Choosing power transformer.

[Danelectron]

What are all criteria for choosing (sizing) a power transformer for a guitar tube amp design?

I have been reading a lot of publications on amp design. New and old and am finding a lot of good info on circuit design, tone sections etc., but not much on the big picture. Is it something like deciding on a given output power, selecting the tubes to accomplish this and adding up requirements from tube data sheets?

Can anyone refer me to a good publication that breaks down what happens in an amp mathematically while showing current paths in the circuit?
These questions may be vague, but then so is my understanging at this point. I have had some electronics ed but it has been a while and I'm working filling in gaps. Had a lot of math too, but that has really gotten buried. We used to do temperature analysis of circuits at given time increments and how this affects individual components in a given design. Like in car radio circuits and such. Any more I don't know how much of that is done. It seems like the idea is how big of a fan or air conditioner to use. Anyway I digress.

Ahy help is greatly appreciated. Thanks, Dan

[Diablo]

Here's my approach. Decide how many watts of power you'd like. Decide which tubes you'd like. Decide how many of each tubes you plan to run. Add up the heater current draw for all of the tubes. Add 20-100% more current to be conservative on the heater draw. Decide on how much B+ voltage and current you need from the transformer. You can use similar power amps using similar output tubes as a guide. Preamp tubes don't draw much power (ignoring the heater circuit). Power tubes do. Decide if you want a tube rectifier. If so, you need 5V heater current supply from your transformer - 3 amp covers most tube rectifiers.

PT General guide:

2 - 6L6GC power tubes, 40-50 watts power, 440-480 V B+, around 200 mA.

2 - 6V6GT power tubes, 18-22 watts power, 400 - 430 V B+, 120-150 mA.

If your PT has extra capacity for current, then it will run cooler and last longer, and will be a stiff power supply with constant voltage. It will also weigh more and cost more. Take a look at replacement power transformers for some of the popular/classic amp designs to get an idea of reasonable current capacity for the wattage.

[Danelectron]

Diablo,
That has pretty much been my M.O. to date, but I can't help but wonder how much the voltage drops across components other than the tubes and different values etc. tax the power transformer.
My intention is to try to get a clear picture of the engineering involved in amp design. In the long run to be able to calculate all current flows and voltage drops through each component in a design, build it and see how close my math is. Is that overkill from a design standpoint? Also, I'd like to be able to calculate frequency bandwidths in the tone sections and a obtain a clearer understanding of impedances in the power and output sections.
I know tube amp design isn't an exact science due to relatively wide tolerances, but I'd ultimately like to be able to design something and come out with reasonably accurate and predictable actual measurements.

I came upon these questions when looking to build a design with 2 KT66s. I was going to model it after a fender twin schematic (5D8 or 5E8)that used 2-6L6s, and began to wonder about the differences in current requirements. They use different rectifier tubes and neither schematic gives any actual specs of the transformers.
I would also like to be able to take a transformer such as one I salvaged out of an old organ and be able to design a circuit around it that would be functional and use optimally the available power. How does a person figure out how much an unknown transformer is capable of? Impedances and such. I have measured all the voltages and know it supplied 2-6v6s and 20 or 30 (at least as I was told) preamp tubes.


What are comfortable operating temperatures?

What determines impedance of a transformer?

I have many more questions, but think this may make my vague intro to this thread a little clearer.

Thanks for your attention, Dan

[yunger]

how critical are filament supply ratings? say the circuit required 2.4A and the PT was rated at 2A...what would be the effect?

[Danelectron]

Yunger,
Good question and right in line with mine. I can say the result would be that the transformer would run hotter, but how much?
I would think 20% is significant. Whether the lower current supplied merely makes the tube output lower or actually does damage to the tube or related components. If the filament has to operate at a given power, the lower current will result in a higher voltage across the filament. I imagine there is a certain percentage over advertised ratings that the transformer can operate safely, but given the variable tolerances of the heaters it could be risky not staying right within the specs of the tube/transformer. As Diablo suggests it is a good idea to figure on supplying more voltage than is necessary for the tube sections and compopnents.
I'm not trying to supply an answer, just adding food for thought that may draw an answer.
Thanks for chiming in.
Hang in here. I'm hoping for some clarification.

[Diablo]

If the transformer winding can't deliver the nominal current rating of all of your tube heaters, the voltage supplied will be less than the 6.3 volt nominal rating of the tube heater filaments. So, just build in a little safety factor in the transformer spec for the nominal amperage to keep the tube heater voltage within the proper range.

Here's a handy link for tube amp design.

How to design valve guitar amplifiers

[Danelectron]

Diablo,
That IS a good link. Clear explanations with math where applicable. In his publication as well as most of the others I've been reading it does seem that tube amp design is pretty loose. A common link to everything I've been reading is " if that's what your looking for, try this". Predictable, but not definable. I tend to over think things.
As far as trying something, what would your prediction be if a person used an unknown power transformer that is physically about twice the size as a marshall 100 watt output transformer, and attempted to run 4 6L6s? The voltages are all there, but if it can't supply the current what would be a worst case scenario. I may never know or be able to figure out what the specs are, but I may surely find out what they're not. Science at it's most basic form.The real challenge in designing something around this is that there is no bias tap. I will have to use another transformer to set up a voltage divider for this.
Whatever happens I'll have fun with it and learn a few things.
Thanks for staying tuned. Dan

[Old Tele man]

Diablo,
That IS a good link.

...the author is a contributing poster here too.

[Diablo]

Diablo,
As far as trying something, what would your prediction be if a person used an unknown power transformer that is physically about twice the size as a marshall 100 watt output transformer, and attempted to run 4 6L6s? The voltages are all there, but if it can't supply the current what would be a worst case scenario.

You could always try running the amp first on 2 6L6 tubes and measure the B+ voltage under full load (amp cranked up), and run it for some time to see how hot the transformer gets. Then if it handles two tubes OK, stick in 4 tubes and repeat the test and see if it delivers the same voltage under full load and how hot the transformer gets. Worst case scenario is if you plan on running the amp at full volume for a long time, you may overheat the transformer windings. If you intend to use the amp for light duty, half volume etc., then you won't tax the transformer. 4 6L6 tubes makes for a loud amp by modern standards. Back in the 1970s, folks wanted that kind of power and more in a guitar amp....not recently though.

[Danelectron]

Old Man Tele,
That is awsome. I've been going over the preamp design section and the design process is getting clearer. I want to really nail the concepts here before moving to the power section. I intend to purchase the book. I have the utmost respect and appreciation for anyone who will give of themselves (as I'm finding a lot of here) without concern of personal gain. I know how precious time can be so feel compensation is in order. What is his log-in, maybe we can prod him for his power section book .

Diablo,
That is good advice. I intend to go ahead with the build and if the power unit isn't up to snuff it will be worth it to go ahead and get a proper one. This thing is so large I'm getting the Tim Allen (Tool Time) syndrome. I'm tempted to do curls with the thing or fire a wrist rocket at it.
I do like to play loud. I have a Marshall TSL 100 and like to max it. Probably not good for my ears, but Ohhhhh yeah. Such a good ring I don't even notice my tinnitis. Standing in front of the cabinet and feeling the low notes hitting me in the chest. When I was younger I used to work on a farm in the summers and would haul an old peavy duece I had up to the hay barn and blast away. I could only imagine the tender steaks comin' off the neighbors cattle. Sweeeeeet! Played a lot o' the Clash back then. Anyway, enough of that. This is going in a 2x12 combo cabinet with an external jack to hit the marshalls 4x12 at will.
Thanks for the responses and I'll keep you all posted on the progress.

[kldguitar]

Power transform need

It is simple problem and complicated problem.
Simply problem means you only supply power to tube plate and lighter, it is o.k. Commonly second Class of PT has two output, one for light(6.3VAC), One for B+( it is depend on difference tube you used).
Complicated problem means people must balance between power and cost. The more power is , the better the sound is , the high cost is. Please watch our amp GT30 power transform
4xEL84, load voltage 260V, Current 0.35A power 64 w. It is two times.
More power can supply larger power for dynmtic needing.It is reason why people still like tube not transistor amp.

[Chuck H]

Bonus KLD. And +1 on your post.

Nice to see you contributing. It feels good to say welcome aboard.

Chuck

[Danelectron]

I just had a thought. I measured 715 volts on the secondary HT output. Could I place a resistance(say 350R) across this to get 200ma of current flow and monitor temperature to get a possible scenario? I could at the same time hook up 1.5 ohms across the heater tap to get 4.2 amps of current flow so I could simulate the heat generated in the transformer by the heaters also. The heater tap doesn't worry me so much because this thing operated 20 or 30 preamp tubes plus two power tubes. I don't recall exactly how many of what. Would this be a realistic test? These resistance values would allow just over the minimum necessary to operate these tubes (if my math is O.K) so would at least tell me if the transformer is in the ballpark. If it runs cool enough I could try slightly lower resistances to see how far I could go. I have a bunch of wirewound high wattage / high value resistors that I could parallel to get down to 350 ohms with well over 200 watts to be within necessary parameters.

[streamz1]

It is simple problem and complicated problem.
Simply problem means you only supply power to tube plate and lighter, it is o.k. Commonly second Class of PT has two output, one for light(6.3VAC), One for B+( it is depend on difference tube you used).
Complicated problem means people must balance between power and cost. The more power is , the better the sound is , the high cost is. Please watch our amp GT30 power transform
4xEL84, load voltage 260V, Current 0.35A power 64 w. It is two times.
More power can supply larger power for dynmtic needing.It is reason why people still like tube not transistor amp.

Where can I get a suitable power transformer replacement for GT15-H

[jazbo8]

Duncan PSUII

Can anyone refer me to a good publication that breaks down what happens in an amp mathematically while showing current paths in the circuit?

I'm surprised that someone has not suggested the Duncan PSUII yet... The program does exactly what you are asking for, plug in the current requirements with the rectifier and filter topology you have in mind, and run the simulation. If done properly, it can give you a good idea of the transformer specs (finding one that actually meets the exact spec is something else, there are always some compromise to be made...)

While, it's not a "publication", I think it's better in someways since it's interactive. But, if you are just looking for reference material, Merlin's Power Supply design book is also recommended.

Jaz