What's new
What's new

Phase Perfect Protection

converterking

Stainless
Joined
Apr 17, 2004
Location
Kolding Denmark
Since a couple of members have had trouble with their Phase Perfect feeding high voltage into their home electrical service, and Phase Perfect has not offered a solution, is there a way you could make something to automatically shut down the Phase Perfect in the event of a failure?
 
Are you talking about a "failure", or are you talking about the fact that the PP can feed back the energy of slowing down a motor into the supply line and cause a higher than normal voltage?

That is not a "failure", but rather the opposite effect to the voltage drop that a motor can produce when starting, or under load. Typically found with CNC machines, many of which can pass power back to the supply,
 
I have a PP and have not experienced this alleged back-feeding. It's hard for me to see how it would happen, or, if it did, why it wouldn't be the same with a single phase system. I suppose you could have an over-running motor load like would occur if you quickly shift a Reeves drive to low speed with a big inertial load, but (a) the same thing should happen with a 1-ph motor with no PP, and (b) I would expect the line side to be big enough to swallow the back-fed energy. Is the issue regenerative motor braking into a wimpy supply? The only other thing I can think of is if there's a fault that dumps the DC bus voltage (which has to be higher than the input voltage) onto the input. Is that possible?
 
I think converterking is talking about what fried Wheelieking's heat pump and other stuff in his house. Can't let my wife hear about any of this stuff as my shop is fed from our house panel, and I already hear enough from her about the dimming lights.
 
Since a couple of members have had trouble with their Phase Perfect feeding high voltage into their home electrical service, and Phase Perfect has not offered a solution, is there a way you could make something to automatically shut down the Phase Perfect in the event of a failure?

Yazz. Out of my "Hell box", actually. Couple of the Mercury-displacement relays part of my approach to a rapid-isolation fix so that even IF the P-P is too damaged to RESPOND to a control input, it is cut-loose, load AND supply end. Already do this with my "4Q" (regenerative) DC drives, line-side only being "good enough".

That said, such "health monitoring" and protection goods certainly exist as ready-made.

Telco's and IS/IT data center facilities have a "critical need", what with sometimes TWO separate utility company feeds, more than one gen set, and split battery plants for multiple UPS. Some two-dollar component fails, it is just NOT meant to take major banks and such offline, etc.

So... it would be better by far to buy it, "rated for the task" and "to meet code". Inch-hoorance issues as well as common-sense and actual-not-theoretical safety.

Beside, most of y'all's JOB as pays the bills is making "stuff" OTHER THAN electrical switchgear anyway.

I've just been twigged to the need/benefit of having such, unusual or no - and have not yet researched what is in the market.

Get enough "PM Eyes" onto most any problem, and solutions WILL be found!

Thanks for starting the thread!
 
Ah.. the capacitor problem..... had forgotten that one.

Well, that was surely spikes on the line from when the capacitors were not replaced on schedule (PP seems to overwork them). The 290V was the spikes fooling the meter. No matter, same thing either way.

They DO say to replace them, and if you do not do it, well, then, you might have an issue. I think PP has a problem there, they seem to be over-working the capacitors with current, and might need to do something differently.

It would not be any problem (I think I said this before) for PP to detect this problem, and shut down, latching in the shut-down mode. It is a detectable problem. many VFDs have a detector for a very similar problem in VFDs, which does not cause a problem up the line, it is just an indicator of too much load, or a bad capacitor.

Not sure it is really practical to do the detection independently, although it might be. It would probably shut down on nearby radio stations, and so forth.

One COULD do a fairly simple filter on the single phase line, that should cut the spikes, and thus the damage. Passive, very little to go wrong. Filters of a suitable sort are commercially available, if perhaps not super cheap.

Remember, the PP gets its power from the line, so the thing CANNOT produce an actual high line condition, there would have to be an energy storage (or source) out at the PP somehow (rotating motor, or good capacitors, but the caps are bad, so...)
 
Remember, the PP gets its power from the line, so the thing CANNOT produce an actual high line condition, there would have to be an energy storage (or source) out at the PP somehow (rotating motor, or good capacitors, but the caps are bad, so...)

Per our offline chat just now, there might be a way it could generate peaks above nominal line voltage. Hard to NOT do so, under one possible failure mode, actually

I submit that there are "enough" electronics resources in a P-P that the regulation control circuitry could run it as a "power oscillator" outside of its normal operating ranges if/as/when trying to do business-as-usual with a partial capacitor-array failure.

The surviving capacitors, now unbalanced, could live long enough to provide the resources to support that.

Only a SWAG. But an experienced SWAG. Our minders never much did ask us to fix stuff that was not "broken", so failure-mode detective work "R us".

:)
 
So I take it that there isn't a solution to cut power to your Phase Perfect before it burns your house down.

Nonsense. Not hard. Few knew there was a "potential" need, is all. Because there hadn't often BEEN a need.

There's plenty of sensor gear that needs no more than a half-cycle of 60 Hz worth of time and a link to any of many common contactors it can drop.

"Power fail / restart" circuit, 1968-69 GA SPC-12 "puter would detect in exactly that time frame, used whatever energy was still in the filter caps to force storage of the program-counter and contents of all CPU registers into NVRAM (Wang magnetic core), then restore it all and carry-on the march when the power came good again.

"UPS" goods and Automatic Transfer Switches have the logic as well.

NOW that we know there is a possibility? All we need is to use any store-bought "overvoltage" monitoring circuit. Common enough hardware, and not necessarily expensive.

The current-generation P-P "probably" already have it built-in. An external add-on would be good for redundancy, even so.
 
An EE could design a simple crowbar circuit

Any JRHS kid who can read could implement a simple crowbar circuit. And did. Many.

"Simple" is exactly what they are, after all. Also "BFBI" brutal.

I'll probably first ASK Phase Technologies about my new, "white cased" one.

The older technology, no longer supported (maybe we now know another reason whyso?) "Blue Case" one, I should add one of my stashed high-amperage Mercury-Displacement contactors, a hold-in circuit with an overvoltage trip-out. White-bread simple.

:)
 
Per our offline chat just now, there might be a way it could generate peaks above nominal line voltage. Hard to NOT do so, under one possible failure mode, actually

....


Peaks, spikes? ABSOLUTELY. Not only CAN it do that, with the capacitors messed up it pretty much HAS TO.

What it cannot do is a long term surge.

The spikes and peaks will be much worse. It has some form of a filter circuit in it, and if the capacitors are bad, then the inductors can produce somewhat unlimited peak voltages. Unlimited until something gives up, such as all the stuff that the other thread was talking about.

IIRC, that thread did talk about things being "odd" for a while before the "big blow up". Very possibly that was from one capacitor starting to fail, or failing. Then the others later failed also.

I am a bit surprised that the PP itself survived. But other things may have failed first, and held down the spike voltages. The reading of high AC volts was probably wrong, induced in the meter by spike voltage.

There are Fluke meters, and even some Simpson 260 meter versions, that will read completely wrong if there are peak voltages present. I had a 260 that I was reading 400V DC with, only the 400V was the output of a PFC boost circuit. The meter actually read something around 800VDC. Clearly the AC reading portion was getting involved due to the high frequency (about 132 kHz). On that same reading, various Fluke meters were obviously reading wrong as well. I finally located a stone-hammer-simple ancient 260 that read it as the correct 400VDC.

So the reading of way above 240VAC is not unreasonable on account of the high frequency spikes that were no doubt present, and which fried the poster's stuff.
 
So what can you do to monitor this and shut down the Phase Perfect?

Read what has been written the FIRST 3 or 4 times and JFDI?

This is a trivial exercise for telco, data center, other power management guys, but we can't MAKE it any simpler than "add a store-bought over voltage protector". IOW "Code complaint", NEMA labeled, UL rated, all that stuff, AND NOT "DIY", easy-peasy or not.

I wish I could just give you a part number, but with Google returning 16 MILLION Mike-Foxtrot hits?

I'd better go and eat something first, else risk falling off my perch from starvation whilst looking!

:)
 








 
Back
Top