This buffer works fine on final output, but kills gain in middle of circuit

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Mbira

Well-known member
Joined
Jun 4, 2004
Messages
2,422
Location
Austin, TX
Hi folks,
I got this really cheap buffer with these VU meters I picked up on Amazon.  I'm planning on using a Sifam for my project, but I wanted to experiment with this first. 

There is a "high" and a "low" input that you can feed the buffer (unbalanced). 
VU.png


They hid what the IC is.  Here is what the schematic is:

buffer.png


I'm wanting to put the VU at this point in the circuit  (I want to switch between seeing gain after gain knob and gain on the output XLR):
G9_VU.png


As most of you know, I'm really new at trying to do anything beyond just "paint-by-numbers" of following a schematic. 

The Buffer works fine when I put it on the output XLR, but when i put it in this stage it sucks up about 90% of the gain.  I'm guessing the issue is that it has a 2K input resistor while the following stage (the output knob) is 47K.  So I'm thinking that I'd really want a really high value resistor going in to my op amp.  Is that correct?  Like...I want it as high as it can go and still work, right? 

So-assuming I'm correct (and I'm probably not), how do I figure out what opamp I can use and the highest value of resistor I can put there? 

To be clear-i was just testing with this buffer, and I'm planning on rolling my own (with a known op amp). 
 
Try changing the opamp to a non-inverting amp (grounding the input end of 2k, "unground" pin 3 and use it as input, no resistor to gnd needed for current return, because R27 does it...
Tht way you hve a high impedance buffer (really a high gain amp, but it buffers also), that won´t chnge the signal, the polarity of the metered signal will change, but it
doesn´t matter at that point.
Could be the gain is too high, so replacing the 200k with a 100k+100k trim could be something to think about

Cool, that you go past painting by numbers (not that anything is wrong with that), it seems a steep learning curve to learn the basics, but afterwards it feels like really easy (not that I am too far ahead of you ;) ). Dive into the concept of "impedance"/"reactance", there´s a quote (from PRR?) somewhere, that everything in electronics is basically just voltage dividers, that one was helpful to me... :D
 
From the schematic you posted it looks like the input impedance of the buffer is about 2K. This is fine for the 600 ohm output but will seriously load the circuit at the mid point you highlighted. To monitor level at this point you need a really high impedance input; several megohms would be preferable. So you need to use a FET input op amp in the non-inverting configuration so that the input impedance is defined by the bias resistor.

Cheers

Ian
 
As has been pointed out you need a high input Z here.
It's not really practicable to just increase the 'input' resistor since you'd need to scale the feedback resistor the same to maintain the same gain. It would work in theory but you can get into problems with very high resistor values eg contamination  of the resistor body becomes significant in defining the actual resistance value etc.
To keep things simple and use the circuit as is I'd suggest to simply add a non inverting unity gain op amp stage in front of it to buffer the signal. R27 will define the input Z as 1M0.
It doesn't really matter whether the op amp is FET or BiPolar input type - the 'input resistance' is multiplied by the open loop gain of the op amp. Or to put it in terms of the op amp pins - pin 2 (inv input) is driven to the same voltage as pin 3 (non-inv input) where your signal goes. Thus no current is drawn - well it wouldn't be for the ideal opa mp with infinite open loop gain. Real opamps have very high (obviously not actually infinite) open loop gain.
FWIW I probably would opt for a FET type to minimise the current noise introduced into the circuit.
 
FET opamp is definitely preferred. With the one position of the HPF you have a DC path to P1 pot which would then see a lot of the opamp input bias current.
 
The easy fix to try is scaling up resistor values.  You can easily get 2M for the feedback which would support a 20k input R. Noise may be a little higher which won't affect the meter, but buffer output could have some hiss.

From a blank sheet, a non-inverting topology will be lowest possible loading. Bifet op amps better yet. I am not the tube guy around here but you might want to consider clamps on the op amp input to protect against voltage spikes.  A simple diode clamp from op amps inputs to (op amp) PS rails. A series input R even with non-inverting topology will help protection work.

JR 
 
Converting to different opamp config, or different opamp, is thwarted because here the opamp is single supply, input at ground. Several opamps will work this way. When external input swings negative (below opamp ground) the output dutifully swings positive; when external input swings positive, the output tries to go negative but can't. (This also means the diode-to-ground is a wasted penny.)
 

Attachments

  • VU-buff-tube.gif
    VU-buff-tube.gif
    53.8 KB · Views: 62
PRR said:
Converting to different opamp config, or different opamp, is thwarted because here the opamp is single supply, input at ground. Several opamps will work this way. When external input swings negative (below opamp ground) the output dutifully swings positive; when external input swings positive, the output tries to go negative but can't. (This also means the diode-to-ground is a wasted penny.)

PRR, can you clarify a couple points: Are there special specs that I need to consider for the NPN since I'm plugging in to the Plate voltage? 

Is this NPN taking the place of the VU buffer, or is it ahead of the buffer?  In this case, I'm not sure if the circled VU is referring just to the meter or the whole buffer.  Thanks.
 
JohnRoberts said:
What PRR drew is a mosfet source follower.  Yes it takes the place of a buffer/

JR

Thank you.  I'll look that up.  PRR is the king of replacing 20 parts with 1! 
 
The source for the mosfet source follower has high output impedance so I'm  concerned about his frequency response  because of Cgd and other capacitances.  There is also very long DC settling time.
 
moamps said:
The source for the mosfet source follower has high output impedance so I'm  concerned about his frequency response  because of Cgd and other capacitances.  There is also very long DC settling time.
I'd be more concerned about the feedback he has created from the SRPP back to the previous stage, warping the level control
 
Sorry for the messy drawing. It's adopted schematic from one project, so  some parts may be redundant, but I think it works.
 

Attachments

  • vu buffer.png
    vu buffer.png
    28.8 KB · Views: 37
moamps said:
Sorry for the messy drawing. It's adopted schematic from one project, so  some parts may be redundant, but I think it works.

Thank you for this.  Is the "selector" anything more than a DPST Switch?
 
Yes, it is DPST if you like to monitor just one channel. Otherwise, you can use rotary BBM type switch if you like to monitor more channels. 
 

Latest posts

Back
Top