Output Transformer Help

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
GA-8T is two 6BQ5 (EL84) at 15 Watts. Proper load is 8kCT to whatever your speaker impedance is.

8k OTs of this size are used in "Marshall 18W", Fender DeLuxe (most two-6V6 Fenders).

You can easily find an OT for the Gibson GA15 which is a variant GA-8.

GA8T-O Gibson | GA-8 | Output Transformer GA-8T — single 8 Ohm tap $106.00
Gibson - Part 3
 
Last edited:

Gnobuddy

Member
2016-03-01 4:10 pm
How do you measure 8kct on an op trans?
If you have a signal generator (or other steady source of sine wave test tones), you can simply apply a small AC voltage at a reasonable frequency (say 300 Hz) to the primary, and measure the exact AC voltage across both primary and secondary with either an oscilloscope, or the AC volts range on your DMM. You should have an 8 ohm resistor wired across the secondary (speaker side) when doing this.

Many affordable DMMs are inaccurate at frequencies above 60 Hz, but the inaccuracy should be about the same for both primary and secondary measurements, so they will largely cancel out. Many guitar output transformers won't work well at 60 Hz, so we don't want to use 60 Hz for the measurement.

Once you have both primary and secondary voltages, divide the primary voltage by the secondary voltage to get the voltage ratio, which is the same as the turns ratio. For example, if you measured 1.05 Vpp to the primary, and 33 mV across the secondary, your turns ratio is (1.05 / 0.033), or 31.8:1. We might reasonably round that to 32:1.

Now square the turns ratio: 32 * 32 = 1024, so the square of the turns ratio is 1024:1.

Finally, multiply 1024 by the speaker impedance. Say you're using an 8 ohm speaker; 1024 * 8 gives you 8192, which is the nominal primary impedance.

That's in ohms, so it's the same thing as 8.192 kilo ohms. Our measurement is not super accurate, so round up, and you can see that what you have is a nominally 8k primary.

By the way: if you measure from the primary centre-tap to one end of this particular hypothetical transformer, instead of from one end to the other end, you'll find that you get about 2k in this example. Yup, if you measure half the winding, you get one-fourth the impedance. That's the way transformers work! :)


-Gnobuddy
 

wg_ski

Member
2007-10-10 5:21 pm
Almost any guitar or hi-fi amp transformer good for over 5 watts will stand 6 volts at 60 Hz on the secondary, especially on the 16 ohm tap. Feed it 6 volts off your heater supply winding (or any little 6V trafo) and measure the turns ratio. I’ve measured them by feeding the primary directly off the 120V, but that requires more care and a light bulb limiter (use a very low wattage bulb in case of a short).

If you have a Woofer Tester you can load it with 8 or 10 ohms and measure the primary (or half of it) directly. The resulting Z curve will also give you what you need to calculate both the high and low frequency roll-offs.
 

Gnobuddy

Member
2016-03-01 4:10 pm
...transformer...6 volts at 60 Hz on the secondary...
The problem with using a frequency below the transformers capability is that you'll measure an inaccurate turns ratio. Guitar transformers have no reason to have a frequency response flat down to 60 Hz, and few manufacturers are going to spend lots of extra money to achieve bass bandwidth they do not need.

For anyone planning to do the measurement the way wg_ski suggests, please be aware that this method creates dangerous amounts of voltage across the primary, nearly 200 volts for an 8k:8 ohm transformer. Enough to shock you badly, or worse.

If you must use 60 Hz (not recommended), it's much safer to feed the 6.3 V AC into the primary, rather than the secondary. Modern DMMs will have no trouble measuring the small (few hundred millivolts) AC voltage from the secondary.
...feeding the primary directly off the 120V
IMO that falls firmly into the category of "Extremely Bad Idea (TM)". Please don't do this. :whazzat:


-Gnobuddy
 

Printer2

Member
2010-04-02 6:34 pm
If you have a problem measuring the high voltage on a primary excited by a low voltage on the secondary, maybe you should not be working on tube amps. Just saying. (But it is a good thing our little buddy mentioned it :)). 60 Hz will give you a quick indication of the winding ratio, if you are off 10-20% it is not going to change much, you have an idea what the transformer is doing with what tube and what load.
 

Gnobuddy

Member
2016-03-01 4:10 pm
If you have a problem measuring the high voltage on a primary excited by a low voltage on the secondary, maybe you should not be working on tube amps.
But who even thinks about the possibility that the primary wires lying loose on the bench are about to shock the bejesus out of them when they connect that heater power to the secondary? I know it didn't occur to me, until I did the back-of-the envelope calculation, fortunately before I grabbed the two ends of the transformer winding and held them against my DMM probes with my bare fingers. :eek:

That was my point - if you know you're about to be dealing with 200 volts, sure, you can deal with it safely. But if you don't know, and don't take appropriate precautions, its a different story entirely.

That said, why do the dangerous version of the measurement, when its just as easy to do the safe version? Apply the test signal to the primary, and you have nothing to worry about.
....winding ratio, if you are off 10-20% it is not going to change much
Well, a +/- 20% error (which is only a 1.6 dB error in frequency response) in winding ratio creates a +/- 44% error in calculated primary inductance. With that much error, an 8k primary could measure anywhere from 5.6k to 11.5k.

So how will you know if you're dealing with an 8k OT, or a 6.6k OT? Or, for that matter, a 5k one, or a 10k one?

No disrespect meant to anyone, but honestly, I don't really see the point of making a measurement if it's going to be so inaccurate that the results are virtually useless. Why bother?


-Gnobuddy
 

PRR

Member
Paid Member
2003-06-12 7:04 pm
Maine USA
www.diyaudio.com
The problem with using a frequency below the transformers capability is that you'll measure an inaccurate turns ratio. ...

No; or not much. The main thing is that the winding sucks more current. The voltage ratio stays good far below the nominal bass cut. Until the DCR gets in the way, typically 5X-10X lower than nominal bass limit.
 

Printer2

Member
2010-04-02 6:34 pm
But who even thinks about the possibility that the primary wires lying loose on the bench are about to shock the bejesus out of them when they connect that heater power to the secondary? I know it didn't occur to me, until I did the back-of-the envelope calculation, fortunately before I grabbed the two ends of the transformer winding and held them against my DMM probes with my bare fingers. :eek:

That was my point - if you know you're about to be dealing with 200 volts, sure, you can deal with it safely. But if you don't know, and don't take appropriate precautions, its a different story entirely.

That said, why do the dangerous version of the measurement, when its just as easy to do the safe version? Apply the test signal to the primary, and you have nothing to worry about.

Well, a +/- 20% error (which is only a 1.6 dB error in frequency response) in winding ratio creates a +/- 44% error in calculated primary inductance. With that much error, an 8k primary could measure anywhere from 5.6k to 11.5k.

So how will you know if you're dealing with an 8k OT, or a 6.6k OT? Or, for that matter, a 5k one, or a 10k one?

No disrespect meant to anyone, but honestly, I don't really see the point of making a measurement if it's going to be so inaccurate that the results are virtually useless. Why bother?


-Gnobuddy


But that is why I said it was good that you mentioned it. As far as the error, I meant your final impedance calculation. I have seen a few graphs where the small signal response of a transformer is a lot wider than the large signal. As PRR mentioned, unless you are sucking current it does not drop off as soon. I just assumed that anyone with a transformer is trying to find out if it would work with what tube. Sometimes ballpark is good enough.
 

estev

Member
2017-03-12 3:43 pm
Thanks Everyone
I was able to determine that the transformers I have won't work for my amp so I’ll have to buy one.
I put 6v on the secondary and with the results on the primary divided by the 6v, squared that #
and then multiplied by speaker ohms.
This is fun. Thanks again.
Steve
 
Thanks Everyone
I was able to determine that the transformers I have won't work for my amp so I’ll have to buy one.
I put 6v on the secondary and with the results on the primary divided by the 6v, squared that #
and then multiplied by speaker ohms.

This is fun. Thanks again.
Steve
Ok .... and? :confused:

Please post actual numbers found.

Thanks.
 

estev

Member
2017-03-12 3:43 pm
Gibson Discoverer GA-8T requires 8000ohm effective load resistance.
I applied 6v to secondary and got 33v at the primary.
Divided 33v by 6v = 5.4 winding ratio
5.4 squared =29 impedance ratio
29 x 8ohm speaker = 233Ohm elr.
Way too low for my amp.
Hope I did this right
Thanks
 

Gnobuddy

Member
2016-03-01 4:10 pm
...5.4 winding ratio...
...233 Ohm...
Way too low for my amp.
Way too low for any tube guitar amp I can think of. You need a winding ratio of just under 32:1 to get an 8k primary impedance from an 8 ohm speaker. (And that gets you a primary voltage of 200 volts if you apply 6.3 VAC to the secondary.)

Perhaps the transformer you measured was a power transformer, and not an output transformer? The winding ratio you measured is about right for a 12-volt power transformer with a 240V AC primary, or a 24 V power transformer with a 120V VAC primary.


-Gnobuddy
 

wg_ski

Member
2007-10-10 5:21 pm
It could also be a 20 watt 70 volt line matching transformer. That would produce those results. And might even have a secondary tapped and labeled as 4-8-16 ohm. Does the primary appear to be center tapped? And is the “center” at half the voltage, or some other ratio?