Troubleshooting a High Frequency Mixer: Oscilloscope Output Analysis

[nst.john]

Hi all,

I am making a frequency mixer to add together channels of an amplifier I have at 52.892kHz. I am attaching a picture of the schematic and the physical circuit I made along with the output on the oscilloscope I am measuring. Instead of the three outputs shown in the schematic, I am just using one, and same for the five inputs I am just testing with one right now. Here are the pictures and I wanted to know why the oscilloscope image is A) much less than the input voltage of 10Vp-p and B) not a clean sine wave, but is flat on the top peaks. Thank you guys for your help and hope to hear back soon!

 

[nst.john]

Sorry for all the extra pictures my computer was being weird.

 

[berkeman]

The circuit appears to be an adder, not a mixer (a mixer multiplies the input signals). Also, the pinout of the TL072 looks strange. Can you post a copy of the schematic showing the two opamp symbols instead of the IC view?

 

[sophiecentaur]

The circuit appears to be an adder, not a mixer

He will have heard terms such as "Mixing Desk", which is really an 'adding desk'. But that sounds like some sort of computer.
Even Engineers can't agree amongst themselves so what hope has the general public got?

 

[davenn]

The circuit appears to be an adder, not a mixer

it is the most common method of "mixing" multiple audio inputs
As Sophic suggested, the description has always been dubious at bestDave

 

[berkeman]

Thanks for the clarification, guys. Given the OP's base frequency of 53kHz, I wasn't thinking in the context of audio mixing, but I think you are right.

The opamp pinout still looks odd to me, though.

 

[davenn]

Given the OP's base frequency of 53kHz

yup not exactly audible audio range huh

The opamp pinout still looks odd to me, though.

you are correct ... I hadn't looked that close

wonder where the circuit came from @nst.john it has several incorrect pin connections and that circuit will never work wired like that

here is the correct pin connections for a TL072 ...

pin 1 being top left as usual
Dave

 

[berkeman]

I think the TL072 is a dual: The TL071 is a single, but the OP seems to be using the TL072...

https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/datasheets/Opamps/072dia.gif

 

[davenn]

I am making a frequency mixer to add together channels of an amplifier I have at 52.892kHz

and as an aside ... what are you using 52 kHz for ? ... that's a way up in the ultrasonic range
tho the TL072 can handle that freq., if you used the capacitor values in that circuit of yours, I suspect that their values are probably too highDave

 

[davenn]

I think the TL072 is a dual: The TL071 is a single, but the OP seems to be using the TL072...

https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/datasheets/Opamps/072dia.gif

crap .. I screwed up there hahaha I should have known better
mite as well delete 1/2 my last post comments

 

[jim hardy]

The opamp pinout still looks odd to me, though.

it's usually helpful to draw the amplifiers inside the box
here's a snip from the center of his wiring diagram , i added them

i like hand drawn schematics
That way your brain is thinking about how it works .

 

[Baluncore]

Looking at Jim's circuit in post #11, I see two irrational circuit features.
1. The 10k pot on the RHS is across the op-amp inputs. There will be zero output.
2. LHS, Pin 2, the inverting input has a series resistor, there is no feedback from the output.

Both non-inverting inputs are grounded which is sensible.
There must be layout problems with both the inverting stages that should make a non-inverting amplifier.
It certainly explains the loss of signal.

 

[Baluncore]

This looks like it might be a better circuit.

 

[nst.john]

The schematic was off of instructables.com but it's just not working with the dual op amp. If I made the circuit that @Baluncore made would it be able to work? Thanks for all the help it really is great.

 

[Baluncore]

If I made the circuit that Baluncore made would it be able to work?

It should. You should be able to modify that bit of your circuit to match my diagram. The LHS is an inverting summing amplifier with a gain of –1. The RHS is an inverting buffer amplifier with a gain of –1.

My circuit modifications are the closest sensible circuit to the layout we were given.
I would need a link to the instructables.com circuit to know what was intended.
Build it, if it works it will confirm we understand what is happening.

Then consider moving the potentiometer from the output to the point between the amplifiers so that gain would be more independent of output load.

 

[jim hardy]

^^^^^^What Baluncore said...

2. LHS, Pin 2, the inverting input has a series resistor, there is no feedback from the output.

I wanted to know why the oscilloscope image is A) much less than the input voltage of 10Vp-p

How much output did you see? Can you put a number on it ?
I don't trust those computer generated numbers on o'scope screen because they might have come from a Microsoft program(see my signature).

i see about half a division , now many millivolts is that ? Can i believe ~20 ?
EDIT - make that a whole division peak to peak, ½ a division up and ½ a division down(i knew i had something wrong - jh)

Now - why was there some output ?

At 52khz your opamp no longer has infinite gain(well, it's no longer a quarter million to be more precise)
meaning it can't hold its inverting input so close to zero as it could at lower frequency
http://www.ti.com/lit/ds/symlink/tl072.pdf , page 10 top left

open loop gain is down to around only 75-ish !
and there's 90° phase shift
so it's no longer able to absorb all that 1/2 milliamp input drive current. Some of it flows through that top 10K resistor toward the other opamp . I'd guess he too is unable to absorb it all.

Why it's asymmetric i do not know but,,,
It might be informative to see what it looks like with square or triangle wave input.
That's because when you shift phase of a sinewave it doesn't change shape
but a square wave ... try it and see?

We all learn by our mistakes, so never let one go to waste !

Have fun,
and my hat is off to you for building this thing .

old jim

 

[Baluncore]

Why was there any output ?
As Jim wrote, the time delay through the op-amps at higher frequencies will result in a phase shift and non-zero clamping of the op-amp outputs, hence some output signal will appear.

I think some of the small voltage seen as output may also be ground voltages due to signal generator and analogue ground currents. Where was the oscilloscope signal ground connected ?
Both op-amp outputs would be biassed by their input offset voltages, which might explain the amplitude asymmetry through power supply and ground currents.

The only way to advance this project is to modify the circuit and test it again. If that does not get a real output then we need a full link to the original circuit diagram and another photo of the components.

 

[nst.john]

Here is the URL for the instructables page:
http://www.instructables.com/id/Birth-of-Man-Mixing-Board/

Thank you for all you're help

 

[jim hardy]

..

Here is the URL for the instructables page:
http://www.instructables.com/id/Birth-of-Man-Mixing-Board/

Well you found the trouble

he says in step 8

The circuit for this mixer is based almost entirely off an incredibly awesome schematic found here.

his step 8 does not faithfully represent the awesome schematic there,,, http://www.all-electric.com/schematic/simp_mix.htm

if he'd drawn in the little amps he'd have seen his mistake
can you email him ?

 

[Baluncore]

Here is my recommended mod that reduces sensitivity to output loading.

 

[davenn]

..Well you found the trouble

he says in step 8

his step 8 does not faithfully represent the awesome schematic there,,, http://www.all-electric.com/schematic/simp_mix.htm

View attachment 103950

if he'd drawn in the little amps he'd have seen his mistake
can you email him ?

nice detective work
there's always a good explanation

instructables.com is about the worse place ever to get electronic projects from
so many of them are total crap. People put up whatever they want and there is no vetting system to weed out the rubbish or get errors correctedD

 

[jim hardy]

Lastly

i don't see so well as i used to
tell me if I'm off the beam here...

 

[davenn]

Lastly

i don't see so well as i used to
tell me if I'm off the beam here...
View attachment 104087

agreed

 

[Junaid Shahid]

The circuit appears to be an adder, not a mixer (a mixer multiplies the input signals). Also, the pinout of the TL072 looks strange. Can you post a copy of the schematic showing the two opamp symbols instead of the IC view?

I hope these resource will help you :) - TL072 Pinout Details

 

[Rx7man]

Is there any chance the polarized cap on the input coupling could be part of the problem as well?