A "Transparent" Buffer?

[Transmogrifox]

C3 and C4 in my schematic are the same as C1 and C10 in the surfy bear schematic.

Just connect the wiper of R15 (50k tone pot) to where schematic says "SURFY_BEAR"

If Pin 3 is connected to VB,
then Pin 1 connects to Pin 2.

Then I assume you are using pins 5,6,7 for the mixer.

You can put 3 10pF caps in series to get 3.3pF, but for a breadboard test you can do without it altogether.  The purpose of this cap is to reject RF noise and to prevent weird things from happening above audio frequencies.

[Loose]

C3\C4 is clear to me, they connect to the mix pot, but that's the surfy bear's wet output.

I'm talking about the wet input, in the original surfy bear schematic the signal is split before Q4, so my question is how do I implement that with the new schematic. Do I keep it split before Q4?

Thanks

[Transmogrifox]

Just leave the wet input the way it is in the surfy bear schematic and then take the mixer input also off the input jack.

Improvements would be to remove C12 and change R21 to something like 10 Meg.   Although these aren't strictly necessary, the input impedance will be lower than expected if you don't:
Mixer input Z = 427k
Parallel R21 = 1Meg
Then with C12 in place you would have R16 parallel R20 = 1.1 Meg

So 427k || 1M || 1.1M = 235k, which is probably ok, but it's simple enough to get it up. 

Reducing the surfy bear input side to only R21 to ground and changing that to 10 Meg will get you to
427k||10M = 410k, which is much better.

[Loose]

I put together the mixer on breadboard but it didn't work, so I decided to just build it on a perfboard and avoid the breadboarding errors.



It still doesn't work, there's no signal even on the dry side of the pot.
I tested it with an audio probe, leaving the input cable connected and using the output cable to probe. I'be found that I get no signal on pin 6 and on mixer 2, I did get a signal on pin 7 and on the out connection.

Voltages seemed weird, 10v on pin 7 and 2v on pin 6, so I removed the reverb in wire (that leads to Q4 on the surfy bear schematic) and voltages were ok at 6v, but still no signal.

Any ideas? Does it make a difference that pins 1,2 and 3 are not connected to anything?

Thanks again for all of your help Transmogrifox!

[Loose]

Update: I changed the chip and now I get a signal. However, when I connected the wire to Q4, the signal was gone again, and again the voltages were 10v on pin 7 and 2v on pin 6. Perhaps I need a resistor or a capacitor that would go before Q4?

BTW I forgot the 100k resistor to ground that you've added on the output, is that necessary?

Thanks

[Transmogrifox]

You shouldn't be connecting to Q4 anywhere.

Double check your connections.  Also the way you are describing things you might be connecting off-board stuff to the wrong place.  That would also explain weird voltages.

See if this helps make more sense how you hook it up. Should be the only signal connection to the Surfy Bear board is to C1,C10.  Then you will be tapping VDD and ground from somewhere on the board.  Mixer input comes directly off the input jack.

[Loose]

I removed the connection to Q4 and added a connection at the input and it works! Sounds great at first listening! THANKS!
I already had C12 removed before that, so it's not an issue.

So regarding the 100k resistor to ground at the output that you've added, should I add it in as well? What's its purpose?

And regarding the input resistors to ground, so currently I have 4.7M on the mixer side, and 1M on the reverb side (although that one is after the 100nF cap, I don't know if it makes a difference). In your previous post you suggested changing the 1M to 10M, doe that essentially equals around 3.2M in parallel?

Thanks again

[Transmogrifox]

Glad it's working now.  If you like it then go with it.  Further down is more explanation of the why's but if you like it then your ears are the boss

The purpose of the 100k on the output as well as the resistors to ground on the input (before the caps) are for the purpose of keeping static charge from building up on the audio input and output lines.  If you are using a true bypass circuit around it then you might hear popping noises when you switch on/bypass without the resistors.  Mostly just a best-practice thing.

If you aren't switching this then you could remove the 1M to ground on the input and forget about the 100k.

So...if you are switching, you need only 1 resistor to ground on the input jack side, and only one resistor to ground on the output jack side.

If you aren't switching then you don't really need any on the jack-side (I'm careful to say that because you DO most certainly need something in the location of R21 to keep the FET referenced to ground.

Here's the run-down of parallel resistors to ground if you don't change R21:
Jack-side input : 1M to ground
R21 (shows up as parallel to ground as far as the guitar is concerned):  1M
Mixer input:  470K

1M || 1M || 470k = 242k

Change R21 to 10M,
1M || 10M || 470k = 309k

Change the input-jack-side resistor to ground to 4.7M:
4.7M || 10M || 470k = 410k  <--this is probably the best

[Loose]

I am using true bypass switching. You wrote that if I am switching I only need 1 resistor to ground on the input jack side, but here we have two.
Is there a way to get close to 410k without changing R21, maybe by changing or adding to the resistor to ground on the mixer side? I am getting a new version of the surfy bear board with SMD components, and when i'll add the mixer circuit to it I would rather do most things outboard than replacing SMD components.

Thanks

[Transmogrifox]

The best way to get >400k is to change the mixer resistor values to 1M.  There is a possible noise or stability issue based on gain and phase relationship I saw in simulation at high frequencies, but only if extra parasitic impedances on the perfboard add something that was not in the simulation.

Most likely 1Meg all-around is ok.  The main reason my original mixer design used 100k is because this put the surfy bear output at approximately unity gain at 50% on the pot.  This really isn't a very important relationship and going into 1Meg will be more like going into a following stompbox or high impedance input amp.  It was just nifty is all.

You could also use that unused op amp as a buffer in front of everything as a non-inverting unity gain buffer and then you could get input impedance up to >1Meg.

In reality if you're happy with the sound the way it is then don't split hairs over input impedance.  Your ears are the boss.

[Loose]

I tested it again and noticed the same high end loss that I've had when I breadboarded only the buffer part of the circuit. I built the mixer without the 3.3pF cap so that can't be the reason.

If I understand correctly from the previous posts, these are my options:

1. Increase R21 to 10M
2. Change the three 470k resistors  to 1M
3. Add a buffer in front, which would make the circuit non-inverting. (Would an inverted phase cause any issues when playing with other pedals or when splitting the signal to more than one amp?)

Is that correct?

Thanks

[Transmogrifox]

Option #3 requires no changes to your existing mixer circuit and only adds 1 resistor, 1 capacitor and a wire between pins 1,2 (which you need to do anyway just to keep the unused op amp stable).


It's the same inverting phase as before but it makes it so you don't have to put really large resistors into the feedback to get input impedance up.  I expect the circuit as shown above will sound transparent because simulation does not show any high frequency roll-off until well above audible range.

Hopefully that helps.

[Loose]

Thanks man!

Can you please explain what "only one place" means on R3?

[Transmogrifox]

It means don't duplicate this in off-board wiring.  In other words if you have a 4.7M pulldown on your TBP switch then don't add another one here.

[Loose]

But i should keep r21 on the surfy bear board?

[Transmogrifox]

Yes, R21 is necessary because it's part of the transistor bias.  What I mean by don't duplicate is don't put more than 1 resistor to ground on the INPUT JACK side of the circuit. 

[Gus]

find a 6G15 schematic
Note that the circuit you built does not have the cut off pedal.  The cut off grounds the signal at the recovery stage and also keeps the plate voltage from "moving" so the reverb side of the mixer control does not have a signal.
You circuit always has a signal at the mixer control side connected to the reverb recovery amp

Loose did you try the above?

[Loose]

Gus: No I haven't. Could you please elaborate? It's not very clear, would having a cut off pedal affect how the circuit changes my dry tone when the effect is on? Would it also change the issue in which turning up the wet side of the pot reduces the dry side volume?

Transmogrifox: I've made the changes to the perfboard per your schematic, now there's a bit more high end than the bypassed signal. What should I adjust in order to keep the signal the same?

Thanks a lot!

[Gus]

Looking at the 6G15 schematic
I posted about the cut off as a first simple thing to try
The cut off grounds the output of the reverb pan and also the input to the recovery amp
If this is not grounded a signal is always "moving" the wet side of the mix control if the tone control is not set to min on the 6G15
If you ground the output of the pan and input of the recovery amp the plate "moves" less you need to keep in mind the Rp of the recovery amp triode as the source follower try to wiggle thur the mixer and tone control
Also the output of this circuit looks like it was made to connect to a high input resistance like 1 meg(C1 and C10 being 220pf for 440pF that and C17 being .1uf the mixer being 250K? hard to read)

If you are looking to bypass the effect disconnect the top of the mixer control then short the output of C17 to the output jack wiper

[Transmogrifox]

now there's a bit more high end than the bypassed signal. What should I adjust in order to keep the signal the same?

If it's just a tad brighter then change the 22n series caps to something bigger, like 47n.

If it seems it's in the high "ice-pick" range brighter, then put a cap in the feedback (across the 470k).  Start at 10p to 20p and work your way up toward 50p or so.

Simulation tells me the most likely cause is you're hearing what amounts to maybe 1 dB low cut on the lower strings of the guitar.  If it's anything like simulation the only frequency response variation I can see is within 1 dB anywhere within guitar frequency range.

To me this is inaudible, but some people can definitely detect 1 dB.

The other thing that may be happening is it's just your guitar reacting to a different input impedance than your amp.  For example if your amp is ~500k and you switch to this buffer that gives you 1Meg, then it may result in a perceptible boost in the high range. 

Another thing may be simply your guitar against cable impedance.  If you true-bypass then you have double the length of cable and extra cable capacitance will darken your tone.  This you could compensate by putting 30 pF from the input of the mixer to ground (parallel the 4.7M)

Several ideas.  Hard to know which one because it's hard to know exactly how much bright and what frequency range you are hearing.