Variable Phase AC Mains Switching Source for testing amps' in-rush current, etc

[nevetslab]

Every once in a while, I find myself wishing I had a Phase-settable AC Mains power switching box. I did have a simple AC Mains Power Panel which had a 30Amp Contactor (Relay) inside, and had built a small timer to control the circuit that turned that contactor on and off, which allowed power on/off cycling of amps in production. I just never went beyond that to add a phase-adjustable circuit so one could dial in WHERE on the AC mains cycle the contactor would close.

At the moment, I had a second of two Fender Twin Reverb amps that both came in last week from inventory, both with blown 4A S/B mains fuses. One amp did have a power tube socket that wasn't getting heater current thru to the power tube. Had to burnish the forked terminals pins 2 & 7, as well as increase the contact tension to restore the tube plugged in. But, I never did find what caused the 4A mains fuse to blow.

This second amp, same deal with the blown fuse, though this time, the fuse cap in the holder wasn't fully inserted and turned closed.. I don't know if it had been that way for some time, and arc'd, blowing the fuse. No issues with the tube sockets on this one. Not finding any abnormal current draw with either amp. That always bothers me, NOT finding what caused enough current to flow to blow the fuse.

Added note 1/15/19: The fuse holder with a fuse in the fuse cap will NOT pass current, even if pressed all the way in, UNTIL it is rotated & locked in place. So, no way for arcing if the fuse holder somehow came loose.

[TimmyP1955]

Sometimes it's venue power, or the sound company's distro. We've had loose a neutral connection give things fits. We've been fed 240V - twice by the same folks!

[SoulFetish]

Every once in a while, I find myself wishing I had a Phase-settable AC Mains power switching box. I did have a simple AC Mains Power Panel which had a 30Amp Contactor (Relay) inside, and had built a small timer to control the circuit that turned that contactor on and off, which allowed power on/off cycling of amps in production. I just never went beyond that to add a phase-adjustable circuit so one could dial in WHERE on the AC mains cycle the contactor would close.

At the moment, I had a second of two Fender Twin Reverb amps that both came in last week from inventory, both with blown 4A S/B mains fuses. One amp did have a power tube socket that wasn't getting heater current thru to the power tube. Had to burnish the forked terminals pins 2 & 7, as well as increase the contact tension to restore the tube plugged in. But, I never did find what caused the 4A mains fuse to blow.

This second amp, same deal with the blown fuse, though this time, the fuse cap in the holder wasn't fully inserted and turned closed.. I don't know if it had been that way for some time, and arc'd, blowing the fuse. No issues with the tube sockets on this one. Not finding any abnormal current draw with either amp. That always bothers me, NOT finding what caused enough current to flow to blow the fuse.

Added note 1/15/19: The fuse holder with a fuse in the fuse cap will NOT pass current, even if pressed all the way in, UNTIL it is rotated & locked in place. So, no way for arcing if the fuse holder somehow came loose.

The way I see it (if I understand you correctly), you have two issues to contend with. There is the issue of zero-crossing switching which causes large current spikes in the transformer until the flux in the core stabilizes. The other is the remanence in the core causing large inrush spikes, depending on the phase relationship of where the power was switched off and then reapplied.
There is much I still have to learn on this, so I apologize if this is a dumb idea. But, I wondered if one could use a high hysteresis device, like core memory, as a phase control to optimally apply power on. Couldn't it be wired so the remanence in it's core maintained the same phase as the transformer during power off? You could then amplify the voltage induced during the opposite phase in the mains cycle as a control voltage. I dunno... just a thought.

[Helmholtz]

There is the issue of zero-crossing switching which causes large current spikes in the transformer until the flux in the core stabilizes.

This I don't understand. If the core is not in a magnetized state (remanence), the high primary inductance prevents large turn-on spikes. Induction counteracts rapid current increase.
There is increased turn-on current though (not spikes), caused by the filter caps charging.

The other is the remanence in the core causing large inrush spikes, depending on the phase relationship of where the power was switched off and then reapplied.

^^This is the main reason for blowing fuses/triggering circuit breakers (in the absence of faults). If power is switched off near the mains current maximum, the core stays magnetized and turning on power drives the primary into single sided saturation for several cycles. Core saturation means low inductance and high (magnetizing) current.
Mostly seen with high power toroidals, as their cores have no air gaps (other than laminated cores).

But I understood that nevetslab is not looking for an inrush current limiter (suitable plug-in devices are readily available) but for something that allows to find and reproduce the worst case turn-on phase instant.


Regarding unexplicable fuse blowing there might be another reason. I think that standard fuses have a limited lifetime caused by fuse wire fatigue. One can sometimes see the fuse wire jump during turn-on. And any metal wire can only take a limited number of bends until it eventually breaks.

[nevetslab]

^^This is the main reason for blowing fuses/triggering circuit breakers (in the absence of faults). If power is switched off near the mains current maximum, the core stays magnetized and turning on power drives the primary into single sided saturation for several cycles. Core saturation means low inductance and high (magnetizing) current.
Mostly seen with high power toroidals, as their cores have no air gaps (other than laminated cores).

But I understood that nevetslab is not looking for an inrush current limiter (suitable plug-in devices are readily available) but for something that allows to find and reproduce the worst case turn-on phase instant.

Regarding unexplicable fuse blowing there might be another reason. I think that standard fuses have a limited lifetime caused by fuse wire fatigue. One can sometimes see the fuse wire jump during turn-on. And any metal wire can only take a limited number of bends until it eventually breaks.

Will a power transformer that was switched off at that random chance of it being at the peak of the power cycle, hold that magnetized state over days, or weeks, until it again is turned on, and do as you described? I hadn't thought about that, but, it is one of those things that could be looked at if I ever get to building a variable-phase power service for doing these types of tests.

I haven't checked Littelfuse, Bussman, Bel or other fuse mfgr's data sheets regarding life-cycles. It's always a roll of the dice when you turn on the juice as to how much current will flow.

[Helmholtz]

Will a power transformer that was switched off at that random chance of it being at the peak of the power cycle, hold that magnetized state over days, or weeks, until it again is turned on, and do as you described? I hadn't thought about that, but, it is one of those things that could be looked at if I ever get to building a variable-phase power service for doing these types of tests.

Yes, demagnetizing requires a reverse current/field.

[SoulFetish]

This I don't understand. If the core is not in a magnetized state (remanence), the high primary inductance prevents large turn-on spikes. Induction counteracts rapid current increase.
There is increased turn-on current though (not spikes), caused by the filter caps charging.

Your right, spikes was the not the correct term. "Surge" is what I meant. But I was talking about the effect of zero switching which is described in this paper:

AppNote_0513_Beware_of_Zero-Crossover_Switching_of_Transformers.pdf

[Helmholtz]

Your right, spikes was the not the correct term. "Surge" is what I meant. But I was talking about the effect of zero switching which is described in this paper:

[ATTACH]51979[/ATTACH]

Thanks, interesting article. Makes sense. I wasn't aware that even in demagnetized state real world PTs (having low saturation headroom) may be driven into asymmetric saturation when turned on at zero voltage.
The rest of the explanation is the same as for magnetized cores.

[nevetslab]

Thanks, interesting article. Makes sense. I wasn't aware that even in demagnetized state real world PTs (having low saturation headroom) may be driven into asymmetric saturation when turned on at zero voltage.
The rest of the explanation is the same as for magnetized cores.

Indeed interesting article. And here I was thinking it was less stressful turning on at / near the zero crossing than at peak AC in the cycle.

[Dave H]

Indeed interesting article. And here I was thinking it was less stressful turning on at / near the zero crossing than at peak AC in the cycle.

You and me both. I thought that was one of the the reasons for SSRs being zero voltage switching.

[Helmholtz]

As can be seen from pictures 1A/1B in the article, magnetic flux increases continuously during the first positive half-cycle as long as voltage stays positive. Its peak value is determined by the area below the voltage half-cycle. This area is maximal when voltage is turned on at zero.
In steady state, half of the flux increase is used up to demagnetize the flux left from the opposite half-cycle.

So it's the first voltage half-cycle after turn-on that determines peak flux and high saturation current peaks, depending on residual stored flux.

[g1]

Is residual stored flux dependent on where in the cycle turn-off occurred? I seem to recall seeing that but couldn't find it again.

[Helmholtz]

Is residual stored flux dependent on where in the cycle turn-off occurred? I seem to recall seeing that but couldn't find it again.

Yes, I mentioned that somewhere above. Residual flux or remanent flux depends on momentary magnetizing current at the instant of turn-off. (AC flux is in phase with magnetizing current). As the phase shift between magnetizing (not total!) current and voltage is always close to 90° (see figure 1A in the article), turn-off at voltage zero crossing means max. remanent flux. If this causes saturation at next turn-on depends on relative voltage polarity.