Does output transformer make a difference?

[Austin]

The original OT was not faulty, just too heavy and I am not going to design an amp with the Fender transformer. It is a huge Bass transformer if you are familiar with the Bassman 100. Also I have a suspicion that the original transformer was designed for bass ( less core efficient), it is way over kill for guitar and the inductance must be higher. That's the two reason I replace with the Magnetic Components smaller 40W OT. I was hoping to get a little compression at high output, but I did not notice that either.

To my understanding, a larger transformer will be MORE efficient than a smaller one not less. The added inductance will enable you to have have better low frequency reponse as well.

Transformer Energy Losses - Wikipedia, the free encyclopedia

I would try to get the compression you are after by adjusting the load to the power tubes.

[Alan0354]

To my understanding, a larger transformer will be MORE efficient than a smaller one not less. The added inductance will enable you to have have better low frequency reponse as well.

Transformer Energy Losses - Wikipedia, the free encyclopedia

I would try to get the compression you are after by adjusting the load to the power tubes.

Haha, that is in my other post trying to mismatch the load to adjust the power and want to see what is the pit fall and how far can I push it.

[Austin]

Haha, that is in my other post trying to mismatch the load to adjust the power and want to see what is the pit fall and how far can I push it.

The Edge... there is no honest way to explain it because the only people who really know where it is are the ones who have gone over.
Hunter S. Thompson

[J M Fahey]

For one, I find human ears ( at least mine) is not reliable. Memory particularly is quite inaccurate. Unless you have identical amps only different in the OT and you can switch side by side, it is almost impossible to compare. Listening to the recording is not very accurate either. Even if you have identical amp setup, each amp might still be a little different. Also, it is impossible to pick the strings exactly the same everytime. The strings sound different every note you pick.

Fully agree.
Sound memory lasts seconds, literally.
To compare those 4 transformers I listened to *only* the first 10seconds of each track, then the next, back and forth 5 or 6 times.
Only way to confirm that something "I thought" I had heard on one but nor on the other was actually so ...or not.

[Steve Conner]

Compression in push-pull OTs: I think what happens is that when the output stage is overdriven, it can take on a DC imbalance, if the two sides clip with different duty cycles. This net DC current can saturate the transformer no matter what the frequency of the signal.

The OT is just one small piece in a tonal jigsaw puzzle. I think all of the makes discussed in this thread would work just fine. The only thing I've seen from Mercury that might justify a higher price is a weird stacking pattern of laminations in their JTM45 clone. I believe it's meant to make the saturation more gradual.

[Steve Conner]

Compression in push-pull OTs: I think what happens is that when the output stage is overdriven, it can take on a DC imbalance, if the two sides clip with different duty cycles. This net DC current can saturate the transformer no matter what the frequency of the signal.

The OT is just one small piece in a tonal jigsaw puzzle. I think all of the makes discussed in this thread would work just fine. The only thing I've seen from Mercury that might justify a higher price is a weird stacking pattern of laminations in their JTM45 clone. I believe it's meant to make the saturation more gradual.

[R.G.]

Compression in push-pull OTs: I think what happens is that when the output stage is overdriven, it can take on a DC imbalance, if the two sides clip with different duty cycles. This net DC current can saturate the transformer no matter what the frequency of the signal.

Good point. It's not the transformer so much as what drives it.

And the genesis of a new theory about how to get output compression at any/lower levels. Hook up a third circuit element so it can pull DC through a winding - any winding - on the OT, and this introduces a net DC offset into the core's operation point, which both offsets it to one side of the BH curve, and reduces the incremental inductance.

Similar setup to getting SE operation in a smaller core - use pushpull, but use one of the output devices as just a DC offset, don't feed signal to it. Lets you keep DC in the SE signal output device, but offsets the BH curve back down near center so you can use more of the total BH curve and hence get much smaller iron.

Hmmm. Magamps.

Only magamps use the differential inductance (mostly) for control, and (mostly) arrange to cancel the controlled signal excursions in the windings.

Lessee. Offset BH curve one direction. One polarity signal adds to this and pushes over the curve to saturation. The other signal alternation opposes the offset and pushes it away from saturation. Sounds asymmetrical. Dynamic, of course, so musicians will like it like trolls like shiny things. But I think the effect still is dependent on volt-time integral adding/subtracting from the offset, so it'll still be frequency dependent, won't it?

[Austin]

Good point. It's not the transformer so much as what drives it.

And the genesis of a new theory about how to get output compression at any/lower levels. Hook up a third circuit element so it can pull DC through a winding - any winding - on the OT, and this introduces a net DC offset into the core's operation point, which both offsets it to one side of the BH curve, and reduces the incremental inductance.

Similar setup to getting SE operation in a smaller core - use pushpull, but use one of the output devices as just a DC offset, don't feed signal to it. Lets you keep DC in the SE signal output device, but offsets the BH curve back down near center so you can use more of the total BH curve and hence get much smaller iron.

Hmmm. Magamps.

Only magamps use the differential inductance (mostly) for control, and (mostly) arrange to cancel the controlled signal excursions in the windings.

Lessee. Offset BH curve one direction. One polarity signal adds to this and pushes over the curve to saturation. The other signal alternation opposes the offset and pushes it away from saturation. Sounds asymmetrical. Dynamic, of course, so musicians will like it like trolls like shiny things. But I think the effect still is dependent on volt-time integral adding/subtracting from the offset, so it'll still be frequency dependent, won't it?

If you have 60ma running through half the core from the one se tube and 60 ma from some source on the other side to compensate the tube side idle current, the signal varies on the tube side going from 60ma at no signal to close to zero at maximum possible signal swing. At full swing on the tube the transformer will be out of balance again and saturate on the dummy side. So this idea isn't perfect either.

[Alan0354]

Good point. It's not the transformer so much as what drives it.

And the genesis of a new theory about how to get output compression at any/lower levels. Hook up a third circuit element so it can pull DC through a winding - any winding - on the OT, and this introduces a net DC offset into the core's operation point, which both offsets it to one side of the BH curve, and reduces the incremental inductance.

Similar setup to getting SE operation in a smaller core - use pushpull, but use one of the output devices as just a DC offset, don't feed signal to it. Lets you keep DC in the SE signal output device, but offsets the BH curve back down near center so you can use more of the total BH curve and hence get much smaller iron.

Hmmm. Magamps.

Only magamps use the differential inductance (mostly) for control, and (mostly) arrange to cancel the controlled signal excursions in the windings.

Lessee. Offset BH curve one direction. One polarity signal adds to this and pushes over the curve to saturation. The other signal alternation opposes the offset and pushes it away from saturation. Sounds asymmetrical. Dynamic, of course, so musicians will like it like trolls like shiny things. But I think the effect still is dependent on volt-time integral adding/subtracting from the offset, so it'll still be frequency dependent, won't it?

Now, this is potentially an out of the box idea!!!!

Only one problem with this is the hum. As you know the B+ is full of saw tooth wave from the full wave rectifier. The noise is cancelled with common mode rejection in the transformer push pull circuit. If you just pull current by say a resistor to ground on one side, you imbalance the circuit and the hum comes back. This is like you unbalance the two output tube.

But, if there's a will, there's a way. Good stuff.

[Tom Phillips]

Now, this is potentially an out of the box idea!!!!
Only one problem with this is the hum...

Yes it's a good idea and I'm sure someone has done it as it has been written up in the literature. Seems like there was also a discussion about someone trying to patent parts of the idea. Re: The hum issue. The set up does require better filtering just as is required to produce a quiet Champ type circuit

[Alan0354]

Yes it's a good idea and I'm sure someone has done it as it has been written up in the literature. Seems like there was also a discussion about someone trying to patent parts of the idea. Re: The hum issue. The set up does require better filtering just as is required to produce a quiet Champ type circuit

Do you know the patent number or any article? I am kind of new in tube amp stuff.

[Tom Phillips]

Do you know the patent number or any article? ...

I can't remember the exact articles at the moment. I believe that it is discussed in one of Kevin O'Connor's TUT series books and recently I read about the concept elsewhere. I'll post if I run across it again.

I didn't mention it earlier but any patent application is probably bogus. There have been many discussions about the things modern companies try to (and often succeed) in patenting that are based on previously published technology. They are sometimes successful because of defects in the US patent system.

[R.G.]

If you have 60ma running through half the core from the one se tube and 60 ma from some source on the other side to compensate the tube side idle current, the signal varies on the tube side going from 60ma at no signal to close to zero at maximum possible signal swing. At full swing on the tube the transformer will be out of balance again and saturate on the dummy side. So this idea isn't perfect either.

It's not perfect by any means. But it can help get a smaller transformer.

If you have 60ma from a SE power tube, the current in the trannie can (conceptually at least) swing up to about twice that and then down to zero and still have a kinda undistorted signal. The transformer has to be designed so that 60ma average in the windings and 120ma peak at some low frequency does not saturate the iron. So the operating point of an SE trannie is halfway up one half of the BH curve.

If you use some other current source to introduce an equal and opposite, but DC only, counterbalancing offset, then the net field in the core with no signal is zero. The output device is still turned on at 60ma, and can swing up by 60 and down by 60, but you can now change the winding ratio so the output device can swing a lot more current and get more power out, or alternatively can work at lower frequencies and not lose output power to the BH curve.

At full swing, the SE output tube is still providing 60ma average through the winding, and never more than 60ma from the dummy side +/-60ma from the output tube. The transformer does stay balanced, to the degree it's set up well. You can't necessarily just use a PP output transformer for this, as the higher DC currents will still heat the windings resistively. So it'll be bigger than a PP trannie for the same power, but probably between half and a fourth of the mass of an SE of the same power.

Only one problem with this is the hum. As you know the B+ is full of saw tooth wave from the full wave rectifier. The noise is cancelled with common mode rejection in the transformer push pull circuit. If you just pull current by say a resistor to ground on one side, you imbalance the circuit and the hum comes back. This is like you unbalance the two output tube.

But, if there's a will, there's a way. Good stuff.

Yep. There are ways. The dummy side needs to be a high impedance current source to avoid resistive loading of the active tube through the transformer, so you're probably going to use an active device there. You could remove the signal to one output tube, but why burn up a tube running DC? Use a power MOSFET and some active circuitry to make it act like a real current source. This will not then load down the active tube.

And the ripple on the B+ is exactly the same problem as in an SE amp. But since you're getting nominally twice the power, it may not be too bad. And since you're going to be using some active device for the dummy side, why not feed in an equal-but-opposite ripple to help cancel the ripple. Do it in the active device on the dummy side if you can't do it in the transformer. I can think of a couple of ways to do this, but it might be as simple as hooking a BFC from the B+ line to the side of the dummy winding that lets it counter the main ripple.

Yes it's a good idea and I'm sure someone has done it as it has been written up in the literature. Seems like there was also a discussion about someone trying to patent parts of the idea.

Yeah. It was several years ago, which probably means it was 10. Something like KBF? KFB? amplifers? They were going to revolutionize the amp world. As they said in Spinal Tap, they're currently residing in our "Where are they now?" file.

The US patent system allowed someone to get a patent on the idea of using a series regulator and series resistor to fake a dying 9V battery. The application went in several months after I put it up at geofex. Silly me. I thought it was, as they say, "obvious to one skilled in the art." There are so many examples of Silly, D*mned Silly, and Downright Embarassing patents issued that it ruins my respect for the patent system. Such as that ever was.

[Austin]

I was reading about Mesa Boogie "Lonestar Special" amps and they can switchh from push pull to single ended with the flip of a switch somehow. They don't reveal how they do it but I bet it is like what we are talking about here.

Class A - Exposed and Explained by Randall Smith

[Gregg]

Get a Mark V schematic and you'll see how it's done.