Is possible to plan the construction of an OT for PP & SE operation ?

[kagliostro]

I mean if is possible to plan an OT as to be used in PP and SE configuration

I think to a core with laminations arranged for SE use (E on one side and I on oter side plus gap between it)

with a PP primary with the CT external as to be able to disconnect it and connect the two half windings

in parallel (paying attention to the phase) so it will be possible to have an 8k / 8ohm PP OT

and changing the connection of the two half windings a 2K / 8ohm SE OT

Do you know if this thing has never been proven ?

Some opinion about it ?

K

[shooter]

I did some digging awhile back, it was not very fruitful.  Can't find the schematic but there was an amp called an "oddblock"?  I think it was a PP using a SE tranny, might be worth the google.

[kagliostro]

Also I'm not able to find such schematic

may be it was an SRPP amp ?

K

[shooter]

YUP, was srpp, I found the PS schematic and one tap was labeled "to srpp amp"

[kagliostro]

OK, an SRPP use an SE transformer


K

[HotBluePlates]

I mean if is possible to plan an OT as to be used in PP and SE configuration

Yes, it is possible. It is also somewhat wasteful in the push-pull configuration.

That is because the air-gap needed to allow for the standing direct current of single-ended operation also means you need a larger core for the same power output at low frequency, compared to push-pull operation.

... with a PP primary with the CT external as to be able to disconnect it and connect the two half windings in parallel ... to have an 8k / 8ohm PP OT and ... 2K / 8ohm SE OT ...

There's no need for special tricks; assuming the OT can handle the unbalanced d.c. of single-ended operation, all you need to do is disconnect one end of the push-pull winding from a tube plate. See below.

Imagine an 8kΩ plate-to-plate push-pull OT, with a center-tap to B+ and an 8Ω secondary winding. Each side of the push-pull output stage sees its half-winding from the tube plate to B+ at the center-tap, or 4kΩ. If you disconnect one side of the push-pull by removing the output tube from its socket, the remaining side sees a load of 2kΩ from its plate to B+ at the center-tap.

The reason is turns ratio and impedance ratio; the impedance ratio is the square of the turns ratio. For 8kΩ to 8Ω, the impedance ratio is 1000:1 and the turns ratio is 31.62:1. As long as some current is flowing through both halves of the OT primary, all turns of that winding are in use and the total impedance is 8kΩ plate-to-plate, or 4kΩ from plate-to-B+ (meaning each output tube sees its half-winding from its plate to the B+).

If you stop all current through half the primary winding (perhaps by removing a tube from its socket), the remaining tube still sees its half-winding from its plate to B+ at the center-tap. This half-winding has half the turns of the entire push-pull primary, which gives a turns ratio of 31.62/2 = 15.81:1. Squaring this, the impedance ratio becomes ~250:1, so when you attach an 8Ω speaker load the transformer reflects 8Ω * 250 = 2kΩ to the remaining half-primary.

Incidentally, this is also why a class AB amp has a load-per-side of 1/2 the plate-to-plate impedance for small signals but transitions to a load-per-side of 1/4 the plate-to-plate impedance for large signals (when the opposite side is driven to cutoff).

I mean if is possible to plan an OT as to be used in PP and SE configuration ... 8k / 8ohm PP OT ... 2K / 8ohm SE OT ...

Earlier, I said the resulting transformer would need a bigger, heavier core due to SE operation which would be wasteful in PP mode. Let's say you envisioned 2x 6V6's in push-pull with an 8kΩ primary with an output of ~20w. If you pull one of the 6V6's, the remaining tube sees a 2kΩ load which will output less than 4w due to the non-optimal load. If you already had a typical 20-25w push-pull OT on hand (with no special construction), you might try it and see if use this way results in too little bass or power through-put. That is, the SE setup would be for reduced power anyway, so a further power reduction due to core saturation may not be a problem.

If you have a different scenario in mind, or need more than a small fraction of the push-pull power output when in the single-ended mode, you might need to talk with a transformer designer to specify the transformer properly so you don't have to over-build and over-pay.

[shooter]

Something else to consider since you will have 2 tubes in PP, when switching to SE, switch to PSE.  You lower the valves "needed impedance", you gain more overall SE power.  Most of my SE builds have been PSE and the musicians do notice the "extra power", or, wow, it sounds fuller that way! FWIW

[kagliostro]

Thanks HotBluePlates & Shooter

@ HotBluePlates

The idea was to start from the spec of an SE OT (that is bigger than a correspondent in power PP OT) and to have inside a PP winding arrangement

I was thinking that to use the two partial windings of the PP primary in parallel is a necessity to permit to all the power to be transfered to the
secondary

This to have in the same amp a 18W PP arrangement (2 x 6v6 for example) and a 18W SE arrangement (1 x GU50 for example), not the disconnection of a tube from a PP to work in SE with the remaining one

---

@ Shooter

You give me the idea to arrange a PP power section with the ability to swap in PSE using the two tubes of the PP paralleled (arranging the PI for that purpose), but I must think about the OT spec requirements

May be it will be necessary to have also a 4ohm secondary on the hypotetic transformer with 8k / 8ohm

the use of the two 6v6  in PSE (if my math is correct) requires near 4k load, so connecting an 8ohm speaker to the 4ohm intake will reflect 4k if the two semi primary windings are paralleled (must have the CT connection external to permit this arrangement in phase)

I'm not sure, but I think that if the two partial primaries remains connected as for PP but used in PSE, the signal in the counterphase windings will cancel, or not ????

Franco

[tonepumpamps]

Here is a link to something I found awhile ago regarding this.
http://www.ampbooks.com/home/amp-technology/amp-patent-7173488/


Looks quite interesting, of course Mesa Boogie already has a patent on it 

[shooter]

I believe you'd have to "switch" to common drive signal and common plate.  I also thought of an UL 40% tap as a "CT" for PP, but that's as far as I went.

[kagliostro]

Thanks for the link Tonepumpamps

Franco

[jjasilli]

Unless I'm missing something it seems that the Mesa Boogie patent is little different from the KOC "single-ended mod", which uses a pot rather than a switch to progressively short signal to ground for one PP power tube.  However some Forum members recently have noted that it doesn't sound very good.  Any takers here?


Another point:  in the recent "matrix stereo thread": http://el34world.com/Forum/index.php?topic=18195.msg185401#msg185401


some research (Peter Bench?) discloses that with a few watt amp, you can cheat & get away with using  a 15W PP OT in SE mode.  Maybe that will help.

[kagliostro]

Ciao JJasilli

Thanks for the link

However the idea is to build a "special" OT as to perform PP and SE not a circuit to swap between PP and SE

Franco

[HotBluePlates]

The idea was to start from the spec of an SE OT (that is bigger than a correspondent in power PP OT) and to have inside a PP winding arrangement

I was thinking that to use the two partial windings of the PP primary in parallel is a necessity to permit to all the power to be transfered to the
secondary

This to have in the same amp a 18W PP arrangement (2 x 6v6 for example) and a 18W SE arrangement (1 x GU50 for example), not the disconnection of a tube from a PP to work in SE with the remaining one

Will you only use one mode or the other? If so, 2kΩ:8Ω is exactly half the 8kΩ:8Ω winding. In your case, you'll have the transformer winder use a core which allows 18w single-ended, which will be much bigger than an OT which only does 18w in push-pull. If you only use one mode or the other, you could have the GU-50 connect only to one end of the primary and use the center-tap. If you're planning external switching between modes, this is the simplest option.

If you need complete independence (different supply voltages, for instance) and will use both modes , then the GU-50 will need a winding separate from the push-pull winding. You'll need a 36w SE core in this case (or whatever the transformer winder suggests; there may be an in-between option). Driving both windings, phasing will matter, though your winder will be able to tell you correct phasing of the 2 primaries.

Looks quite interesting, of course Mesa Boogie already has a patent on it 

Mesa's only patenting the switching, and it only applies to you if you market an amp with that switch. Franco does a lot of home-brew projects.

The noteworthy thing of either Mesa's plan or KOC's plan is that you use an ordinary push-pull OT. The idle current of the non-driven side offsets the idle current of the driven side to result in no need for an air gap. The downside is most often maximum SE output power occurs with a load numerically equal to the push-pull plate-to-plate load delivering maximum output power; these plans will result in a load half as big. But that's usually not an issue because the player wants SE mode mostly to reduce output power, and secondarily to avoid even harmonic cancellation in the OT.

[kagliostro]

Thanks HotBluePlates

these plans will result in a load half as big.

in this eventuality a 4ohm intake to which swap the connection of the 8ohm speaker will solve the problem ?

Franco

[HotBluePlates]

Thanks HotBluePlates

these plans will result in a load half as big.

in this eventuality a 4ohm intake to which swap the connection of the 8ohm speaker will solve the problem ?

There's no problem to solve. You simply get less than the maximum possible output power for the single tube with that supply voltage. But that's the point, especially in KOC's case where you have a pot to continually vary between SE and PP.

[kagliostro]

OK, Thanks

Franco

[jjasilli]

FWIW: with KOC's SE mod, both PP power tubes are ON (though one has less or no signal).  Hence, plate-to-plate impedance remains the same, for matching to the OT primary.

[kagliostro]

That is a point of advantage to those circuits

Franco

[kagliostro]

Thanks for the info Drgonzonm

Interesting, the Lundahl company is a very important one

Franco