A "Transparent" Buffer?

[Loose]

I did use a TL072 for both the buffer and the mixer, just wanted to make sure about the extra pins. Thanks!

[Loose]

I breadboard only the buffer part of the circuit to test it. I get barely any output from the effect, but if I unplug the power from the pedal I get the output back.
I removed R16 from the Surfy Bear PCB and soldered a wire to it to use as VDD.

Any ideas what could be wrong?

[Kipper4]

And your voltages are ?

[Transmogrifox]

Your layout doesn't show anywhere you connected ground wires into the breadboard ground traces.  Also R9 is shown as if it's going to the VDD trace on the breadboard, but should be going to ground.

If you have it literally wired according to the layout then then R1 and R9 don't have a ground connection.  Having R9 floating would make it not work.

Make certain you have ground wires going into both blue traces of the breadboard.

The layout looks like it will work as long as there is a ground connected where the layout seems to show it going.  You do need to connect grounds and power into your breadboard ground and power traces if you expect to get ground and power out of them.

If it's just a matter of the drawing leaving things out that you actually did on the BB then I can't imagine why it wouldn't work unless external connections are wrong (some of those "oh duh" things we do).

Also it would be a good idea to put a jumper between pins 6 and 7 of the op amp to make sure the unused device isn't unstable.

With a DMM you should be able to ring out the problem pretty fast or give us something to work from .  With black lead on ground you should measure the following with the red lead:
Pin 8: 12V
Pins 1,2,3,5:  6V
If you tie pins 6 and 7 with a jumper then you will see 6V also on pins 6.7.

R9 grounded side:  0V
R1 both sides:  0V
R2:  6V 

[Loose]

The drawing doesn't show it but I did connect a ground wire. R9 is grounded as well. (I connected the outer rows on both sides to ground)

I will check the voltages.

Thanks

[Loose]

So I plugged in the pedal to test it and smoke came out... Can't be a good sign.
I unplugged it and plugged it again later and no smoke came out (possibly because that component already got burnt).
No burn marks, and I got voltages reading from all of the transistors, but i'm still not sure if everything is ok.

Regarding the reading of pin 8: That is connected to Vdd and it reads 4.6v. Should Vdd read 4.6v or 12v?

I made sure that the power supply actually outputs 12v.

[Kipper4]

Welcome to the blue smoke club.

Did you make sure that you checked the right power supply rails.
Check your supply for centre negative? Centre positive?
Check your supply is outputting a DC voltage not AC.

If your using a 12Vdc supply then
Pins 1,2 and 3 should measure 6V
Pin4 0V
Pin8 12V

This may vary a bit with meter loading but it should be in the parish.

Google the data sheet for the chip.
What chip are you using btw? Some dip 8 packages are single op amps which have a different pinout to a dual op amp.

All of the above questions are based on the breadboard diagram you posted.

I Hope you bought some spares.

[Loose]

I assume you didn't read the whole thread since there's a lot in it

Yes the power supply is correct, this is an effect that was working properly. I just decided to modify it.

Before we get into the chip voltages, I need to know how many volts should Vdd put out. Here's the schematic:
http://cackleberrypines.net/transmogrifox/Misc/surfybear-jfet-reverb-r3.pdf

Currently Vdd puts out 4.6v, is that the problem? Vdd is feeding the opamp pin 8.
+12v in the schematic does put out 12v.

Thanks

[Kipper4]

Your right I didn't read it all.
Forget what I said. I should have read through.
Hope you get it sorted.

[Loose]

Thanks regardless Kipper.

Another piece of information: when connected to opamp pin 8 Vdd reads 4.6v, but when I disconnect it from pin 8 it read 6.4v.

[Transmogrifox]

VDD should read really close to 12V.

When something smoked probably this was VDD getting shorted to ground somehow.  Either way high current in the 22 ohm resistor feeding VDD will turn it into something much higher than 22 ohms.

Replace that resistor (R5) and then at least you will have 12V at VDD.

Hopefully it was just a momentary short where a couple wires touched.  In that case everything will probably be ok once you replace the resistor.

Just double check your wiring on the breadboard.  It is easy to be off by one row and not notice it.

For example, if you had VDD connected to pin 7 by mistake that would make smoke and would require a new op amp.

[Loose]

You're right R5 was totally burnt! I missed it somehow, I think some of the alligator wires touched like you said. Thanks!

I replaced it and now all the voltages are correct, but I lose some output when the buffer is on. Any ideas?

Thanks

[Transmogrifox]

How much do you lose when it is on?  Where is the volume knob on your guitar set?   If it's subtle or if the volume knob isn't at 100% then it might simply be a result of losing a little gain against the impedance looking back into your guitar.  You could improve this by increasing the 470k resistors to 1Meg, or 2.2Meg if necessary (this is why a breadboard is so nice).

If your guitar volume is wide-open and you're hearing significant attenuation then something else is wrong.

Let's start by assuming the circuit is operating correctly because your DC voltages are all measuring correctly.  Here are some simple things to try:
Crank guitar volume wide open then do an A/B listen test
Change the 470k resistors 1 Meg,
Change to 2.2 Meg. (make the capacitor about 1/2 its value or you'll start to notice the loss on the high end).

If that makes it better then go with it.

[Loose]

I made sure everything is sitting properly on the breadboard and that the wires have proper connection and the problem was solved
Breadboarding can be annoying.

It sounds good! About 85% close to the dry signal, is there anyway I could make it even closer? There's a bit less clarity, highs, mids and punch when the buffer is on.

Also, when using this kind of mixer circuit, is the buffer even necessary?

Thanks

[Transmogrifox]

The buffer in front isn't strictly necessary.  The reason deep blue does this is because the phase reversal on the mixer.  Best practice is to design a pedal so the phase is not inverted, so deep blue uses an inverting buffer followed by an inverting mixer for a net zero phase change from input to output.  In practice this makes very little difference unless either you depend a lot upon the sustain you get from resting the headstock of your guitar on the amp (positive feedback) in which case it will change how the guitar sustains.  Or if you intend to split FX chains and mix the 2 parallel chains, then a phase reversal will result in cancellation of anything common between the two FX chains when they mix.

Unless you intend to do either the above it isn't strictly necessary to correct the phase.  In fact a lot of circuits don't make any attempts at correcting phase.  If it's a phase-inverting effect, then that's just the way it is.

As for clarity, start by removing the RF blocking capacitor.  I think of a -3 dB at 20 kHz inaudible to me, but this does mean about -1 dB at around 7 kHz.  Maybe this small high end roll off is audible, but generally the guitar pickups and cable are rolling off at around 5 kHz, so I typically wouldn't think of this as being particularly audible.  This is small enough capacitance already that the breadboard inter-trace capacitance makes the high-end roll-off more extreme than a calculation on nominal values implies. 

I think increasing input impedance is possibly another cause.  Try changing the resistors to 2.2 Meg, but keep in mind if inter-trace bread-board capacitance was a problem before, then it will be a bigger problem with a bigger feedback resistor, so the sound you hear on the breadboard might not be representative of what it might be in a good layout.

If this all is a problem, then maybe the best thing is to make the mixer with smaller valued resistors (like 33k or something) then buffer both the guitar and the surfy bear with the simple JFET follower or noninverting op amp buffers to get the required input impedance.  High-end roll-off, noise and stability are the trade-offs with increasing input impedance in an inverting unity gain follower on an op amp.

[Loose]

Thanks man. I want to test just the mixer without the buffer. Can I just remove the IC1 section (resistor and capacitor), or do I need to change any of the values in the mixer? Also, do I need to remove either R2 or R4?

Thanks

[Transmogrifox]

If having phase inverted doesn't bother you here is how to do it in 1 simple stage.

[Loose]

Thanks for the schematic!
So does the mixer only part functions as a buffer as well? Since they are constructed the same.
Could you also explain why you chose this opamp over the tl072? And wouldn't it be more beneficial to use a singal opamp here and save the dual for other applications?

[Transmogrifox]

Yes this combines the functions of the buffer and mixer into one unit.  It's basically a mixer with input impedance comparable to a typical guitar amp.

For the guitar side have 470k in parallel with 4.7Meg for input impedance = 427k, which should be high enough to prevent tone sucking off your pickups (if you are happy with the deep blue clean path, then you will be fine with this).

Input impedance presented to the Surfy Bear is 470k (face value).

The op amp chosen is a matter of convenience since I was using LTSpice to run the circuit simulation.  I picked a Linear Tech part that has similar specs to the TL072 so my simulation results would be similar to what you would get by swapping in a TL072.

You can use a TL072 and tie off the unused op amp to VB as previously described.  In fact almost any audio grade op amp will work here.

And, yes, if you happen to have a single op amp laying around then use it here and save the dual for better uses.  The cost difference between a TL072 and a TL071 is almost nothing so it wouldn't make sense to make an order of single op amps just for this circuit unless you are planning to go into production and sell them.

[Loose]

I wired the breadboard like the schematic except for the 3.3pf cap, the smallest one I had was 10pf.

The only thing i'm missing is where to connect the surfy bear's input (Q4), should it go after R3?

Also, since i'm currently testing it with the tl072, should I do anything with the empty pins 1 and 2, or can I leave them not connected for now?
Pin 3 is still connected to VB, I can change that if needed.

Thanks