push pull ot in a single ended amp

[JoshBernstein]

Ive got a 40 watt push pull ot, and a 5f1 with a blown ot. Is there any way to rig this thing in there so that it will work, or should i just toss this tranny and buy the correct one?
Thanks!

[martin manning]

SE OT's have a small gap in the magnetic circuit so they can carry a high idle current without saturating the core. A 40W p-p OT would be oversized so much that it would probably carry the Champ's 40 mA idle current without overheating, but in p-p the magnetic flux from each side is cancelled by the other side.

The other consideration is the primary impedance.Is this the one from your AC30 project? If it is an 8k p-p OT it would probably work ok using the outer primary leads.

OTOH, you can get a new Champ OT for about $25...

[JoshBernstein]

Alrighty then, probably won't use it. Yes it is the one from my ac30 project, just trying to figure out something to do with it.

[David Root]

If the primary impedance is OK, you can do it, it's called ghetto UL SE. I did one a few years back with a KT66 in a HoSo56 kind of circuit. See schematic below. Powerwise your 40W OT should work fine in a Champ.

If the Zpri is not right, use a Hammond 125E like I did and pick your Zpri!

[TUBEDUDE]

Has anyone tried to hacksaw the stack on a PP OT to make a magnetic gap to be more SE compatible? I'm not sure it can be done without shorting the laminations, but have wanted to try it.

[Firestorm]

I've spent some time with SE amps. The Zpri is just a matter of power transfer. You get a theoretical max of 0.5X the plate's wattage (hence 15W with a 6L6GC, 12.5W with an EL34 and 7W with a 6V6). Provided you drive them hard enough. 6L6s sound odd in SE. 6V6s are the standard. EL34s sound wonderful.

The OT, though, has to carry what transformer people call the bias current. Or the DC will saturate them. Thus you lose the low end. Not all bad since SE will not cancel 60Hz hum. A Fender Champ has on OT about the size of a shot glass. No response below 100Hz.

There are SE transformers for Hi-Fi. Huge beasts. They will pass low frequencies if you can kill the hum.

[jazbo8]

If the Zpri is not right, use a Hammond 125E like I did and pick your Zpri!

Even though the 125E is an Universal PP OPT, it is rated to handle 80mA of DC current. Do not try to run SE with other PP OPT's unless they have similarly DC current spec.

[Firestorm]

We have to separate myth from fact. Appropriate in the Christmas season since Luke is almost all myth. (Although, George Harrison observed, if you believe it, then it's true.) We can talk about that somewhere else.

Tubes will try to do what they want to do. If the laws of physics weren't in the way, they would pull infinite current from the cathode. That doesn't happen. They will try to develop maximum power across the load. If the load is wrong, they do the best they can. If the OT is poorly designed, the windings short. If well designed, you lose bandwidth.

Marshall started out with terrible OTs. Now we're all freaked out.

Use fat enough wire and no harm will follow.

Bandwidth, though, matters.

[Firestorm]

Oh, if you actually care about about Luke, ask me in Garage Talk. I have a degree in that crazy stuff.

Otherwise, Happy Christmas, Happy Saturnalia, Happy Solstice, etc.

[Phil_S]

I'm thinking the hacksaw question is tongue and cheek kidding. Just in case it isn't, I want to assure you it won't work. Go find a diagram of how the lams are stacked in SE vs. PP. In SE, the E's all go on top and the I's all go on the bottom, allowing a piece of cardboard or other non conductive shim to go in between to create a gap of the proper size. PP is stacked with E's and I's alternating in orientation.

Spend the $25 for the appropriate OT. Good choices abound for a Champ-level OT. Some things aren't worth the effort. However, if you are so inclined, please feel free to disassemble your OT. Then restack the lams for SE, calc the proper size gap, find the proper size shim, and complete the assembly using the proper varnishing and baking required. I'm sure you'll get a very nice result (no, you won't.) While you are at it, you might as well clip off the center tap.

Seriously, though, there is all sorts of decent reading available on the internet on this topic and it isn't hard to find. Once you see what it takes to design and wind an OT (more complex than a PT) you'll have a better appreciation of your guitar amps and your hi-fi stuff, too. Also, your mind is likely to change about why OT's cost so much. Really, some of them are a bargain considering what's required to make a good one.

[R.G.]

OTs get complicated fast because they are critically dependent on the magnetic field inside the iron and air around the iron and because everything involved in their operation is a variable.

There is a short answer to your question. You might be able to rig the transformer to work, but to do that you'll need to fake the operating conditions it was designed for by using half the primary for the output tube as you're thinking, but using the other half of the primary for a "bias current" to keep the iron from saturating magnetically. This preserves the transformer as it, but gets more electrically complicated.

Magnetic materials have a limited range of field they can support. Push them over this limited field strength and they quit acting magnetic. This is what saturation really means: the iron is fully saturated with magnetic field and can't hold any more. These materials can have fields in two different directions, positive/negative or north/south, whatever you want to label it. And there is a saturation limit in both directions.

A given lump of iron can only "hold" a certain field intensity in each unit of mass/volume, so the bigger the lump of iron, the more total magnetic field it can hold before saturating in either direction. The amount of power that can be pushed through a given lump of iron depends on how much of the total magnetic field you "work". If you can work it from full saturation positive to full saturation negative, that's all it can do at a given frequency.

Stay with me, we're getting there.

A Single Ended (SE) transformer only uses one direction of magnetic field. By its nature, it only uses the iron from zero magnetic field to saturation in one direction. Worse, to avoid distortion, the **iron** has to be "biased" halfway to saturation in the direction it's being used, so the available magnetic field is cut in half again as far as signal is concerned, so SE transformers are only capable of 1/4 of the power throughput of a push-pull (PP) transformer for equal weights of iron and copper. That's the best it can do. Other considerations can make it worse.

Beyond that, if you simply put a voltage across an inductor, the current ramps up to infinity. Well, OK, it ramps up until the coil melts or the voltage source limits. The electronics feeding an SE OT has to limit the current it lets through the OT to the amount that just balances the iron at half of saturation. A PP OT is set up so that it's balanced at zero, which is an easier value to hit.

If you take a PP OT and hold one side to the desired "bias" point to the iron, and then drive the other side with signal, the bias-point half pulls the iron back to a net zero, and you can drive the other side with signal and not saturate the iron. This is fairly obvious once you've been extruded through a magnetics and transformers course, but for some reason the US patent office awarded a patent to some fellows a decade or so back for writing it up. Their circuit simply removed drive from one side of the OT and called it "single ended" in that mode. That's OK as far as it goes, but it is not the same as what we normally think of as single ended.

But it is the answer to your question. Rig the OT so it is held at some electrical bias point by pulling on the non-signal side of the OT to keep the signal side from saturating. It still doesn't get you the impedance matching you want, which is another discussion entirely.

The answers you've been given are good. If you want to tinker, you can tinker with that one. If you want the shortest path to a working amp, go get a different OT that is designed for the SE circuit.

[TUBEDUDE]

Thanks Phil S,
I remember reading about the cut laminations in an old Glass Audio mag. I didn't think about the lamination construction being different in the stack depending on class of operation. Thanks for the education.

[Firestorm]

I build a parallel SE amp (only two tubes, and even that was a nightmare). I wanted a Champ times 2. And one that could do EL34s or 6L6s. There are huge SE OTs with sufficient gap to handle the current required. But they're designed for Hi-Fi tubes like 300Bs.

I used two of the Hammond's, wired parallel. Spent a long time twisting them around to eliminate induced hum.Turns out if you flip them and offset them, they're guiet.

[TUBEDUDE]

Years ago I built a SE mp with a 6DQ5, (my first build with plate cap), and an OT from One Electron. It was good for 14 watts in enhanced mode (driving the screen grid). I later changed it to run a KT88 in regular SE for about the same power. I still had some 120 Hz. hum with a lot of filtering.
It was a.testbed for several things, active tone control and a Mu amp front end.

[strelok]

I had that same issue with my kt120 based DLM. The only way I could eliminate the hum was to use an inductive Pi filter at the input of the powersupply. Before I had 140uf of filtering on the first node and it still hummed! The final configuration I ended up with was 50uf-3H choke-50uf then fed into the OT/screens supply. This brought any hum down to well below audiable levels.