News:

SMF for DIYStompboxes.com!

Main Menu

Buffers again

Started by VPIF, February 21, 2013, 08:17:31 AM

Previous topic - Next topic

VPIF

Hi.

Just about ready to order parts for my ambicious (to my skill level) splitter/mixer. It will be stereo, but for this purpose stereo means 2 x mono. I cannot seem to find any good information about splitting/buffering line level signals online, so I have three short questions, which are kind of three different versionsof the same question:

1. Splitting the effects loop (line level) signal. I will need to split the signal five ways. Would it make sense to buffer the split, buffer the returns, or buffer both the split and the returns?

2. I have drawn (copied) my first schematic (I know there are small errors - the IC is supposed to be TL072 etc.) It is the AMZ Super Buffer. Should I use v1 or v2 to buffer the split of the input (if any)? Please reference the drawings below.

3. I am planning on using the GGG mini mixer to mix the returns. Should I use the passive version, the buffered unity gain version, or the buffered version with (i think) +10dB gain?

Thanks :)





PRR

  • SUPPORTER

Keppy

Quote from: VPIF on February 21, 2013, 08:17:31 AM
Hi.

Just about ready to order parts for my ambicious (to my skill level) splitter/mixer. It will be stereo, but for this purpose stereo means 2 x mono. I cannot seem to find any good information about splitting/buffering line level signals online, so I have three short questions, which are kind of three different versionsof the same question:

1. Splitting the effects loop (line level) signal. I will need to split the signal five ways. Would it make sense to buffer the split, buffer the returns, or buffer both the split and the returns?
Buffer the sends, or the impedance of the following devices will appear to be about 1/5th of any one of them alone.

Quote2. I have drawn (copied) my first schematic (I know there are small errors - the IC is supposed to be TL072 etc.) It is the AMZ Super Buffer. Should I use v1 or v2 to buffer the split of the input (if any)? Please reference the drawings below.
Use v1. v2 has no output caps, and thus would put 4.5v on all your outputs. Or just do it the simpler way that Paul suggested. BTW, I hope one of the errors you're aware of is that the opamp inputs are reversed in the drawings (including Paul's).

Quote3. I am planning on using the GGG mini mixer to mix the returns. Should I use the passive version, the buffered unity gain version, or the buffered version with (i think) +10dB gain?
I haven't looked at it, but a buffered mixer should work better for devices of differing output impedance, and I doubt you'll need the gain.
"Electrons go where I tell them to go." - wavley

R.G.

Quote from: VPIF on February 21, 2013, 08:17:31 AM
1. Splitting the effects loop (line level) signal. I will need to split the signal five ways. Would it make sense to buffer the split, buffer the returns, or buffer both the split and the returns?
Audio signals do not in general need split in the same way that RF signals need split. Audio signals need to be buffered sufficiently so that any paralleled loads can be driven without loss.  There are few special exceptions, but this is the general idea. In the context of audio, "buffer" means "make the signal able to provide enough current to any expected load so the signal is not loaded down by what's connected to it in any significant way."

In this case, whether you need to buffer the split or buffer the return or both depends on something that's not shown - what's being driven by the outputs and mixed by the returns. If those are typical guitar pedals with an input loading (impedance) of 1M or so at audio frequencies, either minimal buffering or perhaps no buffering at all is needed. If it's a line level input, that does not come from a guitar, so you might need no buffering at all with pedals being driven. If what's being driven is audio line level stuff, most consumer stuff is about 10K input impedance, and you might need one opamp to drive up to about 5-6 inputs.

If what's being driven is 600 ohm line inputs or - gulp- speakers, then you may need a whole lot more buffering than this schematic shows.  :icon_lol: 

The things being driven matter to the buffering.

Splitters as such are only needed where the loading on one output can possibly be so much that it drags down the other outputs, and in that case a real splitter to limit any effect on other outputs may be needed. This is the reason RF generally needs splitters (albeit in a different way).

As to buffering the returns - maybe. Again, it depends on what is doing the returning. If it's able to drive the mixer input without much loss, you can go passive. Since we don't in general know what is going to be connected, it might make sense to buffer each return at the mixer, just in case. Opamps are cheap. But for most audio equipment and most pedals, a passive-input mixer would be fine.

I guess this is another ways of saying that the devil is in the details. It's not possible to provide good buffer/splitter/return/mixer designs without some idea, albeit crude, of what's being driven between the sends and returns.

Quote
2. I have drawn (copied) my first schematic (I know there are small errors - the IC is supposed to be TL072 etc.) It is the AMZ Super Buffer. Should I use v1 or v2 to buffer the split of the input (if any)? Please reference the drawings below.
Your schematic shows the feedback from the output to the positive input. That won't work at all. I believe you have the + and - inputs swapped.

Beyond that, I think you're spending a whole lot of opamps on some unnecessary buffering. As PRR notes, performance is likely to be as good with one opamp doing the driving and some minimal resistance between the buffer and each output. But then, as I said, opamps are cheap, and they usually don't hurt until you get hiss buildup, oscillation, or some other audio oddity.

I sometimes find the naming of things on the internet funny. There is a persistent trend among beginners to pick things named "ultra" or "super" for the name, as opposed to what is necessarily done. The "Ultra Clean Power Supply" pops up over and over, and it's the same schemo as many other circuits that are named differently. I guess guys like shiny things.  :icon_biggrin:

Quote3. I am planning on using the GGG mini mixer to mix the returns. Should I use the passive version, the buffered unity gain version, or the buffered version with (i think) +10dB gain?
As to buffering the returns - maybe. Again, it depends on what is doing the returning. If it's able to drive the mixer input without much loss, you can go passive. Since we don't in general know what is going to be connected, it might make sense to buffer each return at the mixer, just in case. Opamps are cheap. But for most audio equipment and most pedals, a passive-input mixer would be fine.

If you need the gain, that's another issue.
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.

Gurner

#4
Quote from: R.G. on February 23, 2013, 11:58:30 AM

Your schematic shows the feedback from the output to the positive input. That won't work at all. I believe you have the + and - inputs swapped.


Likewise...

Quote from: PRR on February 22, 2013, 02:10:07 AM


that's an oscillator  ;)

VPIF

Thank you so much for your replies!

As of present I know exactly what will be hooked up to the loops:

TC Electronic D-Two
Input impedance:     13KOhm
Output impedance:    40Ohm

Korg SDD-1000
Input impedance:     6,6Kohm @ -50dBm
                               220KOhm @ -25dBm
                              47KOhm @ 0dBm
Output impedance:    1KOhm

Roland SRE-555
Input impedance:     5Kohm @ -50dBm
                               50KOhm @ -30dBm
                               500KOhm @ - 10dBm
Output impedance:    More than 5Kohms (actually says so in the manual)

Boss RT-20
Input impedance:     1MOhm (Now, this might be a problem?)
Output impedance:    1KOhm
Recommended load impedance: 10KOhms or greater.

The effect loops impedances are as follows:

Amp 1 (Fender HRDL)
FX-loop send:   1,5KOhms max
FX-loop return:   54KOhms

Amp 2 (Laney Cub head)
FX-loop send:   Doesn't say in the manual. Only says "Line level" and "-10dBV nominal".
FX-loop return:   Only says "-10dBV nominal".

I would like to build this as simple as possible, without compromising the audio quality. I would build it as a completely passive unit, if that'll work.  I realize that I should probably include phase-switches in the design. I hope this helps.

P.S. Totally agree on the "Super" buffer issue, but through searching the forum I found the schematic, and I seemed to fit my needs. Any suggestions for other suitable buffers (I prefer ones I can find a PCB-layout for) are welcome)

Thanks again!

VPIF

Finally got my rack enclosure today. Just about to order the parts for the guts of this thing, and I guess I'm going with the AMZ super buffer and the unity gain GGG mini mixer, if no one has any better ideas.

wavley

Quote from: VPIF on February 23, 2013, 02:26:09 PM

Roland SRE-555
Input impedance:     5Kohm @ -50dBm
                               50KOhm @ -30dBm
                               500KOhm @ - 10dBm
Output impedance:    More than 5Kohms (actually says so in the manual)



Do yourself a big favor and put a JFET buffer immediately after the input jack, no more changing impedance depending on your setting.  I did this to both my RE-101 and RE-501 echoes and it's one of the best things I've ever done for my tone.  Billy Zoom (the guitar player of X) should get credit for bringing this to the interwebs.

Also, Billy Zoom gets credit for this schematic, it's for the 201 so it's not connector 40 in the 501 and I don't have the service manual in front of me to tell you what it is.

New and exciting innovations in current technology!

Bone is in the fingers.

EccoHollow Art & Sound

eccohollow.bandcamp.com

VPIF

Quote from: wavley on March 26, 2013, 04:17:36 PM
Do yourself a big favor and put a JFET buffer immediately after the input jack

Thanks for the tip!

Would this buffer work just as well immediately before the input jack? I don't really want to tinker with the original unit.

wavley

#9
Quote from: VPIF on March 27, 2013, 03:47:39 AM
Quote from: wavley on March 26, 2013, 04:17:36 PM
Do yourself a big favor and put a JFET buffer immediately after the input jack

Thanks for the tip!

Would this buffer work just as well immediately before the input jack? I don't really want to tinker with the original unit.
In the case of the 501/555 it will because it doesn't have the crappy bandpass filter at the front end and it certainly helps any of the other models.  Of course it's a completely reversible mod because you can build it right on the input jack and move the input wire to the output of the buffer, plus you have more headroom because you have 17 volts available to you vs the 9 volts we usually use.  If you use a 2SK246 FET going into the HA1457 on a 501/555 then you almost have a Boss FA-1 without the tone stack.

edit: If you're going to own a Space Echo you better get used to tinkering with it or you're going to spend a lot of money paying guys like me to fix it, they need pretty constant maintenance to keep running
New and exciting innovations in current technology!

Bone is in the fingers.

EccoHollow Art & Sound

eccohollow.bandcamp.com

VPIF

Yeah, the Roland has luckily been working great the years I've had it, but I guess I will have to pay someone to service it sooner or later.

This project got quite delayed, but my rackmountable enclosure has finally arrived, and electronic components are also on the way. I will probably be back with some pictures and whining once I get building.

scratch

VPIF, just out of curiosity, what/where did you order for a rack enclosure?
Denis,
Nothing witty yet ...

VPIF

#12
Hi.

Ended up ordering it from a webshop here in Norway.

EDIT: http://evenstadmusikk.no/index.php?main_page=product_info&cPath=343_549&products_id=16420