dechiphering an output transformer

[Phil_S]

Unfortunately, after investing time in an internet search, we tend to think we are almost "there" and we keep going, maybe loose track of time. There is a relatively new cultural phenomenon. We tend to think the internet knows all. But it doesn't and sometimes that is hard to see. It won't tell you how to double your money quick, or who the next president will be, or what the tx specs are. OK then. How you use your time is totally up to you. I'm happy to help and won't hold it against you

BTW, it was a fairly easy task to lift the text to Excel and split to columns using the space as the delimiter. It took me all of 30 seconds.

I checked your math. It was good until the last column. At that point, use what you know, that the taps are 4Ω, 8Ω, and 16Ω. (Omega is Alt+234.) Use those values as assumed correct values. Impedance ratios will yield:
507 @ 16Ω = 8114 or 8.1K
1026 @ 8Ω = 8212 or 8.2K
2061 @ 4Ω = 8245 or 8.2K
Basically, the primary is about 8.2K. It's normal to see some variance from tap to tap by this method.

[martin manning]

Yay!

[nworbetan]

Oh neat, I diagnosed exactly what his math error was without even seeing his math.

[martin manning]

Since the transformer has no particular primary impedance it can be assumed to have either 2, 4, and 8Ω taps and a 4k primary impedance, or 4, 8, and 16Ω taps and an 8k primary. [edit, see below]

[pjd3]

Thanks everyone for taking a look and giving your input.

The original amp had a pair of 6L6GC's, and I had measured 425 plate volts on each tube back in the days. I think I biased it once. I did base my math on the assumption that for the most part a pair of 6L6GC's are calling for around a 4K primary impedance should the secondary taps be outputted to the "intended" speaker. If there is an 8 ohm speaker connected to the 8 ohm jack, then the primary impedance should be around 4K which usually seems to be the impedance for a pair of 6L6GC's.

One of the things I had brain farts about was if the voltage I was applying compared to what I was reading on the Fluke was correctly correlated by me. Not having a scope to see exactly what was going in to the taps and what is coming out of the primary worried me, but, if figured if I'm using the same Fluke meter AC voltage setting at all times to meter input V and output V then I wouldn't have to worry if say, the function generator was telling me it was peak-to-peak but the meter was showing me RMS peak. that was just a mix up in my head but I did wonder if a misunderstanding of what the Fluke was showing me led to something that could have either doubled or halfed the voltage. The inputted voltage and output voltage on the primary were set and measured by the same meter on the same setting.

Ultimately, since this OT was spec'd out to be for 6L6's, I would want to connect what ever speaker I used to the next lowest tap impedance since this 20 watt plexi is using 6V6's which normally would want to see a 8K primary impedance. Putting say, an 8 ohm speaker on the 4 ohm tap ought to give me that since that would reflect to the primary a more suitable impedance for the outputs of the 6V6's.

Thanks again, and don't think me too pathetic in the way I spend (or waste) my time and energy. Damn, you should have seen me in the 70's! you want to see a waste of time and energy?
Thanks,
Best, stay healthy,
PJD3

[martin manning]

Looks like the Alchemist's power level change is accomplished by dropping all the voltages. In any case, 4 and 8Ω taps are available in the original circuit whether it's 4, 8, 16, or 2, 4, 8, so your calculation assuming a 4k primary is correct, and the OT is really 2, 4, 8. If you want to run a pair of 6V6 at 8k, just assume the secondary taps are 4, 8, and 16.

[nworbetan]

Since the transformer has no particular primary impedance it can be assumed to have either 2, 4, and 8Ω taps and a 4k primary impedance, or 4, 8, and 16Ω taps and an 8k primary. [edit, see below]

Yeah you can use a lot of different speaker loads and come up with a lot of different reflected loads on the primary side, I'll agree with that part of what you said 100%.

However I have enough time to split hairs today and I'm still going to call his original error an error. The known quantities at the start of the deciphering were the 4Ω and 8Ω connections. The schematic also shows an unused 16Ω tap. That leaves the reflected impedance as the one quantity that needs to be deciphered. Using an assumed 4kΩ value for the reflected impedance (remember one sentence ago when the reflected impedance was the one quantity that needs to be deciphered?) gave him speaker load that didn't match the already known quantities, and his confusion at that point was exactly the right response.

It's good to know how to use formulas forwards and backwards, and knowing when to use them forwards and when not to is even better.

[martin manning]

The Bogner Alchemist schematic calls out 3 taps of 4, 8 and 16 ohms on the OT secondary. Thats what I expected to see.
However, after applying a low voltage to each tap and measering across the brown and blue wire of the primary, and of course following through with the math, this is what I got:
Orange to Blk = 8.4
Yellow to Blk = 4.1
Green to Blk = 2.08

I did use a 1K resistor in series with the function generator output and measured the tap(s) input voltage at the taps, after the resistor. (I set it for a 1kH sine wave).
So this seemed to bring me to classic values except, the schematic shows a 4, 8 and an unused 16 ohm tap.
Ya think they just used a different OT with 2, 4 and 8 ohm taps? And I did out the math according to a pair of 6L6GC's with a plate voltage around 425 - primary impedance of 4.3K ohms

Nothing wrong there.

[nworbetan]

Nothing wrong except for the part where he plugged an assumed number into a formula where the answer was supposed to go and did the math and came up with different numbers than the known good values he already knew. Yes, he did the math right, but also no, he didn't do the right math.

[pjd3]

NWorbetan,

Can you explain what you meant by "plugged an assumed number into a formula where the answer was supposed to go".

The answers were to be the 3 tap impedances based upon what a 6L6GC data sheet would give for an AB push-pull pair at around 425-450 plate volts which is usually around 4K, maybe a little higher. I used 4.3K. That's the only number I assumed. In order to get the individual tap impedances in this instance, don't you need to divide the 4.3K by each individual impedance ratio?
I'm still not sure on where you saw my mistake.
It it helps any, I completely followed Uncle Dougs method for deciphering the tap impedance's. This method seemed to correlate with a number of others I read in articles and maybe one or two other videos. I still don't get where my boo boo was.

Thank you,
PJD3

[nworbetan]

NWorbetan,

Can you explain what you meant by "plugged an assumed number into a formula where the answer was supposed to go".

I've already kinda explained this once, but I'll rephrase it a bit and see if that helps any. I'm not a math teacher, so I'm doing the best I can here.

The one formula we're working with is:

P = S * T^2

P = primary impedance
S = secondary load
T = turns ratio

You started the thread with a transformer, and you knew that two of the taps were 4Ω and 8Ω, less than 3 hours later the schematic filled in the info that the third tap is 16Ω. So you already know what S is.

You learned how to measure the turns ratio, and you nailed it. So now T is taken care of.

The primary impedance is the answer you were looking for. You know S and T and needed to solve for P.

Unfortunately, what you did instead was plugged 4.3kΩ in to P and solved for S, which was the wrong direction to go.

And I did out the math according to a pair of 6L6GC's with a plate voltage around 425 - primary impedance of 4.3K ohms

Yep, probably operator error. Without seeing your math it looks like you got the turns ratio right, but then to find the reflected impedance you need to square the turns ratio and multiply it by the speaker load, like dorrisant explained.

You did your math correctly, which I gave you credit for right away, but you didn't do the correct math for the situation you were in. You knew the correct speaker loads, and you measured the turns ratio correctly. The only unknown left was the answer you could've and should've solved for. Instead you came up with this confusion because the numbers you expected to see didn't match the math that you did:

The Bogner Alchemist schematic calls out 3 taps of 4, 8 and 16 ohms on the OT secondary. Thats what I expected to see.
However, after applying a low voltage to each tap and measering across the brown and blue wire of the primary, and of course following through with the math, this is what I got:
Orange to Blk = 8.4
Yellow to Blk = 4.1
Green to Blk = 2.08

I did use a 1K resistor in series with the function generator output and measured the tap(s) input voltage at the taps, after the resistor. (I set it for a 1kH sine wave).
So this seemed to bring me to classic values except, the schematic shows a 4, 8 and an unused 16 ohm tap.
Ya think they just used a different OT with 2, 4 and 8 ohm taps? And I did out the math according to a pair of 6L6GC's with a plate voltage around 425 - primary impedance of 4.3K ohms
At first, I thought I thought the Fluke meter was maybe reading Peak ac voltage instead of P-P which the function generator clearly showed (P-P in big letters). But, I hooked the meter directly to the func-generator and set for one volt P-P, the meter read .708 volts which I had to take as an RMS p-p reading.

[trobbins]

There may be significant measurement errors creeping in as I think you are driving each section of the secondary winding with a low VAC. Not that it really matters, but some become concerned if they don't see a close match to say 4k or 8k.

One error could be your meter reading of VAC, as you are reporting eg. 0.5V, which depending on meter accuracy and rounding could be a voltage from say 0.4 to 0.6Vrms.

Another error may be the magnetising current involved in applying the 0.5Vrms. If possible, it is better to apply a voltage to one winding, but measure voltages from windings that are not 'driven'. One way to easily do that is to apply say 12VAC from a transformer to one half of the primary winding of a PP transformer, and measure the voltage on the other half primary winding, as well as the secondary side taps. Then double the measured primary voltage and use that voltage for your turns ratio and impedance calculations.

[Phil_S]

One error could be your meter reading of VAC, as you are reporting eg. 0.5V, which depending on meter accuracy and rounding could be a voltage from say 0.4 to 0.6Vrms.

FWIW, I agree and confess to being asleep at the wheel for not mentioning this. When I do this, I usually use a 6V filament supply as the source, usually around 7VAC without a meaningful load. This is 12-14x your input and will yield output in the hundreds, greatly reducing meter error. Perhaps the only reason I didn't bother was that the computations for all 3 taps were tightly clustered.

This got me to thinking, why are the input voltages different for each tap? And how on earth did you generate such small voltages, i.e., 0.15V?

[pjd3]

The small input voltages on the taps were present for two reasons. One, because the 1K "protection load resistor" I believe was causing a voltage division and Two, because I was using a function generator that only went up to 10 volts. I'm going to go back and approach this from one or two more different angles. Either apply an unloaded voltage straight to the primary and read the taps voltages, or just get a small low voltage AC transformer sent to the individual taps. It occurs to me now that there are a number of different approaches to skin this cat, a number of different angles to come in on it. I may as well try a couple more and see how results match up.

I'm still not totally convinced that a different OT was used than the one toted in the schematic. It wouldn't be the first time something like that has happened. Even the big Fender company has been shown to not always correlate component specs with schematics and drawings. I'm not hanging my hat on that but, that might not surprise me either. Plus, one of the reasons I got into building tube amps was to infact put myself through patches like this to gain insight and experience. And its working.

I know I've said this before but, I did this analysis modeled step by step to a very well done and clearly laid out Youtube vid that proved itself in the end to provide dead nuts accurate results. As far as process, I didn't add, subtract, switch or modify anything except the unit that was supplying the voltage. And my measurements even came out very close to the ones in the tutorial. But, I'm not done with this. It ain't over until its over.

Thank you,
Phil

[samrock55]

Blue plate Brown plate wires is there a certain polarity or doesn't it matter which output tube pin 3 they get soldered on? Its a typical push pull 6l6 40watt amplifier the schematic shows V5 is Blue pin 3 plate and V4 pin 3 is Brown. But that is backwards from amplifier I pulled Transformer out of? Any input is appreciated!