Measuring buffer impedance

Started by effectsbay, March 20, 2010, 01:46:19 AM

Previous topic - Next topic

effectsbay

Hello All

I'm working on a buffer circuit and I'm curious how to measure the impedance to see if the buffer is actually doing anything. To my ear, I'm having a hard time distinguishing the difference with it engaged.. or not, so I wanted to view it via multimeter. So my noob question, is how to best measure to see low and high impedance, and what are the values I should be hitting? My initial test, and I could be doing this totally wrong, is showing 1.048M with the circuit bypassed.. and the 1.1.29ish when it's engaged. Not sure if I'm in the ball park, etc. I basically connected the multimeter from circuit output to return on my prototype board.

Thanks in advance!
hank

R.G.

You can't easily measure input and output impedances, at least not that way, with a multimeter. That's because the components inside the input and output capacitors affect it too.

What you usually have to do is to put a constant signal into it and measure the voltage out of it with the multimeter. When you get this set up so that you're measuring for instance 1.0V ac at the output, then you insert a resistance (a pot is handier than fixed resistors) and vary the resistance until the output voltage is half of what it was at zero resistance. Then measure the resistor. The resistor is equal to the input impedance for AC signals because it has divided the input signal in half.

On the output you do much the same thing, in theory at least. You set the output voltage to some convenient value to measure, then load the output down with a resistor to ground and lower the resistance until the output voltage is half. At that point the loading resistor is equal to the internal source impedance, and you just measure the resistor. This is complicated by the fact that some buffers have quite low output impedances, sub-100 ohm values, and they may not be able to drive a matched load without burning, distorting, or other bad circumstances. You can try doing loading with a load that reduces the output signal by 1/10, say from 1.0Vac to 0.9Vac.

That would happen when the load resistance and the source resistance make Vout= Vac* (Rload/(Rload+Rsource), so that the source impedance is Rs = ((Vsource/Vout)-1) * Rload. In the case of lowering the output voltage to 90% of the no-load value, then Rsource = 0.1111 times the loading resistor.

As an example, if you measure 1.0V no load, then get 0.9V with a 2.2K resistor, then the source impedance is 2.2K*((1.0/0.9)-1)= 2.2K * 0.111 = 244 ohms.

It is quite taxing to do this with very low impedance outputs. Opamp outputs may have an output impedance of 100-200 ohms open loop, but feedback reduces this by the open loop gain, which may be 100,000, to an insubstantial fraction of an ohm. So while you would in theory measure this at fractions of a milliohm, it's impractical to either measure ohms, much less milliohms with a typical DMM, and it's also difficult to measure the tiny drops in output voltage accurately enough to matter if you have a high "multiplier" loading. The opamp will also current limit, so unless you monitor what the waveform is while you measure it and measure a tiny signal, you get bogus results.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

effectsbay

Thanks much for replying to this. I need to read this 20 more times or so to fully understand what you're saying. Just curious, is there a *easy* way to just determine if my buffer circuit is working?

Thanks!
hank

effectsbay

Quote from: R.G. on March 20, 2010, 10:27:30 AM
What you usually have to do is to put a constant signal into it and measure the voltage out of it with the multimeter. When you get this set up so that you're measuring for instance 1.0V ac at the output, then you insert a resistance (a pot is handier than fixed resistors) and vary the resistance until the output voltage is half of what it was at zero resistance. Then measure the resistor. The resistor is equal to the input impedance for AC signals because it has divided the input signal in half.

Would a 10k pot work for this?

slacker

A 10k pot probably won't work, the pot needs to be a bigger value than whatever the input impedance of the circuit is. So if you think your circuit has about 1Meg input impedance then you need at least a 1Meg pot.

This link might be useful http://www.diyguitarist.com/PDF_Files/ImpedanceTester.pdf

petemoore

Just curious, is there a *easy* way to just determine if my buffer circuit is working?
  If you have a need for it it is...
  If it's a Tonestack which is loading the driving circuit, or a long or bad cable, or a wah or FF that responds differently when not loading the driving circuit.
  By comparing the difference of Wah Tone, or the way a TS works or the tone of FF when buffer is in or out...with the non-difference that any DodBossIBanezDano or many other pedals with buffered output.
  Basically if the circuit biases, and there's a near 1:1 boost from transistor input to transistor output...the buffer's working [pull battery to be sure it's not just signal leaking around the transistor somehow].
  Sometimes you can hear 'it' anyway, just straight guitar through a couple 10' cables. 
Convention creates following, following creates convention.