What actually is a 100v transformer doing?

I

ianthesecond

Active Member
I recently bought an output transformer on eBay.
It was labelled as a Parmeko output transformer, and it is but....

All the amps I've owned and built have had an output transformer which matches the valves to a speaker by matching the large impedance load to a small one. For example 5k to 8 ohm.

This one has two output sections, a 100v and a 27.5v. I'm not sure if these are separate windings or even if they're completely separated from the primary. I have no idea what's going on inside this item.
It is labelled as Parmeko Atlantic Series / Model No. 9000/73

I have a fairly good understanding of output transformers, their function, their construction, the materials and the formulae to match the primary load to the secondary load, but please could someone explain what is going on inside this transformer?

It's obviously a real quality item but I'd like to know what application it was designed for and there is no information that I can find anywhere. I'll post some photos later but for now i'll describe the important aspects. It has a primary like a typical pentode push pull circuit's output; in, out, a centre tap and two other taps, five in all.
The secondaries are just labelled "in (27.5v) out"/"in (100v) out", four connections..

This transformer is very heavy, and I was planning on sourcing a large ultralinear output transformer for a 50watt push pull amp I have in the pipeline. If i can work out a way of connecting speakers to this then it might do the job. I don't know, any information about what current/power this could handle, or what exactly is going on inside (it's sealed) would be very useful.
 
Register to hide this ad
Sounds like a constant voltage type transformer for commercial audio distribution / PA. You would then use an additional line matching transformer at the speaker end. 100V is the typical constant-voltage in Europe for this application, IIRC. (70V is typical in North America.) These high voltage speaker systems are used to support long runs with reduced current requirements while also allowing for multiple speakers (with their own line matching transformers) to be tapped anywhere along the line without affecting the overall impedance seen by the amplifier.

You could certainly measure the winding ratios... but those constant-voltage transformers tend to be geared for PA, not HiFi.
 
The 25V secondary on a big line output transformer might be acceptable for direct speaker connection. And some line OPTs have good frequency response, especially when used well below max power rating. Go ahead and measure AC voltage ratios for the windings you have, and let's see what's possible.
 
You'd have to ohm it out, possible the 100V tap is just the far end of the secondary, or it could be it's own winding. As said, sounds like PA or sound distribution gear. Some was good, some was garbage. My best pair of amps started life as Bogen sound distribution amps, and they are excellent performers with a different circuit wired to the original Bogen iron.
 
That's a photograph of the transformer in question, with a pencil in the shot for scale. I made a quick measurement of the coils' resistance and they came out as follows:

I've called the terminals :
1 (Pri. in)
2 (Pri. second tapping)
3 (Pri. CT)
4 (Pri. fourth tapping)
5 (Pri. out)

1a (100v Sec. in)
2a (100v Sec. out)

1b (27.5v Sec. in)
2b (27.5v Sec. out)


1 to 5 = 150 ohms
1 to 3/3 to 5 = 75 ohms
1 to 2/4 to 5 = 45 ohms
2 to 3/3 to 4 = 30 ohms

This looks like a primary with a centre tap and secondary terminals at 30%.

The secondary shows:

1a to 2a = 6 ohms

1b to 2b = 1.2 ohms.

All three windings are separate.

This looks like it would step my signal voltage right down and would work with a loudspeaker directly connected. I don't have an easy way to measure the AC voltage ratios, I guess I'll need to construct some kind of low voltage AC signal generator.
 

Attachments

  • 018.JPG
    018.JPG
    85 KB · Views: 36
ianthesecond said:
... I don't have an easy way to measure the AC voltage ratios, I guess I'll need to construct some kind of low voltage AC signal generator.
Click to expand...

I believe a lot of people just use a small stepdown transformer e.g. 12V, 24V, whatever, to apply 50/60Hz AC and measure the resulting input and output voltages. It's not a wideband measurement but I think it's generally close enough. This is what I've done.

Then, when you have the voltage ratios you can figure the impedance ratios at various loads.
 
Results may not be super accurate if driving the secondary at a known voltage and reading voltage produced at the primary. A more accurate method is to drive the primary (the side with the larger impedance) with a known AC voltage and measure the AC voltage on the secondary, and in that case I'd try to find an AC source that delivers somewhere in the vicinity of 30 to 50 VAC--but lower voltage could also work, depending on the accuracy of your meter.

Now if you had a variac (variable autoformer) that would be the ticket..you could drive the variac from your mains house current, turn down the variac to something suitable, say 50V AC, and you're off to the races.

I just did this recently to validate a transformer I purchased. I drove the primary from my variac, and adjusted the input voltage until I read 2.82 VAC on the secondary. In my case I needed 81VAC input to deliver 2.82VAC output.
 
I drive the primary through my variac and an isolation transformer. Usually I adjust the input to get about 1 volt on the output, then measure both input and output to work out the ratios. I tossed together an Excel spreadsheet to do the math for me. It crunches the voltage in and voltage out to come up with a ratio, and it will take that times whatever speaker side impedance I plug in to spit out a primary impedance.

The actual voltages you use aren't super important but you do want a decently accurate voltmeter for this. The numbers square, so meter problems will skew your impedance calculations quite a bit.
 
Right, voltages can be arbitrary, as long as you get decent readings from your meter. The only reason I set it to 2.82 VAC on the secondary was because that would produce one watt output on the 8 ohm tap, if the 8 ohm tap were loaded with an 8 ohm resistor.
 
kward said:
Results may not be super accurate if driving the secondary at a known voltage and reading voltage produced at the primary. A more accurate method is to drive the primary (the side with the larger impedance)
Click to expand...

Agreed. I should have been more specific to say apply the AC to the OPT primary.
 
OK. I found the filament transformer and it actually has two 6v secondaries. I connected these in series and measured the voltage which actually came out at 14v.

I applied this to the primary of my Parmeko transformer and the results are as follows:

100v secondary gave 2.6v
27.5v secondary gave 0.6v.

The 27.5v winding looks more useful so I continued with:

14/0.6=23.3 (turns ratio)
23.3^2 = 544 (impedance ratio)

This looks like approximately 5 ohm load to a 2a3, unless I'm very much mistaken (notwithstanding the issues with this being a push pull transformer and not having an air gap)

Everything I've read online says don't use 100v line stuff with valves, but this is from 1961 and looks nothing like the usual 100v OPTs I've seen, it probably came from a valve circuit.

I'm going to connect my single ended 2a3 circuit to it with a 8ohm speaker and see what happens. I suspect it'll sound ok because this is a pretty massive chunk of metal and the issues with heat and saturation might not be a big problem. Also the amp only runs 60ma idle current and puts out a pretty tiny couple of watts.
Plus i like experimenting with this stuff , so I'll be back again this time tomorrow with either good news or bad.
 
Last edited:
For confirmation, I figure your 8R load will look like 4.36K on the primary. You connected the 14V source across the entire primary winding, right? But putting 60mADC through that winding will bring it pretty close to hard saturation. Hope you don't permanently magnetize the core.
 
An impedance ratio of 544 would multiply by your desired speaker impedance. For an 8 ohm load, the tubes would see 4352 ohms.

Push-pull trafos don't like to run single ended. They saturate very quickly. Its not about the power level, its purely the unbalanced current. A resistor off the other plate lead to ground to create some current on the other leg will let it work but your power supply will have to provide double the current it does now.
 
OK, well I won't connect it to the 2a3. Thanks for the heads up.

I have a push pull 6v6 amp which needs an output transformer. I don't remember what plate current it uses but maybe this transformer will be better suited to that. I assume 20 watts is less than what it was designed for and if that had an 8k/8 ohm transformer then this should remain kinda balanced with 15 or 16 ohm speakers.
 
Was the 14V applied across your terminals 1 to 5 , and did you measure 14V when connected ?

If so then a 100Vrms output would require a 540Vrms PP input swing from the amps output stage. That's 760Vpp, or 380Vpk from CT to a fully on plate for a PP stage. So the OT would ahve been designed for a valve output stage with a B+ of about 450V.

Given it's a sizable OT, perhaps 25-50W guesstimate, if you had a 6L6GC PP stage and wanted it to operate nominally around a 5kohm PP loading, then the 27V output would be ok to match with an 8 or 16 ohm speaker.

What's the model number - its a bit hard to make out in the photo.

Sometime you can get a better accuracy turns ratio measurement if you apply the 14V from one primary end to the CT, but measure the primary voltage from the other end to CT, and also the outputs. Your output voltage is getting pretty low (0.6V +/- 0.1V), so this technique will help a little. Also suggest measuring the other primary tappings, as the transformer is likely designed for UL.
 
Ah no, it was 14v unconnected. I did think about that actually, but I forgot to check that after connecting them to the primary. I'll do that again today and obtain a more accurate reading.

The model number is 9000/73, the best Google can do is provide a phone number for Parmeko which won't work.

I'll make some more measurements today. Thanks everybody.
 
Yeah I know! Did you read the whole thing? Pretty amazing story with an amazing ending. I read that the day I placed my bid on ebay but that was the only thing i managed to find.
I have the choke to match this OPT. The choke is 5H 250mA and it's currently doing its thing in my Loftin-White although it was rusty when I got it so its black now and not blue Hammerite. I never really thought I'd get any use out of this OPT but now it's looking a bit more hopeful.
 
You must log in or register to reply here.
Back
Top Bottom