Relays and Op Amp Damage

[mattthegamer463]

R4 is the resistor from Vcc/2 to GND in the voltage divider, did you mean R3, the resistance between the divider and the signal path?  That could be higher.

So you're estimating the reactance of the pickups as being 10k for low frequencies and 100k for high frequencies.  Because of that 10k, the high impedance of the pickups can't drive the low impedance input.  But the impedance of an opamp is something huge, right? The JFET inputs are deliberately very high impedance so that it is versatile, and the output is very low so that the op amp won't have a hard time driving into the next thing on the signal chain.  I'm not sure I see how that 10k comes into play.

Should I place a 1M resistor from signal input to ground?  Thats the classic fix for input impedance isn't it?

AMZ shows the input impedance as such:
http://www.muzique.com/lab/imp.htm

[R.G.]

So you're estimating the reactance of the pickups as being 10k for low frequencies and 100k for high frequencies.  Because of that 10k, the high impedance of the pickups can't drive the low impedance input.

Actually, he was trying to illustrate what the highly inductive nature of pickups does to frequency response. Pickups can be modeled as a signal voltage in series with a 4K-10K resistance, in series with a 0.5H - 4H inductor. The impedance of an inductor is Xl = 2*pi*f*L, where f is the frequency. A 1H inductor has an impedance of Xl = 2*pi*1000*1 = 6280 ohms at 1kHz, 62.8K at 10kHz.

But the impedance of an opamp is something huge, right?

Theoretically, yes. The input impedance of REAL opamps varies. Some highly regarded ones are as low as 100K.

The JFET inputs are deliberately very high impedance so that it is versatile, and the output is very low so that the op amp won't have a hard time driving into the next thing on the signal chain.

Should I place a 1M resistor from signal input to ground?  Thats the classic fix for input impedance isn't it?
AMZ shows the input impedance as such: http://www.muzique.com/lab/imp.htm

Putting a resistance across an input can only lower the input impedance that's already there. It can never raise it. You cannot force a low input impedance to be higher by putting a high resistance across it. So no, putting a 1M resistor from signal to ground is not a classic fix for input impedance. That can lower and stabilize a much higher and variable impedance that's causing other issues, but it cannot raise a low impedance.

[mattthegamer463]

Putting a resistance across an input can only lower the input impedance that's already there. It can never raise it. You cannot force a low input impedance to be higher by putting a high resistance across it. So no, putting a 1M resistor from signal to ground is not a classic fix for input impedance. That can lower and stabilize a much higher and variable impedance that's causing other issues, but it cannot raise a low impedance.

So, the issue is probably that R3 is lowering the actual impedance of the op amp input, and so if I increase its resistance it will help to maximize the input impedance?  I'll test this out tomorrow.

[PRR]

Yes, R3, my bad.

For TL07_ that resistor may be 1Meg.


> The input impedance of REAL opamps varies. Some highly regarded ones are as low as 100K.

If you are thinking 5534.... that's an open-loop value. You wouldn't run 5534 open-loop. Assuming gain-margin of only 10 (closed-loop gain of 10 at 50KHz) the effective input is 1Meg.

Working with TL072 in the active linear zone, the input impedance is of course "infinite" for any normal audio purpose. They even work effectively as 13 cent buffers for condenser mikes, where hundreds of Megs would be significant.

[mattthegamer463]

Hey guys, I've been working on this project over the last two weeks or so and its finished.  It sounds good, functionality is good.  I'm posting the schematic, which I think would be safe to use, but the transfer isn't.  DIY Layout Creator just gets so clunky when you're building things with more than a few components, and my installation is having a problem where I can't right click to select anything, and nothing can highlight, so it becomes near impossible to work with.  I eventually did as much as I could and then corrected the issues with MSpaint. 

Anyway here are some photos. (click for largeness)

Layout.  The box at the top is to allow the PCB to straddle the DC jack and the relays can rest on the back of the volume in pots.


Etched and drilled PCB


Top


Bottom, populated


Top, populated.  I totally missed a very important cap in my layout so I had to add it in later (in case someone happens to spot the problem here)


Wired up


Finished




A mode


B mode


M mode


Thanks for your help and suggestions PRR and R.G. Keen.  I hope you guys like how it turned out.