Dynamic output transformer testing (pushpull amp) question..

[Test Point]

Hey all...I'm not sure if this is the place to ask this but I'll give it a shot...
I need to build some sort of "dynamic" test bed to test output transformers used in guitar, etc amps (pushpull)...I fix amps for a living and sometimes I'll get an amp on the bench that the customer needs a quote for repair costs....unfortunately, some of these amps come to me in such a state that I can't tell if the O/T is working until I've spent several hours in the amp getting it to the point of testing it for power out, only to find out that the O/T is dead or not working properly and then have to call the owner with a "you now need an output tranny to the tune of $120 plus installation". You can imagine their response.
I'm thinking of using an old 100 watt Peavey amp I've had kicking around, pull the output tranny out of it and use binding posts to hook an unknown O/T up to it and run the outputs to a load. I would gut the amp except for the drive tube and the output tubes (pull two of them for a 50 watt O/T) and just pump a signal into the effects return jack.
I guess I'm asking if there is an easier way to do this than what I am dreaming up here? I know you can check a transformer using a sig gen output into the primaries and measure the output but I have found that that doesn't always show the whole picture...it'll show OK until you get about 350 or more volts on the primaries and then the insulation will break down or other maladies will show up....
What are your thoughts on this...am I trying to do something here that is easier to do another way???

I am beholdin' to you all

TP

BTW...I know that O/T's have different input impedances for different tubes, 6L6's have different output Z's than 6550's and El-34's etc but for the task at hand, I'm not sure that will make that much difference...it would probably affect damping and frequency response but that isn't what I'll be testing for and probably will be inconsequential to my test...I think.

 

[Gus]

http://www.ampage.org/ is a good place

R.G.'s site has a very good section on repairing tube amps and I think R.G. had a nice simple transformer tester PDF. www.geofex.com

 

[PRR]

If you want a dead/not-dead test:

First ohm-meter everything. Use an old-fashioned needle-meter, NOT a digital meter (which will be baffled by the high inductance). Primary windings will be several hundred ohms, speaker windings will be about an ohm, and neither winding will have continuity to the case (except a few super-cheap models where one side of the speaker may be terminated on the chassis via the transformer mounts). If the plate winding isn't few-hundred ohms, or has ANY continuity to case, the tranny is toast.

Get 240V AC from a transformer (NOT straight from the wall). Put 1K 10W in series with each output terminal. Apply that to the primary (each side of a push-pull). Load the speaker winding with 50 ohms 10 Watts. Put a scope on that. If you get 8 to 30 V of nearly-sine AC, the transformer is most likely OK. If the 1K resistor smokes, the transformer is shorted (or you connected to the secondary).

The output usually will not be perfect-sine, because the wall-power is usually imperfect. But you will soon know what your local line looks like. And most transformer defects will be gross.

If you really have a mule-amp around, with reasonable B+ (like 400V, not one of those Fender beasts that idled at 700V), sure: fuse the B+, put trustworthy terminals where the tranny goes, disconnect any feedback loop (you don't know the polarity of mystery transformers, so the feedback could end up negative OR positive; gitar amps can work without feedback), and play a few bars. The various impedances will give different power output and overload sound, but a real no-go transformer will be quite clear.

Please be careful. Hay-wired 400V lines can be fatal real quick.

 

[CJ]

just inject a signal into the grids of the power tube/tubes and listen to the output.
if you work on fender/marshalls all the time, you could make a jig that plugs into the phase inverter tube socket.
cj

 

[Test Point]

Well, I usually do the ohm meter test on all the O/T's when get one on the bench but in the case that I'm dealing with today, I got an amp on the bench that zorched the board that the output tubes were on...after grinding away all the charred parts of the PCB with a dremel tool and rewiring those runs that vaporized and then coating everything with a liberal coat of corona dope (and waited for it to dry)...I fired amp up and let it warm up...no problems until I tried to pull power from the amp...as soon as I got above 1/2 power or so, it would go into oscillations and start drawing tons of current. If I reduced the load to 32 ohms, it would start to behave somewhat....I clipped in another mongrel O/T I had in the shop and it worked perfectly...so up to about 50 watts it was OK but after that it would go haywire....how would an ohm meter check find something like that? It seemed to be breaking down as the load tried to take more power from the amp.
BTW PRR, I didn't know about the analog/digital meter discrepancy trying to ring out a tranny....I'll have to see if I can find my trusty Simpson 260 somewhere...I'll bet I gave it away :sad:

I'll check out the web pages that Gus passed my way and see if there is any info in there I can use....

TP

 

[Test Point]

[quote author="PRR"]Please be careful. Hay-wired 400V lines can be fatal real quick.[/quote]


PRR....thanks for the warning...I do get bit every once in a while (even after years (30!) of doing this stuff....it's usually due to my being tired or just plain stupidity. I do this stuff at my 9-5 and then come home and do it some more on my own bench...my 9-5 pays my health insurance and my 2-3 hours on my own bench afew nights a week keeps me in the local musician's circles...as well as a little "mad" money in my pocket. I just got done rebuilding an Ampeg SVT (the original)....you want to talk about plate voltage! almost 700volts.... the first set of 6550's I put in there toasted in the first 3 minutes. They just don't make tubes like they used to. When I get done with one of those, I feel like I just rebuilt a V8...and I look like it too...I'm almost too old to lift one of those onto the bench anymore :green:

TP

 

[NewYorkDave]

Amp goes haywire with oscillation and burns up the output transformer, huh?

Newer Marshall, perhaps?

Remove the 1.5k output tube grid stoppers from the PCB and replace them with jumpers. (Use "Zerohms" if you wanna be really slick). Then install new grid stoppers directly on the tube socket pins. You'll be pleasantly surprised.

 

[Test Point]

[quote author="NewYorkDave"]
Newer Marshall, perhaps?
[/quote]

Hi Dave...actually, no, it was a Peavey, late 70's 100 watt "Artist-VT", but it did/does have the 1.5K grid stoppers in there...I'll have to give it a shot on Monday....It's one of those Peavey amps that has a LTP driving the power tubes...I hate those things :evil:

Thanks,
TP

 

[rlaury]

Hi Group
Another good way to test is to, load each side of the primary to CT with the appropriate power resistors. This would equal the plate resistance of your proposed output tubes. Then using a BIG SS power amp with a low impedance output (I use an old Peavey CS400) drive the secondary slowly up to the rated power. Use Ohms Law, P=Esqu/R to determine the required drive voltage. Example: 20 volts would be 50 watts @ 8 ohms. Be very careful! The voltages on the secondary side are real and can pizz you off. This way you can perform freq response test, determine the actual primary impedance ect. The reason for the SS amp is, we need an amp with a very low output impendence < 1/10 for our signal source. Hope this helps.

RonL

http://www.nashaudio.com