Amp splitter help

[trixdropd]

Hello all,

I have built a 5 way guitar splitter / switcher

What I have done, is take the + and the - and connect it to 5 3pdt switches. This allows each switch to get signal. The outputs when turned off are shorted to ground on each channel. Output1  is straight up.
Outputs 2-5 all are transformer isolated via 2 ebtech Morley hum eliminators (2 channels each). What i've done is hard wired the hum eliminators
in place rather than using patch cords. I use my morley bad horsie wah as a buffer before the splitter. As I turn on more outputs, the gain of my signal decreases.
it's fine using 2 amps or 3, but all 5 together is killing my signal. Any ideas?? Rg??? There must be a way to get my guitar into 5 amps without loss.


Issue 2 involves my leds. I ran the positive power to each led, and switched them via breaking the negative connection on each led. When bypassed the leds were still partially lit!! Why??
I then rewired them to break the positive connection to shut them off and it works. Does this cause noise though having the positive voltage running to my switch unable to find its ground until the led is on?

My last question is a general pedal question.
I have found that many times I need to tie my signal ground to my power ground to stop little noise and hum from happening. is this proper? Should there be a resister in between the two? Many of my pedals use relays to switch in loops. How should the grounds be tied together? What if I added a relay to switch an amps channels? Would the ground need to be isolated?

Thanks for any answers you could give.

I'd love to here your opinions R.G., Infact i'd love to learn how to build your splitter on geofex.com instead of buying these morleys...


Jeremy

[R.G.]

It's clear that the signal is getting loaded down, as you already know. You need to take the next step in that reasoning. If it's getting loaded down, then either (a) the "buffer" isn't able to drive the load or (b) the load is so big that no reasonable buffer can drive it or both of those at the same time.

You say:

I use my morley bad horsie wah as a buffer before the splitter.

I'm not all that familiar with the BH wah. Do you know that it's a decent buffer? Some pedals are not, and can be loaded down easily.

Outputs 2-5 all are transformer isolated via 2 ebtech Morley hum eliminators (2 channels each). What i've done is hard wired the hum eliminators
in place rather than using patch cords.

This one is kind of part B. I don't know the input impedance of the hum eliminators. If they're just transformers, it's unlikely they are high enough impedance to not load down the driving source.

There must be a way to get my guitar into 5 amps without loss.

There is. But it will involve making sure of your buffer's ability to drive whatever the five loads are.

Issue 2 involves my leds. I ran the positive power to each led, and switched them via breaking the negative connection on each led. When bypassed the leds were still partially lit!! Why??

Do you really have an LED in series with each LED and just not show the resistors, or do you really have just LEDs tied to a power supply? LEDs are intended to work with a series resistor no matter how you switch the power. Put a 4.7K in series with each LED if you don't already have them.

Did you verify that the switch connections really make contact the way you think? A 3PDT is a square array, and it's easy to get it turned 90 degrees from the way it actually works. Ohmmeters will tell you.

I then rewired them to break the positive connection to shut them off and it works. Does this cause noise though having the positive voltage running to my switch unable to find its ground until the led is on?

No.

I have found that many times I need to tie my signal ground to my power ground to stop little noise and hum from happening. is this proper?

In most pedals, it's unavoidable. Either the + or - side of the power supply is tied to signal ground inside the pedal. And it must be so, at least one place, or no power supply current can flow.

Should there be a resister in between the two?

No.

Many of my pedals use relays to switch in loops. How should the grounds be tied together?

With solder. Seriously, you have no options. All of the signal grounds must be tied together at some place or you'll get intractible hum and noise.

What if I added a relay to switch an amps channels? Would the ground need to be isolated?

As long as the relay coil is isolated from the amp footswitching arrangement, it makes no difference whether the coil switching arrangement is tied to amp ground or not. If what you're asking is whether the relay switching side needs to be tied to signal ground, no it doesn't, and it's a little cleaner if you use insulated bushing jacks for the cords that go to the amp, and leave the contacts side of the switching that connects to the amp unconnected out in your remote box. That lets the amp decide whether one or the other side of the switch voltages are tied to ground or not.

I'd love to here your opinions R.G., Infact i'd love to learn how to build your splitter on geofex.com instead of buying these morleys...

General Guitar GAdgets sells a PCB and kit for doing that.

[trixdropd]

I'm not all that familiar with the BH wah. Do you know that it's a decent buffer? Some pedals are not, and can be loaded down easily.

I can't say for sure. It says it has a 'true tone" buffer circuit. I built an a/b/y with the hum eliminator on output b and it loaded down as well, but the wah
inserted in the chain before the splitter worked. It's just not enough to handle the load of 4 hum eliminators. The hum eliminators are each a complete
circuit with a transformer and some resisters and such. Do you think if there were a few boss style pedals
before the splitter that may fix the loading? What about putting a pedal in between a few of the isolated outputs and their respective amps? Any insight or links
on buffering you can provide would be greatly appreciated.

I don't know the input impedance of the hum eliminators. If they're just transformers, it's unlikely they are high enough impedance to not load down the driving source.

the specs of the hum eliminator is this:  Due to them using trs jacks, I wired ring to sleeve as I used T-S jacks. Without doing this there was major signal loss.
Input Connectors: 2 x 1/4" TRS jacks
Output Connectors: 2 x 1/4" TRS jacks
Input/Output Impedance: 10K Ohms
Maximum input: 8V RMS
Maximum source impedance: 1K Ohms
Minimum load impedance 10 Ohms
Frequency response 20HZ to > 20kHz (- .05dB)

keep in mind I'm feeding 4 of these circuits in parallel along with another output.

Do you really have an LED in series with each LED and just not show the resistors, or do you really have just LEDs tied to a power supply?
LEDs are intended to work with a series resistor no matter how you switch the power. Put a 4.7K in series with each LED if you don't already have them.

They have resisters on each led. It's indicated on my "drawing' (and I use that term loosely). The switches are orientated right because the circuits work properly
besides getting loaded down. I did also verify with a multimeter. I have since rewired the leds so that the positive gets switched and the ground it tied together but
I wanna know what was causing the leds to be dimly lit when bypassed.

I then rewired them to break the positive connection to shut them off and it works. Does this cause noise though having the positive voltage running to my switch unable to find its ground until the led is on?

No.

The reason I ask is because in my relay based builds I found that I was getting a slight buzzing noise if I switched it in and out via breaking ground.
I wired them up to break the positive and the noise went away. I figured switches would work the same. If I can leave these 5 switches the way I
rewired them now (breaking positive) that would be cool.

If what you're asking is whether the relay switching side needs to be tied to signal ground, no it doesn't, and it's a little cleaner if you use insulated bushing jacks for the cords that
go to the amp, and leave the contacts side of the switching that connects to the amp unconnected out in your remote box. That lets the amp decide whether one or the other side of the
switch voltages are tied to ground or not.

I'm not following you all the way here. I use a marshal style plastic jack isolated from my pedals chassis. I use the following circuit with a relay to control it. I try to keep the amps ground
completely separate but wondered if there was a better way.

I'd love to here your opinions R.G., Infact I'd love to learn how to build your splitter on geofex.com instead of buying these morleys...

General Guitar GAdgets sells a PCB and kit for doing that.

I saw it and may buy it. One of the issues for me is I need to be able to have a modular transformer circuit in order to produce a/b/y pedals. That pcb is done as such that I can't make two out of it.
These morleys I literally cut in half if I only need 1. They just cost way too much for what I need em for. I may but one anyways to learn the details of the design anyways though. What I have been looking
for is a part list for the build.


Anyways R.G.,

I have to salute you with a HUGE "hats off" to you. I have made and sold several remote controlled true bypass loopers and learned a ton of stuff thanks to AnalogMike sending me a link to your "A remote Indicating Bypass"
article. I have been searching for a way to talk to you for over a year now... It's a huge honor to talk to a what I consider a "Legend" in the field. Your contributions are greatly appreciated. I'm sorry to come off as such a noob.
I am one of course compared to you, but I am making a bunch of things that work and hopefully with a little more insight, I'll make a bunch more.

Thanks a million,

Jeremy

[trixdropd]

I cant seem to edit my last post.

I did some more reading on your Hum- Free A/B/Y. I attempted to draw out the circuit (I'm not good at reading schematics, i'm good at seeing what hooks up where)to construct a single channel of the transformer isolation
by following the circuit board hookup instructions at GGG.
I see the TL072 tie into each other as well as with both transformers. I don't see how I can half this design. Is it possible?

Another question I had; Is there a need for any resisters on the input? I noticed my morley a/b/y has them for some reason and i wonder the purpose.

Thanks again.

Jeremy

[Auke Haarsma]

TL072 is a dual opamp. So two opamps in one. That may confuse you. But you can also use it to your advantage.

I quickly edited RG's schematic to show what I think you need to use just one output.


You need the part in green and the powersection (bottome left).

The TL072 can serve both opamps in the schematic. So you need just one TL072, instead of the orignal 2. Not sure about the resistors around the LED-section (middle).

RG, if you want this removed, I'll do so gladly. Removed the values, just to be precautious.

[Andi]

I'd think you'd only need one op-amp stage to drive the tranny.

[R.G.]

Input/Output Impedance: 10K Ohms
...
Maximum source impedance: 1K Ohms
keep in mind I'm feeding 4 of these circuits in parallel along with another output.

Given that info, I highly suspect that these are a version of the Jensen transformer application for guitar isolation. Yes, they are too low an impedance to drive many of them with a wah pedal. They need a buffer driving each of them.

They have resisters on each led. It's indicated on my "drawing' (and I use that term loosely). The switches are orientated right because the circuits work properly
besides getting loaded down. I did also verify with a multimeter. I have since rewired the leds so that the positive gets switched and the ground it tied together but
I wanna know what was causing the leds to be dimly lit when bypassed.

OK. I'm not sure. Well, the LEDs lighting shows that there is some leakage of current from the - side when the switches are nominally off, but it's not apparent what that is from your drawing. Most likely it's something that is in your circuit but not in the drawing; all schematics and drawings are our mental approximations to what we think we see. I've certainly made enough mistakes doing them.

I then rewired them to break the positive connection to shut them off and it works. Does this cause noise though having the positive voltage running to my switch unable to find its ground until the led is on?
The reason I ask is because in my relay based builds I found that I was getting a slight buzzing noise if I switched it in and out via breaking ground.
I wired them up to break the positive and the noise went away. I figured switches would work the same. If I can leave these 5 switches the way I
rewired them now (breaking positive) that would be cool.

That indicates - but I can't finger exactly the issue from the data at hand - that you're actually breaking the entire ground, or perhaps causing your power supply to add noise when they're all off. I would probably have to do hands-on tinkering to find it. If you're OK with breaking positive (which literally means the ground side is connected all the time) then that's a good solution.

I'm not following you all the way here. I use a marshal style plastic jack isolated from my pedals chassis. I use the following circuit with a relay to control it. I try to keep the amps ground completely separate but wondered if there was a better way.

No, there is no better way. Completely separate grounds from the amp is a virtue in this application.

I saw it and may buy it. One of the issues for me is I need to be able to have a modular transformer circuit in order to produce a/b/y pedals. That pcb is done as such that I can't make two out of it. These morleys I literally cut in half if I only need 1. They just cost way too much for what I need em for. I may but one anyways to learn the details of the design anyways though. What I have been looking for is a part list for the build.

Your reply indicates that you intend to produce - and possibly sell - this setup. That being the case, you're in a different ballpark in terms of money than a DIYer working for their own amusement. J.D. Sleep at GGG is a pretty decent PCB layout guy these days. Why don't you call or email him and ask him to lay out a custom board for you with only one channel of isolator on it? As we used to say when I was working on inter-corporation IP licensing, there is always a number which will make the deal work. May be too high a number, but the number always exists and you never know until you ask.

Thanks a million,

You're welcome. Good luck with it.

[trixdropd]

Thanks for all the help fellas.

Is there any buffer suggestions for me if I have to add 4 of them to my splitter?

[Andi]

I've always buffered isolation transformers with a standard op-amp follower.

[trixdropd]

I've always buffered isolation transformers with a standard op-amp follower.

I hate to beg but can you provide a link an example of one?
I looked at this; http://www.eecs.tufts.edu/~dsculley/tutorial/opamps/opamps5.html and it looks promising. I don't see any real detail there though.

http://johannburkard.de/blog/music/effects/Build-a-buffer.html Would this be too much work?

Thank you very much for your assistance.

Jeremy

[Andi]

http://www.muzique.com/lab/buffers.htm

[trixdropd]

http://www.muzique.com/lab/buffers.htm

So an opamp, 2 capacitors, and a resister?? looks pretty simple. Thanks A lot for the help Andi and everyone else.

jeremy

[trixdropd]

Ok, So here is where I am at the moment.


So essentially, Input splits to output 1 (straight to output) and then feeds the buffer I made at the bottom of this page http://www.muzique.com/lab/splitter.htm which has 4 buffer channels each feeding a morley hum eliminator on outputs 2-5.

Everything is working pretty good. The original problem of getting loaded down when adding the hum eliminators to the circuit is gone. the problem now, is slight hum on all the buffered channels. It wasn't there before and i'm only hooking up 1 channel at a time. I recorded 5 tracks into the computer into amplitube, after doing statistics, the buffered, transformer isolated channels are average rms 1db lower than the naked channel. What can I do to lose the hum? It's not A lot of hum and isn't a ground loop. It's certainly relating to the buffer curcuit. One thing that was strange in the layout is they didn't specify polarity of c1, so I assumed positive went top like the others. My cap wasn't labeled but had a small indent on 1 side i assumed to be the +. Thanks for any help you can give. 

[trixdropd]

Bump...   

[trixdropd]

Alright, time to make this thing work once and for all...

I wan't to adapt the RG Keen hum free A/B/Y to a 5 channel splitter, 4 of them being transformer isolated.

Rg, If this needs removal, say the word and it will be done. i removed all values...


See any issues anyone? Suggestions??

[Pablo1234]

The schematic you have thier is exactly what I would do. One thing though, you dont need 2 resistors at the input of the second stage buffers, just sum them together.if its a 100k and a 50 k use a 150K instead.

[trixdropd]

The schematic you have thier is exactly what I would do. One thing though, you dont need 2 resistors at the input of the second stage buffers, just sum them together.if its a 100k and a 50 k use a 150K instead.

Cool man, thanks! So you're saying use a 20k resister in place of the pair of 10k's at the input stages of each channel?

[Pablo1234]

yes exactly, the reason the other schematic had 2 was it was being taped between the resistors and you are not needing that in yours so just use the sum of the 2.

[trixdropd]

yes exactly, the reason the other schematic had 2 was it was being taped between the resistors and you are not needing that in yours so just use the sum of the 2.

That makes perfect sense. thanks. I'll design a pcb and put it up for scrutiny before I build it...

[trixdropd]

I have another question. I'm finishing up the pcb for my build. 4 channels will be transformer isolated, and 1 will not. I will post the pcb for anyone who'd like to have a go at it when it's done.

Now i've ommited the switching from rg's layout opting instead for traditional switching (after transformer, before output jack). The schematic doesn't appear to address the maxx1044 power layout. I'm referencing the ggg hum free a/b/y and the schematic. I'm not sure if I can ommit the transistor before the maxx1044. Is it necessary?