL072 (unless someone has a different explanation). Try jumpering both resistors and see if you get a better result.

[ElectricDruid]

The 220k and 1k to vRef is borrowed form the Rebote delay design. I suspect they were used to reduce the current consumption of the TL072 (unless someone has a different explanation).

I thought resistors to ground like that were to reduce output offsets. The idea is that you match the resistance seen on the two inputs. If input bias current for each input is across a similar resistance, the resulting offset voltage is similar. And since it's the differential voltage that matters, we want those offsets to be as similar as possible (if we care about the output offset). Which is why the bottom op-amp has a much larger value, 220K - it needs to be the same as 510K and 1M in parallel, so 220K is pretty close although 330K is better.

Try jumpering both resistors and see if you get a better result.

+1 agree.
For audio, these resistors are basically pointless, since we don't care about small DC offsets, and in both cases the output is followed directly by a couling cap, so any DC offset would be blocked anyway.
Neither R2 or R3 are required at all.

[Domin]

There may be a 100k+100k divider near a 220uFd cap. The grounded end of one of those resistors is not really grounded. Good bright light and magnifier may find it.

https://www.madbeanpedals.com/forum/index.php?msg=107685

I tried to reflow both resistors and nothing changed

[Domin]

The 220k and 1k to vRef is borrowed form the Rebote delay design. I suspect they were used to reduce the current consumption of the TL072 (unless someone has a different explanation). Try jumpering both resistors and see if you get a better result.

Somehow with this type of pcb im not sure if ill risk desoldering it or connect 2 ends with traces

[ElectricDruid]

How far does the signal get? Have we got signal at pin 1 of IC4A? Does it make it through the PT2399?

Since the bias voltage is *supposed* to be 4.5V, I don't see why 4.3V should make it cut out. Something fishy is going on.
According to my calculations, that 100K/220u combination in the Vref divider is going to take over a minute to charge up! So we'd fully expect to see a slowly rising Vref after switch on, and for a good while. But that shouldn't stop it from working, at least not once the voltage has risen beyond a couple of volts.

[Domin]

How far does the signal get? Have we got signal at pin 1 of IC4A? Does it make it through the PT2399?

Since the bias voltage is *supposed* to be 4.5V, I don't see why 4.3V should make it cut out. Something fishy is going on.
According to my calculations, that 100K/220u combination in the Vref divider is going to take over a minute to charge up! So we'd fully expect to see a slowly rising Vref after switch on, and for a good while. But that shouldn't stop it from working, at least not once the voltage has risen beyond a couple of volts.

For now i dont have audio probe, so ill need to make one

[bean]

The 220k and 1k to vRef is borrowed form the Rebote delay design. I suspect they were used to reduce the current consumption of the TL072 (unless someone has a different explanation).

I thought resistors to ground like that were to reduce output offsets. The idea is that you match the resistance seen on the two inputs. If input bias current for each input is across a similar resistance, the resulting offset voltage is similar. And since it's the differential voltage that matters, we want those offsets to be as similar as possible (if we care about the output offset). Which is why the bottom op-amp has a much larger value, 220K - it needs to be the same as 510K and 1M in parallel, so 220K is pretty close although 330K is better.

Ahh, yeah that makes sense. Thanks for the explanation. I guess you could just put a buffer in front then reduce the value of the two resistors to the inverted input like a DM-2. Then you have your high impedance input and minimize any offset from the IC without using the extra resistor.

[ElectricDruid]

I guess you could just put a buffer in front then reduce the value of the two resistors to the inverted input like a DM-2. Then you have your high impedance input and minimize any offset from the IC without using the extra resistor.

Yeah, that would work nicely.

Personally, I'd probably simply not worry about the inversion, and turn that IC4A into a buffer rather than an inverting amp. But I know other people stress about that more than I do, so it's a question of taste!

[PRR]

I tried to reflow both resistors and nothing changed

"Good bright light and magnifier may find it."

[Domin]

I tried to reflow both resistors and nothing changed

"Good bright light and magnifier may find it."

Tried it with my father and we didnt see anything...