Adding mute switches to jensen 4-way splitter?

[bradberry00]

Hello all,

I am planning on building a signal splitter using the jensen schematic (http://www.jensen-transformers.com/as/as014.pdf). Where would be the best place to add a mute switch for each output? Before or after the transformers?

Thanks,
David

[R.G.]

Depends on how you plan to mute. If you're planning to short the winding to mute it, you will have to do it after Resistors 7/8/9, and you must use those resistors. Otherwise, shorting a secondary winding will reflect as a short into the combined primaries and mute them all. If resistors 7/8/9 are at least 10K, you can short the output jack without muting the other outputs as well.

If you're going to open the output to mute it, put a 100K or so resistor across the output jack and open before it hits the jack and 100K resistor.

I don't think I'd open before the transformers at all.

Those transformers were quite expensive last time I looked. The Edcor 10K:10K or the Triad TY141 are cheaper. The super cheapo Xicon $3.00 ones work OK if you do an opamp per transformer, like the isolated splitter at geofex.com.

[bradberry00]

Thanks RG. I plan on using the edcor equivalents for this build. So let me see if I understand this, put a 100k resistor between the tip and sleeve of each output jack and a switch between r7-9 and tip to open the jack. Is that it? Anything I have to do special for an led indicator or will it be just standard bypass switch wiring? I plan on using pots for r7-9.


Thanks again

[R.G.]

There are some trade-offs, of course. If you put in a minimum resistor of 10K ahead of the pots for R7-9, you can still use shorting switching. If  you want R7-9 to be able to go to zero, yes, put the switch in series between the resistor and the jack tip. The 100K is there to keep the jack tip from looking entirely open to the amp connected to the other end of the cable. This is a compromise, and you'll get some noise and hum from the 100K being the only impedance on the jack. It's probably not going to be too bad except in abusive situations. Shorting would be quieter when muted. But then if another output is running, you won't likely hear the minor hum and noise from a resistor-only muted jack.

Nothing special to do for switching, excepting that you want to use different switch sections for the audio and LED, and the LED switch is on when the audio switch section is open, unless you hook up the LED switch to short the LED when it's to be off. But that's probably needless complication.

[PRR]

Well, I'll be disagreeable. With R.G. but also with Jensen.



The 100 Ohm build-out should be *per output*, so that a drag on one path is not a drag on all paths.

There should be 10X this much, inside the box, in series with each output, so a dead-short on any output is tolerable load on the system. (The "10K" transformer has about 1K internally, but I'd like a simple dumb resistor too.)

With these changes, the switch is on the primary, breaks the source *and* shorts the output. The output signal is effectively "dead zero" in series with a few K.

If loads need some funny 47K source, that's their problem. Stick a resistor up their snout. In general the splitter should deliver good low-Z sources, for low line Z and good drive to "any" suitable load.

The above is a fairly generous design for "guitar cord" loads. For 10K studio loads I'd drop the resistors a smudge. True-600 old-school loads need a stronger plan.

[bradberry00]

anyone want to second the above design? And also, phase reversal, would it be possible? a switch that flips the secondaries is it as simple as that?

[R.G.]

I don't know that it's a disagreement as much as a different design approach. And a valid one.

The real way to design this is to know your transformers and loads well, and design to it. Not knowing what will be connected and using different transformers means there's likely to be some "season to taste" involved.

The 100 ohm in series with each trannie is reasonable, but it works in opposition to low frequency response extension - which may not be a problem with any given transformer. I liked at least one of them when I adapted this same circuit to the Xicon transformer, but for trying to protect and keep stable the opamp.

A better approach would be to note that the opamps are trivially expensive compared to the transformers and give each transformer its own opamp. Then you could have one output die entirely without affecting the others at all, unless the opamps went into chain-of-destruction mode.  It is also possible that the designer of this original circuit was using the 100R to keep the opamp stable in the face of the 100uF/100nF capacitive load if the output was shorted; perhaps the resistor was for the opamps all along. I hope this is one of those opamps that have internal diodes to V+ and V-, else they ought to be external.

The series resistors are another oddity. I would not have used them for a tone control as suggested. But once they're there, they make a convenient way to protect things from an output short, especially light of the advice to use them to tailor treble loss.

But these are just my own personal preferences. There are many ways to do it with good results.

[tubegeek]

a switch that flips the secondaries is it as simple as that?

Yup. Primaries or secondaries, either way. The only subtlety (that I know of) is the capacitance to the core ground (if any) and shield (if any) may give you VERY slightly different frequency response with one flip method instead of the other.

[bradberry00]

quick frankenstein schematic, will this work?



thanks for all your help so far!

[tubegeek]

I'm on the team that thinks that a $0.85 op-amp is a wise investment to improves an $85 transformer's behavior.
If those 25K pots are there to trim the output level, I'd double-suggest that you add an op-amp per output so as to provide a gain control there instead of at the output.

The way you have it, the 25K will be forming a voltage divider with the next device's input, and its effect on the level will be less predictable, while at the same time, when you turn the level down (by putting more of the 25K in line with the output) you'll be raising the output impedance, making you somewhat MORE vulnerable to induced hum, and treble loss down the cable.

How about:

Delete the mute switches. Move the 25K as follows: break the connection to the RED wires. Connect the clockwise end of each 25K to the right-hand junction of C1 and C2 where all the reds go now. Tie the wiper of each pot to each RED wire on the transformer. Tie the counter-clockwise end of each pot to the BRN wire of each transformer.

You will then be able to mute each output by either a switch shorting RED and BRN, or by turning the pot to the left with no need for a switch. If you want to be able to BOTH trim and MUTE, choose the first option in the previous sentence, then you can adjust the level and also mute the result. The 25K pots won't load down the op amp too badly. They might interact just a little, I think?

Your phase-flip is OK.

[bradberry00]

The pots are not suppose to trim output level. they are there as per the original schematic to act as treble roll offs.

[bradberry00]

Could someone please tell me why just having a switch disconnecting both primaries of each transformer wouldn't work to mute that output. It would be like it isn't even connected to begin with when switched off. I'm thinking you could just use a 3pd, one pole for each primary and one for a led indicator.

maybe I am over simplifying things.

thanks!

[PRR]

Disconnected is like you plug into the amp but leave the guitar end laying around. It tends to pick-up crap. At best there is a faint hum/buzz out of the "muted" amp. Loud buzzes and outright howls are possible.

This actually works "OK" for studio line-level signals inside a box. I just suggested that on a studio-gear forum.

At guitar level (lower voltage, higher impedances) it is dubious. Sure. Try it! Worst can happen is you say "yuk!" (or "ouch!") and have to re-wire to short un-used inputs.

[bradberry00]

Ok totally understand, thanks. So is it the consensus that the best method for this is to short the output jacks? I just modified the schematic, let me know if this will work.

thanks much!



thanks geofex and jensen transformers for original schematics!

[bradberry00]

Hey, I just wanted to bring this back to the top quick. I finally have some time to build this now and would like to know if there are any errors in this schematic.

thanks much!
-david