Transformer isolated tuner feed

[Buffalo Tom]

Im looking for a good way to (passive) split a contact microphone/piezo-pickup signal to a always on tuner output. Can I use a transformer to prevent that the piezo pickup being negatively effected by the split and the tuner? The transformer is 42TM018 from R.G. Keens Hum-Free(er) A/B/Y Splitter Box. I know that the quality of the 42TM018 transformer is not the best and in the original R.G. schematic there is a op amp driving the transformer but my design is passive. I have seen many passive 2 way splitters with transformers so it can be done (Lehle p-split etc). Will the frequency response on 42TM018 be good enough for tuner feed using instruments like violin, viola, cello, double bass etc? 42TM018 spec: http://www.mouser.com/ds/2/449/XC-600134-1212477.pdf

1. Is this a good way to do it?
2. Is the 10K-0.001uF network doing any good, or should I remove it?

Thanks!

[antonis]

The RC snubber 10k/1nF(or 100R/100nF) is mainly set on power supplies transformer primary winding (or across mains switch, which is equivalent..) to reduce (smooth) inductive kick during shut-down..

In your case, it isn't necessary but it would do no harm..
(depending on tuner input impedance, of course..)

[Phoenix]

The transformer doesn't present some "infinite impedance" or "no load" to the guitar - it reflects the tuners input impedance according to the turns ratio.
So if the primary side has a high turns count, and the secondary has a low one, sure, it can reduce the load presented from the tuner, but it will also step down the voltage from your pickups, and your tuner may not register them correctly if the resulting amplitude is too low. The same can be reversed, low turns primary, high turns secondary, more load presented to guitar, higher voltage to tuner.

With the transformer you've chosen, 10k:10k, the impedance is matched and the pickups are loaded exactly the same as before (actually a little worse thanks to the transformer parasitics).

What's the input impedance of your tuner? If it's low enough that it presents a significant load on your pickups, then a more practical solution would be to either modify the circuit for higher input impedance, or build a small active buffer with a massive input impedance to go between the guitar and tuner. 10Meg is easy and will not cause any noticeable effect.

[GibsonGM]

Im looking for a good way to (passive) split a guitar signal to a always on tuner output.

Or...http://www.muzique.com/lab/splitter.htm  which doesn't require much real estate at all, and you can set up any number of outs that you like...

[Buffalo Tom]

Im looking for a good way to (passive) split a guitar signal to a always on tuner output.

Or...http://www.muzique.com/lab/splitter.htm  which doesn't require much real estate at all, and you can set up any number of outs that you like...

Yes I know.. But I'm looking for a passive solution.

[Buffalo Tom]

The transformer doesn't present some "infinite impedance" or "no load" to the guitar - it reflects the tuners input impedance according to the turns ratio.
So if the primary side has a high turns count, and the secondary has a low one, sure, it can reduce the load presented from the tuner, but it will also step down the voltage from your pickups, and your tuner may not register them correctly if the resulting amplitude is too low. The same can be reversed, low turns primary, high turns secondary, more load presented to guitar, higher voltage to tuner.

With the transformer you've chosen, 10k:10k, the impedance is matched and the pickups are loaded exactly the same as before (actually a little worse thanks to the transformer parasitics).

What's the input impedance of your tuner? If it's low enough that it presents a significant load on your pickups, then a more practical solution would be to either modify the circuit for higher input impedance, or build a small active buffer with a massive input impedance to go between the guitar and tuner. 10Meg is easy and will not cause any noticeable effect.

Thanks for your reply. (my first post is changed a bit..)

[antonis]

But I'm looking for a passive solution.

"Passive" means attenuation(*).. - Greg explained you in more details.. 
(*) not in case of tunnel diodes but they're obsolete nowdays..

If you don't want to "loose", you have to work on "active" side..

[Buffalo Tom]

What's the input impedance of your tuner?

Different standard pedal tuners will be used, TC, Boss etc.

[Phoenix]

Ok, now that you've edited your original post, it makes the thread a bit more confusing, might have been better if you added the details in a follow up post.

Anyway, if the pickup is a PIEZO, then it should have a preamp/buffer anyway (preferably in the instrument), and the preamp should be able to drive a couple or dozen pedals/tuners in parallel without any major hickups. If it doesn't have a preamp/buffer, well, that's really the ONLY practical solution.

EDIT: Just re-read your edited post, if it's a microphone/piezo system, it's already got a preamp. It'll have no problem running a parallel always-on tuner along with pedals or an amp input, all you need is a Y-cable or splitter box/ABY.

[GibsonGM]

But I'm looking for a passive solution.

"Passive" means attenuation(*).. - Greg explained you in more details.. 
(*) not in case of tunnel diodes but they're obsolete nowdays..

If you don't want to "loose", you have to work on "active" side..

FETs and MOSFETS...but not necessarily to boost - maybe only to buffer       The impedance offered by either of them allows you to drive many outputs...so you could chain out to a tuner, talkbox, slave amp, etc etc.   They can drive any tuners without consideration of their input Z.   The current drain of these guys is minimal; a single 9V battery would last a long, long time.

But if you want a passive solution, that is certainly up to you, Tom!  What works for me may not work for you.

[R.G.]

You've uncovered another of those problems that seem like they ought to be simple, but aren't.

Piezo pickups are inherently high impedance capacitors. Unlike the wound inductive pickups on normal guitars, they have big bass loss when you make them feed too low a load impedance. It's even worse when you make them drive inductors. Inductors (like the primary of a transformer) are even worse than low resistance loads for a piezo pickup, because their inductance makes them REALLY low-ohms loads at bass frequencies.

A 10K:10K transformer is a 1:1 RATIO transformer. It is called a 10K:10K only because when it's loaded with a 10K resistive load on the secondary, it produces its specified bandwidth.

In the case of the 42xxxx transformers, they are mostly specified at 300Hz to 3000Hz. Below the nominal bass cutoff of 300Hz, the primary inductance "eats" signal away from the bass. The point of the active drive in the geofex article was to be able to use active devices to force the transformer to lower than its natural bass rolloff by shoving in more signal current as needed to get below the natural cutoff.

A piezo pickup, all by itself driving a transformer is going the opposite direction. A piezo pickup gets higher loss on its own because its ability to drive current decreases at lower frequencies. A transformer needs more drive at low frequencies, not less. It's kind of the worst of both worlds.

It is possible to design a transformer that will work well with a piezo guitar pickup at bass frequencies. But to do this, you have to make it both big and with dramatically more windings, and that then starts to make the transformer's high frequency losses go up, so you lose treble because you made it higher impedance for bass frequencies.

All of this together makes a purely passive transformer drive by a piezo be a very difficult thing to do, and almost certainly not something you can do with a US$3 tiny transformer off the shelf.

If your piezo pickup has a FET follower - and many of them do, for the reasons listed about piezos above - then you can do the transformer setup as you've shown if you get the right transformer, and that starts to be possible. But a passive piezo driving a transformer is very, very difficult, and not something to just throw together with cheap parts.

[PRR]

> Different standard pedal tuners will be used, TC, Boss etc.

Last tuner I used had a motor and neon.

Looking at pathetic specs for today's tuners, they may have a quite-high input impedance (low loading) AND only when stomped (otherwise "true bypass", whatever truthiness is). So plug and play, no problem.

Are you having a problem?

No, there is no good "passive" way to interface a load to a piezo without bass loss. The tuner obviously has a passive buffer between the jack and the guts. (For those prices, it really should!!)

[Buffalo Tom]

You've uncovered another of those problems that seem like they ought to be simple, but aren't.

Thanks everyone for your replies. So if I get this right you are saying that splitting the signal with a transformer the way I have done it on my drawing is actually worse than using just a Y connection??. As I wrote before my goal is that the signal from jack in to jack out should be untouched and the audio on tuner out should just be "good enough" for at tuner. Thats why I picked a cheeper transformer instead of Jensen JT-11P-1 that seems to work for a passive split according to Jensen

[R.G.]

Thanks everyone for your replies. So if I get this right you are saying that splitting the signal with a transformer the way I have done it on my drawing is actually worse than using just a Y connection??.

Yes. The transformer's primary inductance loads the signal all by itself, irrespective of the tuner, and this introduces issues on its own. The Y connection is just the tuner loading, for good or bad.

As I wrote before my goal is that the signal from jack in to jack out should be untouched and the audio on tuner out should just be "good enough" for at tuner. Thats why I picked a cheeper transformer instead of Jensen JT-11P-1 that seems to work for a passive split according to Jensen

Good reasoning, but missing the fact of the transformer primary loading.  I have no doubt that the Jensen diagram and transformer works well, pending actual investigation, which I have not done. But it's the kind of thing that does need investigation before buying a quite expensive transformer. Cheaper ones have the primary loading issue worse than an expensive, high inductance, wide-bandwidth transformer.

If i may make a suggestion - why not put a momentary footswitch in there to connect the tuner directly only while you're tuning? That gets you the best of both approaches and costs a lot less than a transformer setup that's of high enough quality, whatever that is. Othewise, I'd use a buffer out to the tuner. That's not purely passive, but I'd bet a significant amount of money that I can design a buffer that you cannot reliably determine to be in circuit or not. It would need power, but hey, electricity is our friend, right?   

[Buffalo Tom]

If i may make a suggestion - why not put a momentary footswitch in there to connect the tuner directly only while you're tuning? That gets you the best of both approaches and costs a lot less than a transformer setup that's of high enough quality, whatever that is.

Thanks for the suggestion but the tuner out must be on all the time. Im building the boxes for violin, viola, cello, double bass players at a theater I'm working for. They like to be able to see the tuner while they play.

Othewise, I'd use a buffer out to the tuner. That's not purely passive, but I'd bet a significant amount of money that I can design a buffer that you cannot reliably determine to be in circuit or not. It would need power, but hey, electricity is our friend, right?   

Yes Im going to use a buffer on the tuner out instead of a transformer. I see that now thanks to your help. I found a schematic made by AMZ Jack I'm going to try. I hope my translation from the original schematic to "point to point" is right.


One of the reasons why I first wanted the tuner out part to be passive is that the box I'm building also have have a microphone mute function. It mutes both XLR (mic) and jacks (piezo) with the same switch. This function is done with a dpdt relay and the power is coming from the 48 volt phantom supply. And Im not sure how much current I can take from the phantom supply before there is problems. The relay mute+leds is around 2-3mA.. How much current do you think the piezo buffer in this schematic will need?

[highwater]

Your layout looks correct to me.

As drawn, at 9v, maximum current is about 1ma, about 0.5ma with no signal. That would be fine, but phantom power is 48v, not 9v. If you run that buffer at 48v, it'll draw up to ~5ma, which is probably too much (even if it isn't, the corresponding 1/4watt almost certainly will be).

The good news is that you can tweak the buffer for less current consumption. The best way to do that is above my pay-grade, but since it only has to drive a tuner (meaning some treble loss is no big deal) you might be able to just increase R4 to 100k and R5 to 1Meg. The 22Meg bias resistors are fine at 48v though - they'll draw just-over 1micro-amp.

[PRR]

"Phantom" is 48V through 3.4K.

If you are sucking 3mA, you are already down to 38V.

And if you just steal power any simple way, you will short-out mike signal.

[Buffalo Tom]

Thanks for the updated schematic PRR!

[Buffalo Tom]

"Phantom" is 48V through 3.4K.
If you are sucking 3mA, you are already down to 38V.
And if you just steal power any simple way, you will short-out mike signal.

Yes I have done some test now and obviously my original plan is not going to work.. The phantom power needs to stay at 48V when the MIC is active. So what I need to do is to use a 3PDT switch instead. One pole for the MIC mute circuit and two poles for the piezo pickup A/B thing that send piezo to output OR tuner. (All this is passive and require no power from the phantom supply). BUT what I like to do is to take some power from the phantom supply while the MIC is muted and use it for a LED. So what I'm trying to figure out is:

1. How do I take power from the phantom supply without running into problems? (POP!) You changed the schematic with two 100K resistors tied together from pin2-3 and a 10uF cap to ground? Is this the way to do it?

2. I like to use a 3PDT switch if possible but then I have no pole left for the LED. Want to stay away from 4PDP if possible.. Can I use millennium bypass? Cant figure out where to connect the "control" wire. Does millennium bypass draw current when the LED is off?

I realize that the thread now is far away from the original question... Sorry.

[Buffalo Tom]

...