Adding low pass filter to Fortin 33 copy (high freq noise/hiss problem)

[nothing_nothing]

delete

[nothing_nothing]

delete

[ElectricDruid]

Connecting pin 1 to +V on TC1044 does nothing.

That suggests it was already connected. Did you test that?

Best solution I foud is connecting pins 2-6 on TL071 using 1.5 nF cap, I wonder if using 2.2nF, 5nF, or even 10+nF cap will reduce this noice even further?

You're lowering the highest frequency that the op-amp deals with, so yes, it'll reduce noise, but at some point it'll start to rob you of obvious treble too. There's a trade-off there.

That's why stopping the noise being generated at source is a better solution than removing it later by filtering. But you don't always get a choice.

[MikeA]

Just curious, did you ever try this?  It's worked for me more than once.  https://www.diystompboxes.com/smfforum/index.php?topic=128429.msg1236691#msg1236691

[nothing_nothing]

Connecting pin 1 to +V on TC1044 does nothing.

That suggests it was already connected. Did you test that?

Best solution I foud is connecting pins 2-6 on TL071 using 1.5 nF cap, I wonder if using 2.2nF, 5nF, or even 10+nF cap will reduce this noice even further?

You're lowering the highest frequency that the op-amp deals with, so yes, it'll reduce noise, but at some point it'll start to rob you of obvious treble too. There's a trade-off there.

That's why stopping the noise being generated at source is a better solution than removing it later by filtering. But you don't always get a choice.

So wich cap soldered to TL071 will work the best in my case? I feel like 1.5 nF didin't affect my tone at all, only reduced this hiss a lot.
Something like 5 nF won't damage anything?

Just curious, did you ever try this?  It's worked for me more than once.  https://www.diystompboxes.com/smfforum/index.php?topic=128429.msg1236691#msg1236691

I tested it with only aluminium foil, no change at all :| I will try again later with electrical tape

[ElectricDruid]

So wich cap soldered to TL071 will work the best in my case? I feel like 1.5 nF didin't affect my tone at all, only reduced this hiss a lot.
Something like 5 nF won't damage anything?

No, you can try whatever values you like and you won't damage anything.

I can't guess what value would work best because of that transistor in the feedback loop. If it was simply R7/220K and C8, then 1n5 gives you a roll-off at 480Hz! That's definitely going to affect your tone. But you say it doesn't, so the transistor must change things significantly. (Does anyone else know what that's going to do?)
I'd still recommend trying a few values to see which works best. Solder some wires to the two connections for C8 maybe, and then you can solder different caps onto the end of the wires without having to work on the PCB. Once you've decided, you can unsolder the wires and put the cap on the PCB.

[nothing_nothing]

Ok guys, I did some testing with different caps, also soldering 2 wires to the tl071 pins helped me a lot with testing new caps.

With 1.5 nF hiss/noise was almost gone, but I felt like I need something bigger, so I went crazy with 10nF and it was pretty dark, Not bad for  classic rock/blues tones, but too smooth for metal, also hiss was gone.

5nF was almost like a average ts style pedal, good, but just a little to muffled.

2.2 nF was a perfect choice, all the noise is gone and pedal still sounds like original, but without any weird noise.

I am just wondering if I should remove low pass filter that I added at the start?

[anotherjim]

I think Vref could do with a bypass filter cap. It depends on the RC filter before the divider but the divider resistors are high value so the Vref is a high impedance aerial. 100nF Vref to 0v ought to do it. Vref always needs to be clean, any noise on that is passed out as audio otherwise.

[ElectricDruid]

I think Vref could do with a bypass filter cap. It depends on the RC filter before the divider but the divider resistors are high value so the Vref is a high impedance aerial. 100nF Vref to 0v ought to do it. Vref always needs to be clean, any noise on that is passed out as audio otherwise.

There's a 10K/4u7 RC filter before the divider, which gives a 3Hz cutoff, so I suppose they're hoping that there's not noise much going into the divider. Whether that's better than getting rid of the RC ahead of the divider and sticking a cap on the divider instead would depend on the details, I suspect. Adding an extra cap on the divider as well as the RC would in theory be an improvement, but whether it genuinely offers much benefit is also an open question.

In many ways it's easier to stick the extra cap in and see what happens rather than do the calculations, so go for it.

[anotherjim]

Well, I'm wondering if the divider resistance values make Vref a high impedance node (pretty much 100k) and as there's a switcher on board there's plenty of noise around. 100nF is worth 16k at 100Hz which should be less sensitive to noise pickup. You could tack the 100nF across the lower 820k in the divider.
As Vref couples into the audio input, any noise there is audible.

[ElectricDruid]

Well, I'm wondering if the divider resistance values make Vref a high impedance node (pretty much 100k) and as there's a switcher on board there's plenty of noise around. 100nF is worth 16k at 100Hz which should be less sensitive to noise pickup. You could tack the 100nF across the lower 820k in the divider.
As Vref couples into the audio input, any noise there is audible.

Do you have any idea why they've pulled Vref so far from the midpoint, Jim? I don't get that. And why use 100K/820K for the divider? And the values are so high too, as you said. 10K/82K would do the same job better, except then you'd also have to change the R101/C9 filter.

[iainpunk]

Well, I'm wondering if the divider resistance values make Vref a high impedance node (pretty much 100k) and as there's a switcher on board there's plenty of noise around. 100nF is worth 16k at 100Hz which should be less sensitive to noise pickup. You could tack the 100nF across the lower 820k in the divider.
As Vref couples into the audio input, any noise there is audible.

Do you have any idea why they've pulled Vref so far from the midpoint, Jim? I don't get that. And why use 100K/820K for the divider? And the values are so high too, as you said. 10K/82K would do the same job better, except then you'd also have to change the R101/C9 filter.

when working out the DC, i think we have to look at the transistor as a common base amplifier, the maths dictate that it works as expected. its just a quirk of the odd topology.

cheers

[Vivek]

Do you have any idea why they've pulled Vref so far from the midpoint, Jim? I don't get that. And why use 100K/820K for the divider? And the values are so high too, as you said. 10K/82K would do the same job better, except then you'd also have to change the R101/C9 filter.

If we assume that the role of Vref is to have the output sit somewhat midway between the rails

And in most configs, if we have + pin half way, the output is half way

But in this odd config with a transistor, the Vref needs to be that way as that's the only way to get output roughly midway

Rob had done a basic calculation on this circuit some time back. But a rigorous solution to calculate voltages of this schematic is too difficult.

I had just put that thing into LTSPICE to see where  voltages settle.

[anotherjim]

I just passed over the unequal divider due to the BJT but only by the assumption that it's meant to be. Vcc is boosted high and it won't be strong so large Vref resistor values might be for the best.
There is signa okl, but with noise. LP filtering the audio is sure to make the noise less apparent but you can only go so far. If it's only for guitar use, you can have everything over 5Khz cut as deep as you like without making it sound particularly dull, but it still leaves noise in.

[anotherjim]

If it matches the schematic given earlier, I think the power filtering is actually lacking anyway.
The switcher IC really would benefit from filter caps on its supply pin to 0v. 100uF +100nF(ceramic). The switcher's current pulses will be felt on anything else using the same 9v supply otherwise.
For good measure, another 100nF disc should be placed across the opamp supply pins.
This may or may not cure the noise being complained about. If it's due to the PCB layout not keeping audio and switcher ground paths separated or components just too close, it still may not be enough. However, it would be good practice and may well prevent problems you don't yet know about.

[m4268588]

About 'Vref"

[Vivek]

Rob had done some calculations here

https://www.diystompboxes.com/smfforum/index.php?topic=122749.0

[Vivek]

Biasing on the TC Electronics Integrated Preamp at different supply voltages

[nothing_nothing]

If you have any suggestions for a better fix explain it to me like I am 8 years old, "solder this here and there", or just mark it on the photo.

Currently all the noise is gone, but I have lost some top end attack :|

[m4268588]