An XLR Challenge to All !!!

[zpyder]

boggietube-

You've layed out exactly what I'm trying to do.  I started one of the "two other threads concerning this subject that are active right now" and I'm very excited to find that other people are looking into this as well.  My intended use is this: My Boss RC-20XL Loop Sampler has 2 inputs, guitar level, and mic level.  In the live stage setting I'd like to split my vocal mic signal, send one balanced XLR out to the PA through the snake as usual, and send the other (for which I would have a bypass switch do it's not always on) I would send an unbalanced 1/4" signal to my RC-20XL and eventually to my amp.  The way I figure it I should be able to "harmonize" with myself by sampling vocal loops.  Albeit they'll come out of my amp (and then back into the PA through my amp mic), I believe this could work.

Though you want to use your differently, the circuit you described and the one I'm trying to create appear to be the same thing!

My first thought that came to my mind was to use the splitter from ROG's Splitter-Blend, on BOTH the hot & cold XLR input pins (IE two splitters) dumping one of the colds and using it's hot for the FX send, and using the other pair of hot & cold as balanced XLR out.

Gilles-
I've added the splitter from ROG's Splitter-Blend to your schematic:


I have taken some liberties here, so I sincerely hope that I am not offending you, Gilles, or ROG by posting this material here.  If I am, I hope that I am contacted and told "bad bad...bad!!  stop it!!"

Effectively what's happening here (if I understand it correcly) is that first we go into the splitter, which creates two outputs, both with a gain of 1 relative to the input signal (same volume), then one goes directly out to our unbalanced 1/4" jack.  The other signal goes to our 5532 circuit (Gilles) which once again creates two output signals, but in this case one is inverted.  Viola!  We now have both the hot & cold signals we need for our balanced XLR output... Two outputs, one balanced, the other unbalanced.     However, the PA ground is still connected to this ground here.... 60 cycle hum?

I will be building this circuit soon as well... Although I don't know which XLR jacks to buy.  Anyone have a Mouser/Other reference for obtaining XLR male & female jacks?

One question remains for me: Ground isolation.  It seems that the XLR out must be isolated since this circuit is active... Is there a cost-effective transformer solution?

cheers!!
zpyder

[Pushtone]

My intended use is this: My Boss RC-20XL Loop Sampler has 2 inputs, guitar level, and mic level.  In the live stage setting I'd like to split my vocal mic signal, send one balanced XLR out to the PA through the snake as usual, and send the other (for which I would have a bypass switch do it's not always on) I would send an unbalanced 1/4" signal to my RC-20XL and eventually to my amp. 
zpyder

Ah, I must have missed that part about the RC-20XL.
Its a good idea to use one mic for FOH vocals and vocal sampling.
A passive switch box could make that very EZ.

Its important to note the RC-20XL already HAS a mic input preamp.
Its probably a circuit very much like the one posted above.
It even has its own level control so you have separate levels for your guitar and mic.
So for the RC20XL, you don't need another preamp, you have one.

First
Check your owners manual and look for wiring reference
for the input jacks. Sometime manufacturers use the Tip Ring Sleeve 1/4" jacks for balanced lines.
You might not need to unbalance but just switch it. Look for "TRS mic input" as well as the word "balanced".
BOSS USA makes you reg to DL manuals. That sucks.

Second failing that, make a passive switch box.
For a dynamic mic all you need is a DPDT switch, one XLR-F input, one XLR-M outputs, one 1/4" jack and a box.
To unbalance the mic signal, connect pin 2 to the 1/4" tip and pin 3 to ground (pin 1 is ground).
You now have unbalanced, low-Z, mic-level signal. Just what the RC20XL wants to see at it input.
Route pins 2 and 3 through the DPDT switch.
Tie all grounds together (this is where transformer isolation could block ground loop hum)
One switch state just passes pins 2 and 3 to the XLR out and to the PA.
The other switch stage sends pins 2 to the 1/4" tip and pin 3 to any ground node, like the 1/4" sleeve, and to the RC20XL input.

I would use a momentary switch to avoid ever getting confused.
Step and hold and you are recording, step off and you mic is live to the PA.

The fact that the RC20XL has a mic preamp makes it unusual.
You would need an active circuit if you wanted to connect standard effect pedals between the mic and the RC20XL.
But then you would use the guitar input on the RC20XL, wasting its mic preamp, and you still want your guitar connected, don't you.
But the solution there is easy too. Connect effect pedals after the RC20XL and put effects on its output.
You could connect the whole chain to your guitar amp and mic it or into a DI box for recording.

I think I might have posted a link to some commercial product that did just this. Can't remember now.

[Gilles C]

No problem Zpyder, I wanted to suggest that it could be a project to be tried and modified by whoever could be interested, but wasn't sure if someone could be interested.

So, go ahead. I want to compare it on my side with the discrete version from Jensen that I built a while ago. I bought a XLR connector today, and added it to a test jig I am builing to test my circuits. That's what was missing to enable me to test the Jensen circuit with a balanced output.

I have both a Peavy Solo amplifier that has an XLR input (normally for mics), and a Behringer MIC200 preamp that had can be padded down 20 db to lower the input signal. That should give me a good idea of the result.

Gilles

[markusw]

Here are some pics of my box I finalised yesterday eve. Will post the layout soon.







The schem is basically the same as in http://sound.westhost.com/project35.htm fig 3. I just removed the line/speaker switch + resistors as well as the 1 k pot. Now there is just a 100k log pot at the input to control the level of the balanced signal (like in Gilles' schem).

Regards,

Markus

PS: the huge non-polar caps were the cheapest solution for me to get the 22µ (withou using rough foil non-polars  ).

[Gilles C]

Nice...

And you're fast.

Gilles

[markusw]

And you're fast.

Gilles

Hm, not really. I started already one month ago

Markus

[blackbirdneo]

sir will that work well with a bass guitar or do you think i have to modify it like all the other guitar pedals? thnx

[markusw]

sir will that work well with a bass guitar or do you think i have to modify it like all the other guitar pedals? thnx

It works with bass guitar. That's what I need it for

Layout/project file will be posted within the next days.

Markus

[Gilles C]

I was checking the schematics again, and after reading Markus's mention of the bipolar capacitors, I wondered about the need to use such capacitors. Some circuits use them, some don't.

Even Jensen use them in their schematics... but not all the time.

And I have some other schematics that don't use them. 

Any opinion?

Gilles

[markusw]

I was checking the schematics again, and after reading Markus's mention of the bipolar capacitors, I wondered about the need to use such capacitors. Some circuits use them, some don't.

Even Jensen use them in their schematics... but not all the time.

And I have some other schematics that don't use them. 

Any opinion?

Gilles

I suppose if you intend to use the box with phantom power or battery it needs non-polars. If you're going to use it just with battery (with no phantom power on the XLR jack) you can use polars in the regular way. If you intend to use it with phantom power only you probably might use polars the other way round since the opamp's outputs will always be lower than the XLR (48V).
Probably, this is complete rubbish but it was my interpretation.

Markus

[Gilles C]

Thanks. I also thought that the caps may be better the other way because of the 48V Phantom Power. I drew them as in the application note from Jensen. But I realised they didn't use or even added the Phantom Power in the schematic, and almost always use a transformer (of course...).

http://www.jensen-transformers.com/an/an003.pdf#search=%22balanced%20input%22

Hum... Maybe I should use 2 x 100uF back-to-back. I'm not sure I could find non-polar caps easily.

Gilles

[markusw]

Hum... Maybe I should use 2 x 100uF back-to-back. I'm not sure I could find non-polar caps easily.

If you gonna use two polars back to back you might add a 10k from the junction to gnd. I saw this the first time in a Bartolini bass preamp. They used tantals, though. IIRC polars give a better sound when there is a voltage across them. Don't know if it really improves sound, but it might be worth a try. On the other hand it might not be necessary with aluminium electrlytics?

Edit:  actually I just ralised that the additional resistor would only be a benefit if you use phantom power 

Markus

[zpyder]

I would use a momentary switch to avoid ever getting confused.
Step and hold and you are recording, step off and you mic is live to the PA.

Thanks for chiming in PushTone... that was very informative.  And you're probably right, I didn't even think of the RC20XL's mic input stage...

I would like to take this one step further, though.  The major alteration I have in mind to the plan you laid out there is that I would like to include an active splitter circuit.  The reason is that while I want to be able to route my vocal signal through my RC20XL, I want it to simultaneously still be piping out to the PA.  I have in mind to place a splitter in there which is always running, one output always going to the XLR/PA, and then just a switch that either connects the other output to ground, or to my 1/4" out jack... So I am looking at an active circuit.

What transformer should I be looking at?

zpyder

[Gilles C]

I did a little research about isolation transformers, and the closest i could find at Mouser is this one

http://www.mouser.com/search/refine.aspx?Ntt=838-MET-50%09

I was looking for a 20K:150 (12:1) transformer, and found nothing. I wonder if there is one that would fit the job better without being too expensive. Still searching...

The IC version of the DI box circuit I want to try is not ready yet, so I wired the discrete version I already had and never tried with my guitar.

Here it is on my audio test jig (that still need a few more things...)



I will try to test it this week-end.

Gilles

[markusw]

For those interested in a XFMR-less DI box here's the project file:

http://www.aronnelson.com/DIYFiles/up/DI_box_project_file_v1.pdf

At the moment I'm running it with a TL072 which gives me a V+ of about 20V when used with 48V phantom power. Maybe I'll test it with a MC33178 which consumes about 20-30% current compared to the TL072 and thus should allow for 24V V+, but at the moment I'm happy how it works.

Have fun

Regards,

Markus

PS: Please report if you find any bugs in the file.

[boogietube]

You guys rock!!! The challenge has been a success !!! Now many of us can add a D.I. to effects without the hassle of a transformer.
Special thanks to Marcusw, and I'm sure Gilles C will have an IC version soon! And all the rest for some amazing input.
I have one question for marcusw , though.
Where do you get those crazy non-poarized caps and the board mounted pot?

Sean

[Gilles C]

I like the mult-caps job that you did for the NP caps.

Very inspiring...

Last night, I read about the benefits of doing this while searching for nonpolar Caps on the web.

Here is this article by... Robert Keeley

http://www.musicianshotline.com/archive/monthly/pedal_power/0903.htm

Yesterday, I flipped the 2 capacitors C1-C2 on my schematic to have the positive lead on the +48V  side, but later, I realised that if there is no +48V as in my Peavy amp, it would make no good. So the NP caps seem the way to go.

Mouser has somew NP caps

http://www.mouser.com/search/Refine.aspx?Ne=254016+1447464&N=1323038%20254310&Ns=P_SField&RefType=Header


Gilles

[Gilles C]

A little note about the transformer I mentioned at Mouser which was had a  10K:125 ratio.

Well today, I bought one locally. It's a Hammond 107N, with a ratio of 10K:600, but with a center tap. It's cheaper than the Tamara at Mouser, and has the same freq response.

http://www.hammondmfg.com/107.htm

The 107V would have been closer to what I was looking for, but the only one available locally was the 107N, so this is the one I will use for my circuit.

Btw, the only thing bugging me about these cheaper transformers is the freq response which is 300Hz-20Kz

Gilles

[Gilles C]

On a real last note about transformers, I checked and the Mouser 42TM018 has a freq response of 300Hz-3.4KHz, so it shouldn't be a problem using the Hammond transformer...

Gilles

[markusw]

Last night, I read about the benefits of doing this while searching for nonpolar Caps on the web.

Here is this article by... Robert Keeley

http://www.musicianshotline.com/archive/monthly/pedal_power/0903.htm

Thanks for the link!

I have one question for marcusw , though.
Where do you get those crazy non-poarized caps and the board mounted pot?

Sean

The caps were the "cheapest" non-polars I could find. The pot is a Bourns 91 conductive plastic pot. They are pretty expensive but I prefer PCB mounted pots. Also I like the smooth feeling when turning them. Got all of it at rs-components here in Austria.

Markus