Reamper/Pedal-Interface with filters, input needed

[Meathands]

Thanks for the new circuit. Some thoughts:

- Personally I would leave out DO1-DO4. What are the odds someone is going to plug a power amp into an XLR male output jack? I would rather see it at the balanced input, where there's a higher chance of people plugging in something wacky,
- I would also leave out LO1, LO2, CO6, CO7. IMHO it's better to leave the output simple rather than degrading the CMRR with ferrites.
- You could make CP1 and CP2 higher, say 220uF. Why not?
- Why is RF1 100k? It seems high

Sure, you could use that zener scheme to protect the unbalanced input. I see no problem with using that and the FET opamp buffer.

 

[Studio Mollan]

Good input, I'm SO glad you tase the time for this.
Regarding the diodes, caps and inductors on U4, That1646. that exact circuit is from the THAT1646 datasheet. Ive used it like that on other builds. Thats the "recommended setup" so I went with that. Do you think it is hurting the circuit in any way like this?

Regarding RF1. This is the Harrison Ford filter circuit. The original input is 10k for both RF1 and RF2. As I see it they are forming an attenuator to get some more headroom for the resonant filters AND also sets the input impedance for the circuit.
I f I try to keep the input impedance of the filter and the output amp, U2.2, the same I won't get a level difference when I switch the filter on/off. Or? I do NOT know exactly what I'm doing. Might be better so keep the Filter input original and change the input Z if U2.2? OR the different Zs won't cause a level difference anyways?

Thanks again!
/
Emil

Thanks for the new circuit. Some thoughts:

- Personally I would leave out DO1-DO4. What are the odds someone is going to plug a power amp into an XLR male output jack? I would rather see it at the balanced input, where there's a higher chance of people plugging in something wacky,
- I would also leave out LO1, LO2, CO6, CO7. IMHO it's better to leave the output simple rather than degrading the CMRR with ferrites.
- You could make CP1 and CP2 higher, say 220uF. Why not?
- Why is RF1 100k? It seems high

Sure, you could use that zener scheme to protect the unbalanced input. I see no problem with using that and the FET opamp buffer.

 

[Meathands]

There's no harm in keeping the protection diodes and LC filters on the output. I just don't think they are necessary.

Regarding RF1 and RF2, 10k does seem like a better value. If you do that, increase CO1 accordingly. I'm not sure I understand your question about the levels—if you change them both to 10k it will still be the same amount of attenuation, 6dB.

I just noticed something else, you should make CO3 larger. The input Z of the 1646 is 5k. Right now you are creating a 30Hz HPF. I would recommend a 100uF electrolytic. I am for 0.3Hz corner frequency to keep all HPF out of the audio range.

 

[Studio Mollan]

Yes I think I will revert back to 10k. What I meant was this. I wanted to have the same input impedance of both u2.2 and filter input. By altering rf1-2 to 100k I work the same attenuation and changed the input impedance to the same value as u2.1.
And the why: to keep the same load on the preceding stage and so keep levels the same wether the filter is engaged or not.
This so probably overkill anyways so I will revert back to the standard 10k.

There's no harm in keeping the protection diodes and LC filters on the output. I just don't think they are necessary.

Regarding RF1 and RF2, 10k does seem like a better value. If you do that, increase CO1 accordingly. I'm not sure I understand your question about the levels—if you change them both to 10k it will still be the same amount of attenuation, 6dB.

I just noticed something else, you should make CO3 larger. The input Z of the 1646 is 5k. Right now you are creating a 30Hz HPF. I would recommend a 100uF electrolytic. I am for 0.3Hz corner frequency to keep all HPF out of the audio range.

 

[Bo Deadly]

So I would actually keep 1uF film there.

Actually on second thought, I would eliminate that cap entirely. The 1646 wants low Z on the input. I have always DC coupled to the previous stage and do whatever is necessary to keep the offset very low. It's common and good practice to high pass a line output but you should do using the amp feeding it. That would require some re-design of this particular circuit but for-the-record, that's how I would do it if I were designing something from scratch.

 

[Studio Mollan]

Thanks for your input!
The protection stays. It's already laid out on my pcb so i'll keep it the way it is now.

I'm not decided on changing the Cap. Does make sense to keep it as straight as possible. But you are right. 30Hz is cool, especially with the inherent LOFI-ness of this device. Keeping it a film cap also makes sense. Maybe see if I can find a 2u2 cap that fits, or as large a film cap as I can find that fits and doesn't look redicilous. =)
Thanks again!

/
Emil

Those diodes protect the output if it's plugged into a mic input with phantom on which is very possible with a typical console that has XLRs for both mic and line which are probably right next to one another. Plugging into the mic input w/ 48V on by accident would cause mic input coupling caps to discharge into the 1646 output. So they are recommended. Although an alternative way to do it, that I personally like myself, is to use low capacitance TVS diodes shunting to chassis.
The schematic looks like it says "1u film" so there should be no distortion. There's really nothing below 30Hz that anyone is going to hear and any LF rumble could just cause the higher frequencies to clip. So I would actually keep 1uF film there.

 

[Studio Mollan]

FYI. This is now finished and available in the white market.
https://groupdiy.com/threads/studiomollan-tirdi-500-proto-500.79487/#post-1018874