How to optimize when you match power tubes to the output tranny?

[smackoj]

friends and neighbors; I would like to know if there is math formula along with perhaps some meter readings that a tech can use to maximize the match between the OT and the output tubes? To say it a diff. way; if I have a bunch of vintage OT iron and I want to find the one particular OT that most closely mates up with, for example, a pair of EL34s, how would I test the OT and the power tube output to find if they do closely match?

I am savvy to the fact that power tube output is dynamic and that there has to be a 'fudge' factor in this endeavor, but I really want to be able to squeeze the best lemons to make the lemonade....so to speak hermanos.

thx in adv. and Merry Christmas to the brethren of tube tone.

Jack D     

[punkykatt]

I would say a turns ratio test would be your best bet.   I don`t have the formula right at hand, but if you do a google search on out put transformer turns ratio testing you should find it.
Merry Christmas

[PRR]

It's like "matching" a light-switch to the light.

You don't actually match a 0.5A switch to a 60W light. You find a good-enough switch that is not too expensive. Socket-switches may be 6A for a few pennies. Wall-switches may be 15A-20A for cheap. Yes if you are lighting a ball-field with 350 Amps of light you have to go up-size, but will probably use a 500A switch.

You match the OT to the _PT_. Then find tubes good-enough and not too expensive.

In push-pull: Say the PT and rectifier can deliver 400V at 100mA (0.1A). That is 400V/0.1A= 4K. By push-pull leverage the OT end-to-end should be twice that, 8KCT. The full-power input is 400V*0.1A or 40 Watts. In self-bias, the tubes have to dissipate that, 20W each. 6L6 EL34 are ample, 6V6 EL84 are not. In fix-bias the tubes can idle cool but the part-power dissipation calculation becomes a bear.

Here's the simple way. Buy PT and OT from a "standard design". (Fender Deluxe, Marshall 100W, etc). Use tubes as large as the original (EL34 will do 6L6 work and usually vise-versa, etc).

[HotBluePlates]

... if I have a bunch of vintage OT iron and I want to find the one particular OT that most closely mates up with, for example, a pair of EL34s ...

Then the shortest path is to note type and number of output tubes the vintage piece used, and stick with that. You know that will match properly.

If you're designing from scratch, PRR gave the correct method. But there seem to be only so many options out there, and most of them conform to a known-plan. So the easiest method is follow the plan.

You could re-design from scratch, pretending the plan doesn't exist. You may land on a subtle variation, but you'll find most of your "original" designs come very close to a known plan due to the small range of available parts, if nothing else.

[smackoj]

I'm getting the picture from the best minds and I say thanks for your thoughts and direction. If I could just open up this window aka "box" to explain my interest in this realm.

Many of the fellas on this forum have much greater knowledge about tone "stacks" , clipping diodes and OD signatures. For them, that's where the fun is in building and modding amplifiers. That's all good and I enjoy reading the personal likes and philosophies but it's too complex for me at this point in my learning process. I could spend a week trying to copy someone's new twist on the Dumble stack and I would most likely be tired, hungry and out of sorts 'cuz the thing just squals or pops and smokes. But I'm more simple in the fact that I try to find guitar expression more with organic recipes, i.e. I like to hear the sound of the wooden body of the guitar mixed with the sound of high quality pickups (sort of the same family of parts as an OT I believe?) so I'm staying with the basic tenets of the early amp platforms.

In other words, I want to experience the best that the mixture of performer, guitar and amplifier can produce at any given moment in time. So, for me, I want to stick with the belief that the 3 most important parts of the amp for making sweet tones are the speakers, tubes and transformers. Surely some will not agree with these parts as the "most important" but it's a simple concept that I can understand and changing these three items definitely is something I can do with good results on a regular basis.

Speakers are a lot of fun to experiment with but they are bulky, heavy (shipping cost) and costly in general for an enthusiast on a budget.
Tubes are lots of fun to change and do critical listening tests on. This part of the tone search makes sense financially and gives a lot of satisfaction without starting a range war in my home. So, this is the jump off point to the original question about matching the OT with the output tubes.

An OT, in my way of thinking, has a soul of its' own if you will. All that groovy guitar power juice rushes in one side of the transformer carrying a mixture of the player's expression along with electrical host organism. Then, it does some kind of a metamorphasis transformation jumping across the cosmic wasteland and then rushing out the other side heading for the "light".....sorry if I'm sounding all dramatic but I'm trying to make a point that is not easy to explain. So, the musician's expression going into the OT is set free and he/she waits for the sound to arrive on the other side to see if it is pleasing and inspiring, or if it is lifeless, and dissapointing.

So, as senor PRR has said, we match up the OT and power tubes in the same fashion as you would a light bulb and a switch. However, that circuit is not carrying expressive music so there's a large X factor with the music and the OT that a bulb and switch do not share IMO. I do understand that from a KISS pt. of view, using known iron which, for example, originally ran two 6L6s, is the safe bet and pretty much "known" results should follow. But, more along the lines of Hot Blue's contribution which sees the possibility new things by not following a known pathway or "plan" 

I realize that from the standpoint of buying new iron from the outlets that now exist, there is a limit to what can be obtained. But, there is a fairly large pool of vintage iron available and New/Old iron just keeps getting unearthed as people rumage through grandpa's garage so to speak.

So, to conclude, I guess I said all this to just explain why I asked the question in the first place. I want to KISS and I want to explore any and all options to set the 'mojo' free out of this strange thing called a transformer. I don't want to spend weeks modifying a BF Fender tone stack, I want to see if there is a ghost in the machine, or a genie in the bottle without frying any vintage parts and with as much forethought and theory as I can muster to make it at least mildly 'controlled'.


One final ?  if I have a vintage OT that was wound to accept 6500 ohms primary and deliver 8 ohms secondary, but I want to mate it to a pair of 6MB8 pentodes which, for the sake of the example, send 8700 ohms to the primary, can I add resistance after the output and before the OT with good results? I have not seen this done that I know of so I would like to hear other's ideas?  thx

[HotBluePlates]

... An OT, in my way of thinking, has a soul of its' own if you will. All that groovy guitar power juice rushes in one side of the transformer carrying a mixture of the player's expression along with electrical host organism. Then, it does some kind of a metamorphasis transformation jumping across the cosmic wasteland and then rushing out the other side heading for the "light"...

Beware. Your sense of romantic artistry sets you up to fall victim of audiophile snake oil salesmen.

But I don't want to steal the magic and romance from your experiments...

So here's what I propose:
Find a way to tinker with the speaker load you're applying to the OT secondary. Why?

... a vintage OT that was wound to accept 6500 ohms primary and deliver 8 ohms secondary ...

This is backwards of reality. Don't worry, I and everyone else started off thinking this was how transformers work.

The impedance presented by the output tubes is very much less than the output transformer primary impedance. In fact, it has to be to allow tube current to flow. Instead, a transformer works by having a defined load attached to the secondary, which is then "reflected" back to the primary. The reflected impedance is altered, as by a lens, to appear bigger than the secondary impedance by the ratio of the number of windings on the primary compared to the windings on the secondary.

A handy mechnical analogy is that of a lever. If the primary and secondary had the same number of turns, the fulcrum of the lever is halfway between the primary end and secondary end. As the number of primary turns gets bigger than the number of secondary turns, that fulcrum moves more and more towards the primary end, and the small secondary impedance looks like a big impedance to whatever is attached on the primary.

But more to your point:

If I have a vintage OT that was wound to accept 6500 ohms primary and deliver 8 ohms secondary, but I want to mate it to a pair of 6MB8 pentodes which, for the sake of the example, send 8700 ohms to the primary, can I add resistance after the output and before the OT with good results?

To what end?

Actually, it is probably not a good option, as adding resistance means adding a place where output power is converted to heat. Ohm's Law tells you that with current passing through a resistance, there is a voltage dropped across the resistance, and the equation for power tells you the product of that voltage drop, and the current producing it, is the power dissipated in the resistor. That power is subtracted from the output power your amp would otherwise produce.

[HotBluePlates]

Why is the primary impedance you were shooting for so important?

PRR explained that you match the OT primary impedance to the power supply components you are choosing to use (and ultimately, you're picking both for the resulting output power to the speaker that you want). How much output power is needed directly or indirectly dictates things like how much voltage/current the PT will supply, how high the B+ will be, what type and how many tubes are needed, will you operate them class A or class AB (bigger power will require higher B+ which may preclude class A with a given tube), whether self-biasing is feasible or if you must use fixed bias, how much bias voltage is needed, and then what type of phase inverter is practical in light of all the other considerations.

Wth so many factors riding on the required output power, an OT primary impedance is chosen along with the B+ voltage, current and tube type to land at the desired output power. The amount of distortion may also be a key factor in this balance; if 2 different primary impedances are considered, and at the same output power one has a higher level of distortion, then technically that one also has less clean output power.

So many of the old "correct" or "ideal" primary impedances resulted in maximum clean output power with a given tube type and supply voltage. If you deviate from the ideal, you will get less clean output power and the type of harmonic distortion in the output stage may be shifted from odd to even, or vice versa. Catch is, a proper push-pull output stage should cancel all even harmonic distortion created in the output tubes, so less odd order/more even order distortion may sound subjectively like the same amount of clean power. There might be a shift in the quality of the distortion.

But all that is complicated if you have a feedback loop around the output stage, like most 20w+ amps. These tend to minimize/mask output tube distortion until you run the signal level up so big the feedback runs out of steam and the output distorts more abruptly.

Further complication: your speaker is not one impedance, all the time. It is its rated (and lowest) impedance at a narrow frequency range, often around 200-400Hz. Any frequency above/below that results in a higher impedance presented by the speaker, and through reflection, on the primary of the OT. That increases distortion, reduces output power, and is a big reason feedback around the output stage is used in the first place; to make the amp more immune to impedance changes at the speaker.

[HotBluePlates]

So now that I presented most of the problems, what should you do?

Feel free to raise/lower the impedance of the load attached to the secondary, aka yours speakers. With an 8Ω output tap, add a second 8Ω speaker. Put it in parallel for a 4Ω load, or try series for a 16Ω load. In this way, you'll try the stock primary impedance, stock impedance/2 and stock impedance x2.

Maybe using the same type of speaker as your original one will help avoid perceived tone changes, which are really the result of a different-sounding 2nd speaker.

If you don't get earth-shattering tone changes with that big of an impedance change in either direction (-50%/+100%), I'd doubt a 25-33% change between 6k and 8k will do much to your sound.

Then the next place to look is the core size of the OT.

If you have 2 transformers, both rated for 6k:4Ω, one at 20w and the other at 100w, what's the difference?  (Pretend you can find a 100w, 8k transformer)

There may be winding differences, but mostly the wire used for the windings will be a little thicker, and the iron in the core will be a LOT bigger in the 100w part.

Look at the claimed frequency response and what power level it is claimed to be rated at.

Many guitar amp OT's are claimed maybe 70Hz-10kHz, often at 1w or a part-watt. Hammond OT's intended for Hi-Fi use are generally rated for 30Hz-30kHz at full rated power for the part (100w for the 100w example transformer above). The wider the claimed frequency repsonse, at the higher power the claim is made, the bigger and heavier the core will be.

Go the opposite direction. What if you take a 20w transformer with 30Hz-30kHz claimed repsonse, and try to shove 100w through it? Assuming you don't melt the windings (and you probably won't), the frequency repsonse will be more constricted; you lose the extreme lows and highs.

Since guitar amps often have significant output tube distortion where we like to run them, and full response to the extreme ranges of those harmonics will probably sound harsh, the limited frequency response may turn into tone shaping. The transformer itself can also contribute its own layer of distortion, which may be subtle or dramatic depending on just how much bigger the input power is than the intended power handling. This is used as a tonal tool in clones of the Neve 1073 mic preamp (the analogy to guitar amps is somewhat limited because of the very different nature of the transformers used in a mic preamp versus those in guitar amps). Cut to the chase, 1073 users generally get a darker and more distorted sound when overdriving the transformers in their preamps.

So you may draw the conclusion that if you want pristine clean with wide frequency response, you want a big heavy OT that otherwise has the primary/secondary characteristics needed. A complex, distorted amp with limited frequency repsonse may call for a lightweight OT with a smaller core and otherwise the same primary/secondary impedances.

Example: My Standel 25L15 has extremely wide freqeuncy response and is clean as a whistle. Part of that is the circuit design of the amp, and another big part is the JBL speaker, which was originally intended for full range Hi-Fi use, not limited-range guitar amp use. But the 25w output stage also uses a 60w OT rated for 30Hz-30kHz at full 60w of output. I haven't investigated the parasitic characteristics of the OT to see if 25w duty actually extends the frequency repsonse even wider. Why such a big lump of iron, so overrated for the actual use? According to a trustworthy source who reverse-engineered an original 25L15, this OT was the closest match for the size of core used in the original. So this 25w amp uses a bigger, heavier OT than a Twin Reverb (8lbs for my OT vs 5.3lbs for the Twin).

What that means is the OT is removed from the equation in altering the amp's tone. OF course, the speaker I'm using could either be viewed as eliminating the speaker's contribution to the sound, or as a major contributor, because the speaker's frequency repsonse is 30Hz-17kHz. That's an enormous change from typical guitar amp speakers.

So bottom line, you may find after tinkering reflected primary impedance that OT core size is the biggest tonal variable when dealing with just an OT change. If you can ever find pictures (or an in-person example) of super-budget amps of a given power level and more expensive amps, note how much bigger the OT's are in the costlier amps. The small OT may contirbute to the cheaper amp's funky charm.

But amps are a balance of a LOT of competing factors, and the OT is just one of them (and often not a factor with the OTs many people choose).

[smackoj]

Wow! that's quite a lesson in OT theory and practice. Thanks for taking the time to expound on it. I will try briefly to summarize what I am thinking after your in depth replies. I will have to read this more than just a few times to absorb the theory and/or reality of what is really going on with this "match-up" between the PT, power tubes, OT and the whole "power output" part of the equation I wasn't considering. I specifically want to spend some time getting a mental picture of the OT impedence vs. reflective impedence. I can understand how that piece of the puzzle may be sort of the "X factor" , or one major X factor in my original ideas about how to "optimize" the output transformer to improve tone and/or the general quality of the sound of the guitar amplifier.

I should note that in my original question, I was mainly interested in tone improvement and not considering power output and how that is affected by the factors Senor Blue and PRR discuss. And, at this point, I will simply conclude that my whimsical idea that finding that one special transformer that is heads above all the others soundwise, is much less likely due to all of the parts that are connected on both the primary and secondary sides of the tranny that have a lot more effect than I originally understood. And, I should include here other facets that create changes in tone like the feedback loop and just simply adding another speaker and changing the ohm load the tranny sees on the secondary.

Maybe naivete is to blame for my search for that 'special' output marriage of parts and secret mojo that creates the irresitible sound on the yet to be recorded Led Zep 2nd album? I know Blue is leary of 'snake oil salesmen' and I am too, but I do believe in miracles and I still believe that the way the musical notes are transformed from the primary winding, through thin air, and then show up on the secondary side to be enjoyed by the listener, has a spiritual component that cannot truely be explained.....but, that's just one man's opinion.

"I'll be back"  duty calls