Marshall PT current ratings

[Roffe]

I have a replacement power transformer for a 50W Marshall that I'm planning to use for a completely different project. It's rated 690VCT/150mA and 6.3V/5A, are these "conservative" ratings? The circuit will not draw more than 150mA B+, probably closer to 100mA, but I need 5.5A for the heaters. The wire gauge on the heater winding certainly looks up to the task, but will a 10% increase be safe? How much leeway do I have?

[Chuck H]

I have a replacement power transformer for a 50W Marshall

New one? Old one? I think the older transformers are more likely to just drop a little more voltage and get a little warmer with a small added current. I can't say about newer model transformers, though it's probably the same. If you're not going to be maxing the rating on other windings it's probable that the overall heat in the unit won't be threatening. I'd probably go for it if it were my project. But don't blame me if there's a problem Anytime you exceed ratings you takes yer chances.

But how are you drawing that much filament current? The way I count you'll have a maximum of 3A for the power tubes and preamp tubes only draw 300mA @ 6.3V. At 5A you have enough spec left over for 6.6 preamp tubes. Let's call it 7 'cause, ya know. So that would seem like plenty.

[Roffe]

I think it's the "new" style (JCM800-type) transformer. I'm building Pete Millett's low-mu preamp, so the heater winding will power a 6AS7/6080 tube that draws 2.5A, in addition to two EL34s. I think I'll just go for it and keep an eye on the temperature. I do have an extra 6.3V/3A filament transformer in case this doesn't work out, but I'd really like to save some space...

[The Dude]

If you're building anyway, just power it up with the Marshall PT and see what you have for filament voltage. If the filament winding drops way below 6.3V, you'll know it's being stressed too much.

[Chuck H]

If you're building anyway, just power it up with the Marshall PT and see what you have for filament voltage. If the filament winding drops way below 6.3V, you'll know it's being stressed too much.

Yep. If you were Helmholtz or Juan you'd probably even calculate the wire gauge using the winding resistance and the voltage drop figures Indeed The Dude has it. If the wire gauge is small and that 5A spec is a tight ceiling you'll probably get a filament voltage below spec. Otherwise I'd bet you're alright.

[The Dude]

Yep. If you were Helmholtz or Juan.......

Well, I'm not as smart as either one of those guys, so a shortcut it is.

[Enzo]

And a 5A rating means the thing can sit there making 5A all day, so I'd wager the extra 10% won't kill it.

[nevetslab]

And a 5A rating means the thing can sit there making 5A all day, so I'd wager the extra 10% won't kill it.

Of course, pulling 5A all day long will produce a significant temperature rise in the transformer. Over time, the heater voltage does drop until you hit thermal equilibrium. The insulation system in the transformer is probably 130 Deg C. It's not difficult to measure/calculate the temperature rise, if you have a DMM that can measure 10 mOhms. Better accuracy if you can measure lower than that. That was one of the standard tasks in product development when evaluating transformer samples, so we could qualify all of the mfgr's xfmr samples. Just in case, I've attached an article outlining the change-of-resistance temperature rise test procedure.

resistance-method-explained1.pdf

[Chuck H]

if you have a DMM that can measure 10 mOhms.

Even paying close attention to the null figure and set to the most appropriate range my Fluke probably won't steady state well enough for an accurate low mOhm reading.?. I think I've read that some old school tube VOM's will do it.

[Roffe]

Yep. If you were Helmholtz or Juan you'd probably even calculate the wire gauge using the winding resistance and the voltage drop figures Indeed The Dude has it. If the wire gauge is small and that 5A spec is a tight ceiling you'll probably get a filament voltage below spec. Otherwise I'd bet you're alright.

Even better, I measured the wire gauge I took off the end bells when I painted them and got access with my calipers. It’s about 1.65 sq.mm (around 15 awg). That should be enough for over 6A if I’m reading things right?

[Helmholtz]

I think it's the "new" style (JCM800-type) transformer.

Where do you get the 5V heater voltage for the tube rectifier?

Hammond replacement for 50W JCM800:
https://www.hammfg.com/files/parts/pdf/290GX.pdf
Note max. heater rating of 5.43A.

PT specified in the project schematic:
https://www.hammfg.com/files/parts/pdf/270FX.pdf
Note 5A heater rating.

[Chuck H]

Even better, I measured the wire gauge I took off the end bells when I painted them and got access with my calipers. It’s about 1.65 sq.mm (around 15 awg). That should be enough for over 6A if I’m reading things right?

Can't say. I guess it depends on the length because that will dictate the voltage drop. There are a lot of ways to design a transformer winding. The wire gauge must be big enough to handle the current, and then the length would dictate the voltage drop @ current. What I mean is that if a relatively small (but still suitable for the current rating) wire gauge is used, and there is a lot of it then the winding could be grossly over the voltage spec when run at lower than the spec current, and then grossly low on voltage spec when run @ higher than spec current. If a thicker wire is used and there is less of it the voltage variation between low and high current would be less. Different transformers can do it a little differently depending on things that transformer designers know about and I don't. Just noting that these variations exist. So, as mentioned, if the voltage drop @ higher current doesn't render your filaments below spec I think you're good to go. 15 AWG would seem ample. And as Enzo noted, for something like a filament circuit winding the 5A spec is such that it's expected to do this all day, every day for the life of the unit and provide the spec voltage @ that current. 10% more current shouldn't stress a component with that sort of stringency in it's design. Probably not even if it were all day every day. So it really just comes down to what voltage you get. And, of course, checking the temp of the unit under load is never a bad idea. This doesn't usually need to be a high accuracy measurement unless it does seem too hot. Then you might want to know the actual temperature more finite. That is, if it just feels "kinda hot" it's almost surely fine. It'll feel straight up hot before there's a real danger.

[Roffe]

Where do you get the 5V heater voltage for the tube rectifier?

I have a separate 5V filament transformer for the rectifier. I’m not building the power supply from the schematic, I have to drop a bit of voltage with this transformer so I’ve modeled a supply in PSUD with a 5R4 rectifier and a power resistor before the first filter cap. The resistor will also limit inrush current and keep the 5R4 happy.

[Helmholtz]

I have a separate 5V filament transformer for the rectifier.

A separate heater transformer may induce hum. It should be mounted at the outside/on top of a steel chassis and preferably have endbells and/or a copper flux band. Also its orientation may matter.

I think you can use your PT provided it runs cooler than in the original Marshall amp.

The main safety parameter (apart from hipot testing) for a PT is wire temperature. The method described by nevetslab gives an average temperature increase for the whole winding. But it's actually the hot spot temperature that matters. As inner layers will have a higher temperature some extra safety margin has to be added.
In my company we used specially prepared transformer samples with buried thermocouples for evaluation/certification.

[nevetslab]

But it's actually the hot spot temperature that matters. As inner layers will have a higher temperature some extra safety margin has to be added.
In my company we used specially prepared transformer samples with buried thermocouples for evaluation/certification.

We also used that same method, coupling them to temp to DCV converters to drive the multichannel chart recorder.