Using an imbalanced PT.

[Alan0354]

Hi guys, I am starting a new thread in how to use the defective Weber PT that has imbalanced output. This is the data sheet of the transformer: https://taweber.powweb.com/store/025130intsch.jpg. But the two Red wires and Red/White wires are not balanced. They measured to be:


1) With Red/Yellow grounded:
a) Measure the two Red wires, readings are 320VAC and 366VAC resp.
b) Measure the two Red/White wires, readings VAC 250VAC and 294VAC resp.

2) With Blue/Red ( 45VAC) grounded:
a) Measure the two Red wires, readings are 366VAC and 320VAC resp.
b) Measure the two Red/White wires, readings are 294VAC and 250VAC resp.

Weber told me all the ones they have shows the same problem. They are going to sent me another model. But at the mean time, I am trying to make use of this defective PT as it's already in the amp, it's too much trouble to fit a new PT in and it will be a week before I receive the new one........AND the most important, I don't have a lot of confidence on the new one either. So I am going to try to make this original one work.

I am going to use the Red wires that measure 320VAC and 366VAC respectively. Questions are:

1) If I just use it as is, I should get about 420VDC after rectifier. Obviously the 366VAC side is going to carry much of the load. Does that present a reliability issue?

2) I can put zener diodes to lower the voltage. The difference is 46VAC. So the peak voltage is 46 X 1.4 = 64.4V peak. So if I use zener diodes in series to take up about 60V, I should be able to balance the two sides better.

What do you guys think?

Thanks

[Leo_Gnardo]

1) If I just use it as is, I should get about 420VDC after rectifier. Obviously the 366VAC side is going to carry much of the load. Does that present a reliability issue?

If you have some circuit where you can get away with using a seriously compromised transformer then go ahead knock yourself out. Derate HV current by 50% as you won't be getting any from the lame winding. Powers a preamp or maybe a very small champ-like amp.

I'd put this transformer into the recycle bin, or if I was forced at gunpoint to use it, cut the wires off real short, and it makes a dandy paperweight.

[Alan0354]

Thanks for the reply, how about the idea of using zener diodes to knock off the voltage on the high side? If I can get within 10V, that should help.

[g1]

Having zeners will not help the unequal sharing of the load.

[J M Fahey]

I am going to use the Red wires that measure 320VAC and 366VAC respectively. Questions are:

1) If I just use it as is, I should get about 420VDC after rectifier.

Sorry but no, the capacitor is going to charge to the highest peak voltage available, some 366*1,4142=517V .
Capacitor and tube killer.

That it will be a poor regulation supply and will drop much more than expected, is not the point.

Obviously the 366VAC side is going to carry much of the load. Does that present a reliability issue?

Yup, on 2 counts.

To begin with the higher voltage side will supply much more current than the other, because the voltage difference is huge.

Besides, this will unsymmetrically load the transformer core, which will work closer to a halfwave supply than to a full wave one.
Cores *hate* that saturating DC component.

Besides, I wonīt accuse this transformer of being bad or poorly made (besides this winding error) without further proof, but I guess itīs safe to assume that it was made to a price, so Iīd expect to meet specs but not have extra safety margin, so it would not stand misuse graciously .

Unavoidable misuse because of the winding error.

That said, that transformer *can* be used in a normal way, having a functional bias tap, etc. , but my quota of free consulting for Commercial Companies has been exceeded for the Year 2014, so Iīm sorry but Iīll have to keep it to myself.

2) I can put zener diodes to lower the voltage. The difference is 46VAC. So the peak voltage is 46 X 1.4 = 64.4V peak. So if I use zener diodes in series to take up about 60V, I should be able to balance the two sides better.
What do you guys think?

Yes you can and your Math is right.

Wasteful but if thereīs no option ....

The diodes would dissipate on average 60*0.15*0.5=4.5W (because they would work 50% of the time), but since the mass is small and we canīt *really* count on it for thermal averaging, plus we will have start-up current peaks, etc. , Iīd use at least 5W *each* Zeners.

Cost is nil.

Or to play it even safer, Iīd use 4 x 15V 5W Zeners in series with that transformer leg.

[jazbo8]

Forget about the CT, just use a bridge rectifier.

[Alan0354]

Sorry but no, the capacitor is going to charge to the highest peak voltage available, some 366*1,4142=517V .
Capacitor and tube killer.

That it will be a poor regulation supply and will drop much more than expected, is not the point.

Yup, on 2 counts.

To begin with the higher voltage side will supply much more current than the other, because the voltage difference is huge.

Besides, this will unsymmetrically load the transformer core, which will work closer to a halfwave supply than to a full wave one.
Cores *hate* that saturating DC component.

Besides, I wonīt accuse this transformer of being bad or poorly made (besides this winding error) without further proof, but I guess itīs safe to assume that it was made to a price, so Iīd expect to meet specs but not have extra safety margin, so it would not stand misuse graciously .

Unavoidable misuse because of the winding error.

That said, that transformer *can* be used in a normal way, having a functional bias tap, etc. , but my quota of free consulting for Commercial Companies has been exceeded for the Year 2014, so Iīm sorry but Iīll have to keep it to myself.


Yes you can and your Math is right.

Wasteful but if thereīs no option ....

The diodes would dissipate on average 60*0.15*0.5=4.5W (because they would work 50% of the time), but since the mass is small and we canīt *really* count on it for thermal averaging, plus we will have start-up current peaks, etc. , Iīd use at least 5W *each* Zeners.

Cost is nil.

Or to play it even safer, Iīd use 4 x 15V 5W Zeners in series with that transformer leg.

Thanks for the answer.

1) It is funny, I have been using the Red/White wires which is 250VAC and 294VAC resp. At no load the voltage is about 400VDC. But with two 6V6, it is down to 330VDC. I don't even have to strum the guitar, just no signal idle voltage. I just hooked up the Red wires and measure 419VDC at B+. Seems like it follow the lower voltage tap.

2) Is you secret way to balance out by using some sort of voltage divider between the CT(Red/Yellow) and bias taps(Blue/Red), then pick out the half way point as the new CT? The divider can be resistor or better yet, using two inductors and take the mid point. I was thinking about this, but I need the full bias voltage of 45VAC. Well, if I change to cathode bias, then I can do this.

3) I am thinking about using this circuit with zener. Yes, 4 of the 15V 5W zener will be perfect and they are cheap. This is my proposed circuit, Not only I have zener, I use a 250 ohm resistor on each side to help balance the current. I am using a 135ohm resistor to simulate the sag of a tube rectifier, so I just split the resistor and move the resistors towards the rectifiers as shown.

[g1]

You can add all kinds of zeners and resistors but it will not help the imbalanced loading or saturation problems.

Forget about the CT, just use a bridge rectifier.

This is THE solution. The transformer is designed for use ONLY with bridge rectifier. It is mis-labeled, stop thinking of it as a center-tap, it does not have one and was not designed with one. It will work perfectly for anywhere a center-tap is not needed/used.
Use the diagram Enzo posted in the other thread, that is what you have, they have mis-labeled the 370V tap as a center-tap, but it is not one.


Edit: Of course this begs the question, why would you put a 5V tap in a PT designed/sold for bridge only use?

[Enzo]

I have seen nothing to dissuade me of this view. Ever since I first saw these transformers, I have been convinced the description given the winder was bad, and he simply made exactly what was requested, rather than figuring out that something different was the real need. Look at some of the other Weber PTs. A number of them have nicely printed voltage numbers and stuff, and there, clearly edited in, is a "CT" by one lead that was not there before.


I think the winder was not told to make "a power transformer for a tube amp." I suspect he was told to make a transformer with a 5v winding, a 6v winding, and high voltage windings per the drawings. I suspect it was done as a drawing rather than words. And I think Weber made an assumption that the winder did not. And I think Weber didn't/doesn't want to eat all these things or at least sit on them until some usefull application is found.

[Alan0354]

Maybe I should wait for the new one to come in, it's supposed to be the same size and has no problem. I just try to savage the two I have as I don't think they are asking these back.

[jazbo8]

Maybe I should wait for the new one to come in, it's supposed to be the same size and has no problem. I just try to savage the two I have as I don't think they are asking these back.

Hey that's a good way to score some irons, let's place a bulk order then ask for replacements if Weber is willing to let us keep the "defects".

[J M Fahey]

I think that there was not a detailed spec asked for, just some voltages and currents specified, and the winder who probably is a large OEM supplier didnīt even know (nor care) that this was for a tube amp , even less that it was for a tube rectified full wave supply.

I bet each winding supplies the voltage and current specified, so he fulfilled the contract.

That such transformer is unsuitable for that specific task is beyond his duty.

As a transformer winder I can also imagine some choices he made to simplify and speed up production, quite reasonable, but which in this case backfired.

And no, I donīt mean correcting voltage dividers , even less resistive ones.

If used with a bridge rectifier, downspec it to no more than, say, 100mA.

Reason is in 2 diode full wave rectifiers wire is specified to carry rated current only 50% of the time vs. bridge rectified 100% , a big difference.

If it had been designed for bridge rectification (such as other Webers) the max end to end voltage wouldnīt allow for around 700V RMS (almost 1 KV DC) .

[Alan0354]

Update, I build according to my schematic I posted in post #7, I use a 24V and a 36V zener diodes to test. I measured across the two 250 ohm resistor. I got about 13V on both, the current split very nicely between the two legs!!!! I got 370VDC and the center of the OT which is perfect. I might try 150 ohm so I drop less voltage next. If this is a production amp, I would not do it like this, but this is going to be my amp only, as long as I got it balance, there is no reason it is not reliable and I don't think I need to de-rate the current at all.

The two power tube is not matching that well, at one adjusted to 25mA, the other is only 21.5mA or so. Question to you guys is, Is this too much.

When I crank the high gain channel all the way, I do hear a little hum. So I think there is still a slight imbalance. But I think it's acceptable.

I am going to buy the 15V 5W zener diode tomorrow.

I think I can get onto the Marshall type soon. This current one is a channel switching with power scaling, more emulate Fender, the voicing are more of crank Fender with WGS G12CS all American sounding speaker. I am going to get on with gutting the Marshall JCM900 Dual Reverb combo, this time, I am going to do similar channel switching amp, but more Marshall voicing. I am going to get the 50W Marshall type OT and use the PT in the amp. Then likely getting the WGS ET65 speaker to get more Marshall sound. Any suggestion will be appreciated.

[Alan0354]

Hey that's a good way to score some irons, let's place a bulk order then ask for replacements if Weber is willing to let us keep the "defects".

Well, I don't intend to take advantage of people like this. It is what it is right now. If they want the second one back, they have to pay for the return shipment, I ain't going to pay for return shipment. I doubt though, what are they going to do with that, they can't exactly resell it.

That's the reason I want to try making the defective PT work. Looks like I got it working as in the last post.

[J M Fahey]

Good to know.

I don't think I need to de-rate the current at all.

Nice statement but you don't give any reason for it

Don't worry, in *this* case it will work.

What I said earlier is that this is a transformer designed for 2 diode full wave rectification ...which was miswound or miswired at the Factory

So if you use it with a 2 diode full wave rectifier, each HV winding supplies 150 mA 50% of the time and no derating is necessary.

Now if you use just *part* of the windings present there, the unused ones will rest idle ... but unfortunately still using window space, and if you use a section with a bridge rectifier it will be working 100% of the time, which was not expected, so it will heat up beyond normal, that's why in that case it needs derating.

Otherwise that winding will become a hot spot relative to the rest.

Imagine as a comparison using less HV but more filamenbt current, the total power draw will not be exceeded, the mains fuse will not blow, but the filament winding may smoke and burn by itself.
Same thing.

Really can't comment or imagine what will Weber do with their stock of miswound transformers.

As I see it, the supplier :

1) wound 2 tapped symmetrical HV windings , each providing 250V and 320VAC (as measured with Alan's unknown wall voltage)
So far so good and what Weber specified, and what's needed for 2 diode full wave rectification.
And both tapped windings with 150mA capable wire, which means the same wire was used throughout.

2) wound a 44V bias winding.
He *should* have used thinner wire, bias current needs are minimal, but instead he still used the same 150mA capable wire.

Why?

Thinking as a hands-on winder: it's an annoyance/p.i.t.a. to stop the machine, replace the spool with a thinner wire one, reset the lead screw or advance mechanism to a finer thread or pitch , just to wind a few extra turns ... and then resetting everything again for the thicker wire so a logical compromise is to go on with the thicker one already loaded.
So far so good, he's making a better than asked for winding

3) the Problem:

a) he should have joined the start of the symmetrical HV windings and call it center tap.

b) he should have joined to that same center tap one end of the bias winding , and sent the free end outside with an extra wire, called "bias".
And we'd all be happy.

4) instead, he connected one tapped HV winding to one end of the bias one and to the red/yellow wire; then the other HV tapped winding to the other end of bias and to blue/red , effectively killing symmetry and making transformer unusable.

The "magical solution" would be to short together R/Y and B/R wires to recreate the original (missing) center tap .... except that you would be shorting the bias winding and magnetically all transformer windings are in parallel so it's a no no.