Output transformer question

[Phil_S]

I've got an unknown output transformer that I've been trying to figure out, and something isn't adding up. It is a physically small item, maybe 1.5" x 1.5" x 2". It is installed in a chassis that's been cut. There are two noval sockets. It is a PP tranny with secondaries wired to screw terminals for 4-8-16. The relative ohms and turns ratios confirm the 4-8-16 relationship for the secondary taps, but in solving for assumed plate resistance (turns^2 * 2ndary ohms) I'm getting numbers in the range of ~2400 to ~2700.

Maybe I'm off base here, but I was expecting something either in the 6K-10K range or maybe 22K. I have no idea what tubes were in those sockets and wiring was stripped long ago.

There is also a power transformer in the chassis, also physically small that puts out about 288-0-288 VAC without a load, and has what appears to be a 6.3v winding that puts out in excess of 7VAC without a load. This is marked Viking 103-6A; 98 6002. Metering 1K ohms across the primary, I expect a significant drop under load and that would be OK.

I was hoping these would make a good pair for a small amp. I'm thinking something along the lines of a couple of 6SL7's (tweed-ish?) and a self split 6SN7 for the output. For the 6SN7, I'm thinking, that OT isn't right; that I probably need one that's 22K primary. Is there a real downside to trying this?

I'm looking for a sounding board here. Any thoughts would be appreciated.

[Andy Le Blanc]

that does seem off........ what voltage ratios do you get if you apply 1V ac.
to the sec. taps of the OPT.......

[Phil_S]

Well, Andy, your question was enough to prompt me to recheck what I'd done. The first time, I ran ~5v through the primary and checked the secondary. The output, at under 1v, was not reliable on my meter.

This time, I checked by running the input to the secondary. Turns ratios are 25, 33, and 49. The secondary taps are tied to a terminal strip labeled G-4-8-16. This says the primary is 9904 (16), 8787 (eight), and 9786 (4). I'll write off the 8z tap as the outlyer and call this one about 10K.

If, for triodes, R(load) = 2 * R(plate), then it is good for use with a 5K plate. Probably close enough for 6SN7 considering that it was probably wired for 12AU7. At this low wattage, there is probably a wide envelope for error and I'll probably give it a try.

So, sorry for the false alarm and thanks for the help.

[Andy Le Blanc]

Im glad it checks out.... sometimes its the good question that get the good answer....
the question I have about OPT testing is at what frequency
Ive seen plenty of refferences at 60Hz...... but they pretain to PT.....
I guess the right thing to do would be to reffer to the freq. as well....
I havent located a good refference that presents results of a transformer
tested in this manner (1V ac. applied to sec.) with frequency as a focus
it might be valueable to know....

[skyboltone]

Im glad it checks out.... sometimes its the good question that get the good answer....
the question I have about OPT testing is at what frequency
Ive seen plenty of refferences at 60Hz...... but they pretain to PT.....
I guess the right thing to do would be to reffer to the freq. as well....
I havent located a good refference that presents results of a transformer
tested in this manner (1V ac. applied to sec.) with frequency as a focus
it might be valueable to know....

I've noticed that the higher the frequency the higher the turns ratio, the higher the primary impedence. The difference between 60hz and 1000hz though is on the order of a few hundred ohms at the voltages I've been able to apply. My old HP 200C won't put out more than a couple of volts into that kind of load.

[Aurora]

Using any kind of small ( or larger) audio amp should fix the loading problem. Just make sure you measure the voltage directly at the secondary winding being used.

[skyboltone]

Using any kind of small ( or larger) audio amp should fix the loading problem. Just make sure you measure the voltage directly at the secondary winding being used.

Ding! Thanks for that. Duhhhh. I guess really even just a little transistor gizmo. Just didn't occur to me.

Just goes to show that you always think better north of the 65th parallel. At least I always have, haven't I?

[Aurora]

Just goes to show that you always think better north of the 65th parallel. At least I always have, haven't I?

I'll most certainly make that a permanent quote!!

[Andy Le Blanc]

I've got a HP 202c...... I end up tweaking the freq. to get the desired V.....
so its good to know what the error amounts to.......
using another amp is a good idea...... Id mind the output impeadance....

[Phil_S]

Im glad it checks out.... sometimes its the good question that get the good answer....
the question I have about OPT testing is at what frequency
Ive seen plenty of refferences at 60Hz...... but they pretain to PT.....
I guess the right thing to do would be to reffer to the freq. as well....
I havent located a good refference that presents results of a transformer
tested in this manner (1V ac. applied to sec.) with frequency as a focus
it might be valueable to know....

I don't grasp this fully. The material I've been using http://www.radioremembered.org/outimp.htm suggests that you test the thing and do some math. I've found that is seems to work for most of the trannies I've worked with. He does say to test it with one volt, but I don't have a one volt source. I've been using a 5v winding from a PT.

I focused on this, "Now, armed with the turns ratio, we can calculate the impedance ratio and the impedance that will be reflected to the primary with a given load in the secondary. Remember we said earlier that the impedance ratio is the square of the turns ratio...." I don't see how this would change with input voltage.

This source goes on to say, "The impedance load seen by the tube and output transformer is not constant. The frequency of the audio signal will vary over a wide range. The inductance in the windings will have a different impedance at different frequencies. At a certain frequency an 8 ohm voice coil may have an impedance of 10 ohms or at low frequencies it my have an impedance of 4 ohms. This varying load impedance is reflected back to the primary, so the tube, and output transformer must work into a varying impedance range."

[Aurora]

The latter part of your post refers to a loudspeaker as a load for your amp. A typical 8 ohms full range or guitar speaker, will be 8 ohms or slightly less around 1 kHz, which is usually the reference frequency for loudspeaker data. Since the speaker is driven by a coil,( the voice coil), its impedance will rise with increasing frequency, 10-20 ohms for the upper freq. range is quite normal. At lower frq's the speaker will have an electromechanical resonance point with an impedance peak, often up to several 10s of ohms. This is also depending of the cab in use. A reflex cab will usually have at least two peaks, one for the speaker/ air mass and another for the reflex port. This will make the actual load at the tubes vary, sometimes below, but mostly above the design impedance.

As for the transformer impedance ratio, this is certainly turns ratio squared, as said. WIth respect to the above section, this implies constant loads. Are you measuring the voltage directly at your injection point, or do you rely on the 5V spec alone? The 8 ohms winding may load the power transformer, unless it's a rather hefty one. OTOH, if the power transformer you use IS a hefty one, the 8 ohms tap may present a too smal load, thus giving more than 5 V. I'd suggest using a 20-30 ohms series resistor, and measure directly at the injection point, - use this voltage for your calc's.

[Phil_S]

Hmmm.....the PT is a really big one, probably from an organ. It's too large for a guitar amp, at least one you expect to move around. I think the 6.3v winding is rated at 9A (yes, nine). I can't remember what the 5V winding is rated at, but certainly it's no wimp, either.

I typically hook the 5v supply to the secondary and see what come out of the primary. I check to see what the 5V supply is while under the load of the OT. I hadn't thought to load up more than that -- won't that muck up the calculation?

It depends on line voltage of the day and what load the OT presents. Typical readings are 5.25v to 5.35v. I use alligator clips from the source PT and measure voltage where the clips meet the secondary wires.

On more than one "known" OT, the numbers seemed to work out fine. On this particular one, after mending my ways, they made sense. I just did another this morning that was driving one SE EL84. Turns were about 32.5:1 and I came up with 4500:4 ohms, which also seems to make sense.

Do I really need to do something different?

[Aurora]

This particular transformer was somewhat smallish...?
You might be saturating the core...........try the resistor trick, and see what comes out of it..
Otherwise, -I'm out of ideas..........

[Andy Le Blanc]

testing with one volt makes the math easy..... gives a direct ratio......
any voltage will do...... but if theres a high turns ratio you can get a high
voltage as a result.... ie.... if you apply 1v to the secondary of an output
transformer with a turns ratio of 35:1 .... the result will be around 35v...
but if you apply 5v the result will be around 175v which begins to be a
safety concern.....

[Phil_S]

Andy: I agree that 175v is a concern, but figured for the amount of time it takes to read the voltage, nothing bad would happen. I've pushed this through several OT's without incident. Worst case is that I'll let the smoke out

Aurora: I think I'm OK with it.

Thanks to both of you for your commentary.

--Phil