Dave, I don’t think he needs that, from his question.

Perhaps that unit should have a big transformer removed or supported somehow due to the likely hood of it being on its side. Or perhaps some parts of the front should not have any pressure on them. I’m just shooting from the hip, but have seen unfortunate results. Now? to those knowing the PS-150-120

?

Trent, those do look like old pin tips ,but seem to have better restraint for the wires. I have no Idea about original sizes, but if it works….. don

?

Jim, Halden ?too Rick also, and where is Randy the orig owner of THAT transformer?

My stuff in Blue, this is now late.

The voltage drop across a forward biased 5R4 is anywhere from 25 to 60 volts depending on how much current is flowing through the rectifier.? A 25 volt drop results in 97% of the B++ voltage being applied to the 5 volt heater winding while a 60 volt drop results in 93% of the B++ voltage being applied to the heater winding.? To simplify the spitball analysis, I considered both to be a zero volt drop across the forward biased 5R4.

yes ..for the ongoing transformer failure discussion, the voltage drop only complicates things unnecessarily, so ignore it or call it nothing…??

?

Per the schematic, when pin 6 of the 5R4 is forward biased the 5 volt winding is at or near +800 DC equivalent volts.? At the same time, pin 4 is at -1130 peak volts.? Stated again, right above this section of the 800 volt winding sits the 5 volt heater winding which is at or near +800 DC equivalent volts but the reversed biased section of the 5R4 pin 4 is just below it and is at -1130 peak volts.? What do you think the differential voltage is between the 5 volt winding and the 800 volt winding just beneath it?? I think it is just under 2kv, what do you think?

Yes, As before, it is about the end to end voltage of the transformer HV. {I sometimes called this “the very HV”}

?

To complicate matters even more, what is Faraday's second law of induction?? This is why the SSR is not a good choice to control current through a transformer or a motor for that matter.? The zero volt switch causes the maximum current through an inductor every time the switch closes so the current instantaneously rises from zero to maximum.? So the induced voltage is even higher then that of a time varying current.

I’d rather not get into SSR’s until the above gets settled in the minds of others.. or not, ??but as to the response of the transformer output, on a zero volt start, it will draw inrush current and no doubt saturate, and draw high lopsided current, but dv/dt is no more than for switching at the crest, so the output voltage still reflects the input IMO. But I would rather sit on the fence about the overall use of SSRs vs mechanical switches have been tallied. With details too … ?

?

?

?

Rick ,Jim? too

From “after ?I get this”, ???my st…uff in green

?

But with a "zero volt" solid state switch HOW does the current rise instantaneously when the voltage does not? (and of course, because of back-emf, it cannot ever rise "instananeously") ya? "instananeously")??? ???[not THAT fast, maybe quickly?]?????

?

If a mechanical or electromechanical switch were to be switched on every time (during power-up)? at "zero" in the AC sine wave, the voltage would start at "zero" and rise at the rate in the actual sine wave to it's peak. And depending on the inductance etc.....the current would follow. Ya,? but not so simple see below. ?

?

There is no current until the thing is turned on. Ya, that must be faraday’s unspoken zeroth law!

?

If the switch were to be "thrown" AT the actual peak. ......Voltage applied at the peak,? the resulting? current rise [ rate of rise? ] would be greater... significantly? greater than if it was allowed to rise at the 'normal" sine-wave rate. Ya,? and/but soon it[the rate of rise]? would decline to nothing [at 90 degrees] leaving the flux at max normal and then reverse the current and start lowering the flux, thereby ...---- too tough to write! See below ?

So what am I missing here???

Rick,Maybe this helps you, even if it is hard to understand:

pdf? 4 ---

The worse inrush happens when the AC waveform is crossing zero. At this moment, the

voltage levels are extremely low, requiring a lot of current to energize the core. Here is a picture of

what inrush looks like over time:

^this must be a sizable transformer with a high X/R so the high currents are slow to decrease, at 1kva my guess is the inrush is really the first cycle, and by the 3^rd it looks normal note the “DC offset”?????????

and

??

d. Switching at maximum voltage crossing- no residual magnetism

This scenario will have transformer switched on during maximum voltage with no residual magnetism.

Figure 6: Waveform for case ‘d’

This case is an ideal situation as there will not be any inrush current theoretically. Remember that flux ‘B’ in the core is always 90⁰ out of phase with system voltage. Since the switching occurs at peak voltage, flux can easily start from zero and attain its nominal value. There will not be any inrush current in this scenario.

See other stuff too

A note special for Halden relating to residual magnetism /Remanence [this is the right spelling I think]??

see

f. Switching at maximum voltage crossing- maximum residual magnetism in same polarity ??and similar cases

?

?

Trent, nice unit, looks good, but can’t see any “Pin Tip leads”. Are they unusual? ?

?

Halden ?, Jim too

One part that doesn't make sense to me is the one about the 5V windings being "connected" to the B++ windings.? They're only connected when the insulation fails and the arc is active, and only one of them is known to do that.? ?

Maybe ?I’m ?guilty of using insufficient wording ??"connected" to the B++ windings. ?I meant ?"connected ?thru the 5R4” to the very HV windings.

?I still cant think of an accurate simple expression. ?Is that any help?

?

The other part that doesn't make sense to me is "When one 5 volt heater winding is at +800 RMS".? RMS is a way of quantifying an AC signal over time and is not an instantaneous value.

thats probably me too. I agree that it may be easier to use instantaneous values. ??As I see it, there a +800 RMS"

for one half cycle and a -800 RMS" half cycle ?also ?each half cycle of the HV takes its turn at grabbing the filament winding and sending very pos or very neg, with the peaks being ?800*√2? which is 1131.3708499 according to google.

?

My hi-pot testing uses DC.

?

I didn't measure the 800 Vrms directly because I don't want such high voltages around when I'm working alone.? So I used the secondary of another transformer to apply 20 Vrms to the primary and then measured the voltages at various taps on the secondary.? All measurements used a Keithley 2000 DMM, so I'm confident that the RMS values are correct.? Then, I assumed linearity and multiplied my result by 5.ok? Now that I wrote this down, I see it was an error - I was trying to think too fast!??? I should have multiplied by 5.8.? I would use 6 assuming a 120 volt primary but it may well be rated at 115, but we don’t know, do we? ??So, where I wrote 800 in this morning's post, it should be 919.? 1130 should be 1300, 1100 should be 1260, 2230 should be 2560, and I should probably turn my Bertan up to 2600 V next time. ?

you effectively had the turns ratio? measured, ??

I was only wondering if Hammond ?had the same turns ratio, they said ?120 /880volts ?or a ?1:8 ?ratio

I don’t know enough about Hi-pot, especially DC and on used transformers, so…

?

If the 5V secondaries are likely to be 10 turns, then both 5V secondaries and the 6.3 V secondary are likely to be on the outer layer of windings and accessible after digging through a few layers of crisp paper and varnish.?

?

Maybe I could find the arcing spot and add insulation to repair it.? I'm now tempted to open up the transformer and look.

Good , It might be creepage around the very ends, so test before you get the scalpel out , ?oh ?I’m late hope the tipoes fix themselves on the way.

don

?

?

Jim ,

Re ?“It was not reported that a short circuit involved the common ground tap so I am not sure that this is of concern.?“

maybe Halden understands you, but I don’t.?? I cant see any “short circuit involved the common ground tap “ ??but he did mention the HV CT ground in passing. As I read it ?????He did not seem to be concerned IMO.

However, It is important to state where grounding is, otherwise voltages might float very high wrt gnd, and it is tough to convey voltages .

Having that ground means that one terminal cant get above 800 or whatever that really is, so stresses to ground are known, all assuming the core is grounded too and grounding really means “chassied” at near local mother earth.? I’m sure you know all this, but maybe glossed over the wording.. sorry if that’s wrong.

?

I think it was not necessary to speak of? the other windings as the reported failures are from the HV and the 5V [5R4] wdg where ever it is, but when you look at obvious paths to breakdown from the HV, one expects that ??HV wdg to 5V [5R4] wdg ?has a weak insulation path by design, or perhaps by manufacturing QC and this. ??disconnecting? that 5 volt winding? will remove that high stress? before the transformer fails and after.

Hi Halden

Your posting just poped up when there is no load on GIO.

I also have contorted my head trying to get at the crux of this failure, and the reasoning behind it using other wording. Your reasoning and wording seems very good to me, and matches what I read from Jim’s writings. ???But

Jim is using peak voltage so gets over 2kv ?rather than 1600 [800+800] or what I used 1760 ?from Hammond’s ?replacement, I think Hi-pot is in rms values.

When it gets to accuracy, there are a few questions/comments:

My 1720 rms values don’t include any fudging for transformer drop or tube drop.

I wonder what the primary voltage and operating mode ?conditions were when you got 800 VAC; the reason being halli did not state a single rated voltage input such as 120 vac, but instead presented a range of useable voltages 105 to 125, and I cant see any place where the AC input is stated in the voltage chart , nor could I see the operating mode etc, ?

?