OT, the bigger the better? And which one to choose?

[exclamationmark]

Fryette has a 1U 80 watt(!) power amp rack unit that uses 6l6's. The only way way I'd assume that were possible were if a toroidal OPT was used. Mind you, this isn't the type of thing you'd overdrive the crap out of, and is intended more for modelling amplifiers or rack mounted pre-amps.

[Gregg]

Other type of OT can be used as well as seen from pictures here:

http://www.facebook.com/FryetteAmps

[Alan0354]

So you have all the many layer interleaving, but, I have not seen anyone definitively come out and say it does sound better!!! Anyone can say it is superior to cheap OT? remember guitar amp don't really amplifier beyond 5KHz AND if you put is tweeter in a guitar amp, it don't sound good!!! It just give you chirps!!! Regarding to phase margin, it is very easy to compensate without rolling off the highs. This is not an issue. Just read any opamp book and you'll find hundred and one way of taming a closed loop feedback system. You just play with the poles and zeros, or using Laplace Transforms. For open loop power amp( no feedback), you don't worry about phase margin.

So it boils down to the sound for guitar amp....ONLY. How does it sound different particularly when over driven? How does it distort? I thought the distortion and compression of OT is part of the beauty of a tube amp!!!

I am not challenging the expert here, I just want to know and get to the point.

[Chuck H]

Well... We're wading back into subjective territory again. Some players or builders like the bigger, tighter bass and overall more brash sound of bigger iron. Others, not so much. And as mentioned before, it's only one part of a whole system. Some circuits sound better for some people with marginally rated cheaply built transformers. And vice versa. The Hammond 1608 I use in one of my designs is a hi fi OT that is bigger than other OT's used for this type of amp. It also used five interleaves IIRC. I like it very much. Sounds better to me than the Heyboer I ordered for this project. Now... That said...

I also have a small amp I built from spare parts that has a very small, straight wound OT in it. It has a very different sound. But it also sounds very good.

I think the tonal difference between big and small iron has been covered already. I don't use every spice in my cabinet for every dish.

[Alan0354]

Ha ha!!! I am just cheap, I will try my best to use cheap part until I am convinced that there is no other way but to use the expensive parts. Design effort is a one time thing, but expensive parts will be in each and every single amplifier. Hack, I am not working, if I am not spending time designing, I'll be studying or watch tv, or talking on the forum!!! So my time is not exactly money!!!

Soooo...........I started with a cheap Magnetic Components el cheapo 40W Vibrolux/Bandmaster OT that is $46. I will squeeze the last drop out of that until I can squeeze no more!!!

[jpfamps]

So you have all the many layer interleaving, but, I have not seen anyone definitively come out and say it does sound better!!! Anyone can say it is superior to cheap OT? remember guitar amp don't really amplifier beyond 5KHz AND if you put is tweeter in a guitar amp, it don't sound good!!! It just give you chirps!!! Regarding to phase margin, it is very easy to compensate without rolling off the highs. This is not an issue. Just read any opamp book and you'll find hundred and one way of taming a closed loop feedback system. You just play with the poles and zeros, or using Laplace Transforms. For open loop power amp( no feedback), you don't worry about phase margin.

Laplace transform is hard if not impossible to use meaningfully on solid state power amps where you have a fighting chance of being able to calculate the poles accurately, let alone a valve amp.

So it boils down to the sound for guitar amp....ONLY.

And reliability.

[Alan0354]

Laplace transform is hard if not impossible to use meaningfully on solid state power amps where you have a fighting chance of being able to calculate the poles accurately, let alone a valve amp.



And reliability.

There is no difference whether it's a tube or SS amp. It's all about poles and zeros. You adapt with different output impedance and condition.

Laplace transform is very useful in closed loop feedback system. Negative feedback of opamp ( this kind of power amp) are just the very simple version of closed loop feedback system.

I am just not very good at it and I use Bode Plot for closed loop feedback design. Believe me, I am very involve in Physics Forum advising college students and engineers. There are experts there that tame closed loop feedback purely by LaPlace transform like Jim Hardy, I am more the RF transmission line, EM, EMC and transistors. This type of power amplifiers are the simplest of the simplest of a closed loop system. You have no dead time type of thing to deal with, you pretty must do dominant pole compensation for stability. Laplace transform will work beautifully. These kind of opamp type power amp and switching type power supply is the easiest of the closed loop system design.

As for the topic here, if it is just phase margin people are worry about, that's is not even an issue. There are hundred and one way to do pole zero compensation and tame it with the right frequency response. If there is other special characteristics, then it's a different story.

[Chuck H]

Almost moot since the feedback loop in any guitar amp is a very short path. Not at all the "global" (output to input) type circuit seen in any hi fi amp. Guitar amps typically only have a loop between the PI and the OT! The phase margin is not nearly as hard to control as a global circuit.

[Alan0354]

Almost moot since the feedback loop in any guitar amp is a very short path. Not at all the "global" (output to input) type circuit seen in any hi fi amp. Guitar amps typically only have a loop between the PI and the OT! The phase margin is not nearly as hard to control as a global circuit.

True, I remember I experiment with the Bassman 100 with open loop( open the 820ohm resistor) and compare, the volume only goes up about 3 times. This goes to show the open loop gain of the power amp stage is very low. There is really nothing to worry about.

This further proof the point that the higher frequency response and extra phase margin that an expensive transformer with interleaving is moot in this sense. I can sure see it is important for audiophile amp that need to work to 20KHz and down to 20Hz or below. Then the inductance, coupling, phase margin, frequency response become important and you need to pay an arm and a leg for the OT. Again, we are talking about 50Hz to 5KHz here. I just don't think the expensive OT will have much advantage on this.

BUT how about the compression sound? obviously, the OT do compress and must give certain characteristic of the guitar amp, how does the OT with interleaving perform in this sense?

[Alan0354]

I guess let me ask the question in a different way. How were the OT wound on some of the old Black Face Fender Deluxe Reverb, Vibrolux reverb, 59 Bassman. Marshall Plexi, Vox AC30? Are they cheap wound or interleaving wound?

I don't put a lot of weight on the old old tweek stuffs because they likely were the first generation using hifi components. I mainly looking at the second generation successful amps that defined the trend......like the few I mentioned above.

[Steve Conner]

As far as I know, they were all either straight-wound or once interleaved: the cheapest and second cheapest options respectively. If I didn't know better, I'd say that the amps with feedback would be once interleaved, and those without feedback (AC30, 5E3, 5F1, etc) would be straight wound. However, I could be wrong, I believe jpfamps or someone has a recipe for a Vox AC30 transformer.

Compression can be thought of as a dynamic limiting of the bandwidth. As the transformer core saturates (and as I pointed out above, this only happens from DC imbalance, not loud signals) its magnetic permeability decreases. This lowers the transformer's magnetising inductance, causing the bass end to be high-pass filtered. I think it also increases the leakage inductance, low-pass filtering the treble. This effect would be lessened by interleaving since there is less leakage inductance to start with.

Peavey's "Saturation" circuit was a crude model of this bandlimiting effect, controlled by a front-panel knob instead of dynamically shifting with the DC imbalance. It's described in the Hartley Peavey white papers.

Laplace transforms are all very well, but they are a linear model of the system, and the "mojo" of guitar amps is all about non-linear, large-signal effects.

[jpfamps]

Laplace transform will work beautifully.

Assuming you know all the poles and zeros, which you don't.

The "elephant in the room" here is the output transformer. You would need to know the primary inductance, leakage inductances, the stray capacitance under large signal conditions. It's hard enough to find any of these parameters for a hi-fi amp OT from manufacturers data sheet (where someone actually might be interested), let alone for an OT for a guitar amp.

Furthermore the leakage inductances and stray capacitances will almost certainly be different for either side of a push-pull transformer (important when driven into class B loading), and will vary with the secondary taps used.

Regardless, as Steve points out, you are working outside the linear range of the system.

[Amp Kat]

Kevin O'Connor's London Power amps sport toroidal OTs.

Can't really comment on their efficacy as I haven't tried one (nor heard any sound samples), however I can see DC imbalance being a major issue in a guitar amp (or indeed a hi-fi amp).

Speaking to someone who winds toroidal mains transformers, interleaving becomes a major issue due to the fact the hole in the middle of the transformer soon gets full of insulation.

Those Toroidal designs were in Priciple Of Powers book and they were Designed by Menno Van Der Veen who also has a book out called Transformers and Tubes where he discusses his designs and talks alot about Toroidal and frequency in the complex domain. The subject in itself is quite complex and has a lot of Engineering terms and formulars with all the long algebra that goes with.

Van Der Veen teamed up with Plitron and put out the Output Transformers and Power Trannies to support the builds which are the same ones listed in Kevin's book and don't find it a coincedence that Plitron and London Power or both in Ontario Canada. The amps are very clean and as a matter fact could be subbed for Hi-Fi with a few small tweeks. As mentioned above^^ one of the problems or should I say characteristics of the Toroidal OT is DC imbalance and DC servo's were recommended to keep the sides in check but I have used a Plitron Toroidal power supply in an SLO-100 design and was very clean and very powerful but you have to be careful because the wires are very small on the secondary side and you have to use Full Wave Rectification but there is no imbalance problems whatsoever with it and worked almost perfect and sounded very nice. Haven't ventured into the OT yet but I noticed the prices have come way down from where they were and not sure where Menno stands with Plitron at this time.

[Alan0354]

Assuming you know all the poles and zeros, which you don't.

The "elephant in the room" here is the output transformer. You would need to know the primary inductance, leakage inductances, the stray capacitance under large signal conditions. It's hard enough to find any of these parameters for a hi-fi amp OT from manufacturers data sheet (where someone actually might be interested), let alone for an OT for a guitar amp.

Furthermore the leakage inductances and stray capacitances will almost certainly be different for either side of a push-pull transformer (important when driven into class B loading), and will vary with the secondary taps used.

Regardless, as Steve points out, you are working outside the linear range of the system.

Both Bode Plot or Laplace Transform need to know the poles and zeros, without that, there is no way to design, that is always the hard part. But you need to know those.

That said, you don't calculate, you test and observe!!! You drive a signal with the same input driver and/or with the same load, you use a signal generator to drive the system and measure the amplitude and phase of the input and output to get the poles and zeros. This is done when you work with control system. You don't care about the mechanics and the theory of it, you just observe the poles and zeros.

You drive and different signal level and observe the phase shift. Remember in Bode Plot, ALL you need is to guaranty you are less than 180 deg at 0db cross over. Then you pulse the system and see the overshoot. You do it even when there is distortion.

Most of the closed loop feedback control are much more complicated and I did it the same way. There is always one part that you have to experiment and observe. There are feedback control system books that you can study. You DON'T experiment and come out with something that you think is stable. This is how people get into trouble in production. This is ALL done by theory. You see switching power supply from even manufacturers that has to spec minimum current before it gets noisy. That is bunch of crock!!!! They don't know how to design and a lot of them fake it by playing with it. This is called conditionally stable.....that it is stable only in limited range!!! They did not do their homework. I tame switching supply even though I had my engineer did the magnetic and switching design, I did the closed loop control part of it. Most of the power supply engineer think if you put more cap, you drown the oscillation, that is exactly how they gone wrong. Bigger cap seems to make it harder to start oscillating, but ones you pulse and kick start it, it sing like a bird.

"Experiment and see" work for guitar amp, but most other design don't. Tube amp is a very unique thing in electronics, you don't carry the same design philosophy to the rest of the electronic design.

[Alan0354]

Compression can be thought of as a dynamic limiting of the bandwidth. As the transformer core saturates (and as I pointed out above, this only happens from DC imbalance, not loud signals) its magnetic permeability decreases. This lowers the transformer's magnetising inductance, causing the bass end to be high-pass filtered. I think it also increases the leakage inductance, low-pass filtering the treble. This effect would be lessened by interleaving since there is less leakage inductance to start with.

I thought loud signal drive more current thought. Only true class A cancels out and don't matter. With class AB, depend on how deep it is, part of the cycle the current don' cancel out and I thought this will cause saturation!!!

Peavey's "Saturation" circuit was a crude model of this bandlimiting effect, controlled by a front-panel knob instead of dynamically shifting with the DC imbalance. It's described in the Hartley Peavey white papers.

Do you have a link on this one.....please!!! I love to read it.

Laplace transforms are all very well, but they are a linear model of the system, and the "mojo" of guitar amps is all about non-linear, large-signal effects.

Thanks.