Passively splitting a pickup's signal output?

[peskywinnets]

I'd like to split my guitar's output into two - but I'd like to do it passively...which will likely bring some form of transformer into play (?)....now I know diddley squat re the spec of transformer needed to do such a task....but I'm pretty keen that any transformer I use mustn't colour the sound of the original guitar signal....any top tips on the characteristics/specs of the transformer needed?

I'm figuring here that you can get something for nothing - ie if the signal has been split into two...then something must give - what is it? (original signal strength? ....will the transformer primary will load the signal down etc)

[Steve Conner]

If you want to do it passively then the last thing you want is a transformer. A guitar signal is very high impedance, so it's difficult to find a transformer that won't colour it. You're going to Sowter or Jensen, and paying hundreds of dollars for a transformer dedicated to guitar isolation.

The only use of the transformer is to break ground loops, which stops the hum that you sometimes get when you try to drive two amps from one guitar. But try it first without the transformer, maybe your setup won't hum.

What gives is the signal strength, right enough. A classic Fender amp has an input impedance of 136k, so two in parallel have 68k, loading your guitar twice as much. This would be true whether you used a transformer or not: only an active buffer can fix it.

Some modern amps have 1M input impedance to get a hotter tone from the pickups, and some of the old ones will go up to 1M if you stick an empty jack plug in one of the other inputs.

[peskywinnets]

The only use of the transformer is to break ground loops, which stops the hum that you sometimes get when you try to drive two amps from one guitar. But try it first without the transformer, maybe your setup won't hum.

Or to allow two disparate/incompatible circuits (DC level wise) to be connected (where caps aren't ideal)

I do need the solution to be passive - if you're saying don't use a transformer...then what do you reckon I should I use?

After a bit of Googling, I see Jensen have an application note for passively splitting a guitar signal...

http://www.jensen-transformers.com/as/as013.pdf

With the spec of that transformer here...

http://www.jensen-transformers.com/datashts/11p1.pdf (therefore I have a 'spec template' to start searching with!)

anyone familiar with or dabbled using similar transformers?

[pdf64]

'A classic Fender amp has an input impedance of 136k'
I'm sure that's not right Steve.
Plug into the '1' input only of almost any fender since the 'D' series and the input resistor is 1M. The switching system on the input jack sockets will parallel up the 68k resistors, in series with the signal. There is effectively no loss of signal because the 34k resistance is after the 1M.
Plug into the '2' input only, and the jack switches put the 68ks (now connected in series to form 136k) in parallel with the input. The signal to the tube grid is from the junction of the 68k resistors, so forming a potential divider, with only half the input signal getting to the tube (ie -6dB pad).
Plug into both inputs and the input impedance at one input is dependent on the impedance at the other input, but is never less than 136k. Pete.

[peskywinnets]

Also, I'm thinking here that since the primary side of a transformer is not a resistor - & has inductive reactance - so will present a much higher 'AC resistance' to the guitar signal than the amp input resistors you're making comparisons against? (obviously depending of the frequency...higher = more resistance) - in other words, you can't compare the input resistors on an amp with a transformer primary coil resistance (apples vs oranges)

[pdf64]

'Or to allow two disparate/incompatible circuits (DC level wise) to be connected (where caps aren't ideal)'

I think that more info is needed about what these circuits are, at least in terms of their input impedance, to give you a useful answer.
But if they're 1M or greater, then it's possible that they can both be connected passively, via dc blocking cpas, without a buffer of some sort being necessary.

[peskywinnets]

But if they're 1M or greater, then it's possible that they can both be connected passively, via dc blocking cpas, without a buffer of some sort being necessary.

I appreciate the offer of help & this really isn't meant to come across as offish - but as mentioned earlier, I really do want to avoid using a DC blocking caps at the front of the circuit I'm 'tapping off' the signal into - so rather than head off down that tangent, I'd like to find out if anyone has dabbled with small isolation transformers?

As it goes, 'The tapped off' signal will actually be feeding an active circuit...so the resulting level of the 'tapped off' signal isn't that critical (it can be corrected)...when I mentioned passively earlier ...I've probably clouded the matter (ie in the light there'll be an active circuit for the tapped off signal!) I meant using a passive device like a transformer as the initial mechanism to tap off the signal.

[cbarrow7625]

Have you tried just using a Y-cable to just split it passively with no additional components? If the input impedance of each amp and/or effect is 1M or greater, chances are you will hardly even notice the difference in output level. The guitar will 'see' a 500K input impedance instead of the 1M it would 'see' connected to a single device. Not really a huge difference when you consider the output impedance of a typical guitar pickup.

You would still have to deal with ground loops but that can be handled relatively easily with some cable modifications.

[peskywinnets]

Have you tried just using a Y-cable to just split it passively with no additional components?

Alas, I can't do that - the DC potential of a guitar pickup 'hot' output is 0VDC., but the input stage to the 'follow on' circuit is not 0V (a virtual earth arrangement & therefore sitting somewhere around +4V) - so I can't simply Y the two together.

the traditional way to circumvent this DC imbalance is to use a DC blocking capacitor, but using an input blocking cap has implications in the signal path, so i therefore need a way of connecting a small AC signal (sitting at 0V DC potential), into a circuit sitting at 4V DC potential *but* avoiding the use of a cap....which unless I'm missing the obvious likely needs a transformer to be brought into play. But therein lies the rub...what real life examples do forum members have (if any) with passive transformers to tap off a guitar signal? (esp wrt coloration of the original signal)

The fact that Jensen make & sell such transformers, suggests it must be do-able (at least to the extent where it doesn't impact the fidelity of the original signal that much - else nobody would buy 'em!)...and if they can do it, then with a bit of practise & trial/error, then it's likely I can make one too (cos I'm not paying out large spondoolies for what is in essence two coils of wire facing one another!)

[cbarrow7625]

Just curious, what kind of equipment are you using that is sitting at +4V DC? Is the hot sitting at +4V DC or the ground? Do you really think a transformer is going to color the sound less than a capacitor? For the cost of the transformer you could buy a mighty fine Audiophile grade capacitor that might sound "transparent". The transformer will likely just cause a different tonal shift. Maybe some more details about what, specifically, you are trying to accomplish would help create better suggestions.

[pdf64]

DC blocking caps in the signal path are ubiquitous, any system will require them somewhere along the chain, even the most expensive mic pre-amps. The important thing is to avoid types that will colour the signal, eg electrolytic and tantalum. As mentioned, I'd guess that almost any transformer will colour the signal more than a film cap.

[peskywinnets]

The cct is a self made affair - 4V DC just happens to be 1/2 VCC ...therefore the cct uses a vitual ground biased at 4V ....but also the hot is biased at 4V (ie just like in a dual rail opamp circuit where the ground is biased at 0V & the hot is biased at 0V too) - my latest dabbling is with no specific goal in mind other than to experiment with bespoke phasing throught a signal chain.

DC blocking caps in the signal path are ubiquitous, any system will require them somewhere along the chain, even the most expensive mic pre-amps. The important thing is to avoid types that will colour the signal, eg electrolytic and tantalum. As mentioned, I'd guess that almost any transformer will colour the signal more than a film cap.

Agreed, but if your DC path is biased to be the same throughought (from 'signal source' to the end of the chain - 'the speaker '......ie cradle to grave as it where), then there's no need for a blocking cap anywhere along the path. (ie if this were a dual rail affair...the signal path would be 0V...as would the input from the guitar - a dual rail push pull amplifier has a mid point bias point of 0V DC...etc )

Since I'm making an amplifier myself, I can control 'the grave' part of the equation...but it's the 'cradle' part I'm now looking at!

[cbarrow7625]

I assume you must be aware that there are thousands of pedals on the market that use the same biasing supply. All of them have a DC blocking cap on the input. It's just the price for admission using that type of circuit. A lot of folks think that many of those pedals sound great, even with their commercially viable low cost caps in place.

I don't know if it would be possible for your circuit, but often the single 9V battery virtual +/- design can be replaced by dual +/- 9V batteries with a real ground at the center. That could alleviate the +4V DC part of your issue, possibly without requiring a cap.

[peskywinnets]

I think perhaps my 'drift' still isn't quite getting across - I know that DC blocking caps are used ad infinitum everywhere - so the main thrust of me not wanting to use them has nothing to do with their colouration on the signal being fed through them ...but I want to try and eliminate a cap as being a potential source of phase shift (for which I accept most folks wouldn't not care one iota)....else if the phase of the signal coming into my circuit is moving (vs the original pickup signal) it'll be somewhat of a random target to work with!

Yes, i could go dual rail (& thereby eliminate the need for a front end DC blocking cap), but there are much choice available nowadays for single rail parts - plus the end circuit will possibly be going into a device that has a 8V single supply readily available - & I'd like to tap into that.).

[pdf64]

Or from one battery -
New Page 1
The + and - supplies will avoid the need for a cap or transformer on the circuits output also.