Power capability question

This forum is for specialized infomation important to the construction and safe operation of the high voltage electrical supplies and related circuitry needed for fusor operation.
Post Reply
Dan Knapp
Posts: 402
Joined: Wed Aug 06, 2008 8:34 am
Real name: Dan Knapp

Power capability question

Post by Dan Knapp »

I have a question for the EE’s out there. I have a big commercial HV power supply that is designed for three phase input power feeding a three phase bridge rectifier to power its DC rails. I have been operating it on single phase power, but I’m not sure how much HV DC output current it can supply in this mode compared to its rated output operating in three phase mode. Neglecting the slightly lower rail voltage resulting from use of only four of the six diodes in the bridge, it would seem intuitively that the bridge could deliver two thirds of the current delivered in three phase mode. This is probably oversimplification, and the situation is more analogous to the power capability of an open delta transformer capability, I.e. 57%. Can anyone answer this question definitely?
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Power capability question

Post by Richard Hull »

While not a power electrical engineer. I am pretty sure you are operating at 240 volts on one phase. If so, the power electronics and filtering are designed around the co-joint three phases to support the electronics. Most 3 phase around here is 208 volt based for simple industrial use. I would imagine you might only have access to about 1/3 of the net rated power, if that. It all depends on your actual power needs. If I remember your talk a few years back at HEAS you were attacking fusion jointly with a friend or friends a bit differently from the simple systems most often found here.

I have the HV business end of a 208 three phase HV supply that was designed for a positive 60kv 100 ma variable output fed by a 25khz switching supply feeding a giant flyback and stack of epoxy incased voltage doublers. This was meant to be fed with the high frequency power developed at a horribly potent 3 phase supply probably gobbling up about 7kw of 3 phase power when fully loaded. I have been meaning to just get one of those beefy 240 volt HF switchers driving so many of the current guys flyback systems and hook it to my flyback and stack to see to I might get. A few years ago I spent a lot of money to buy 3 of the negative doubler epoxy discs, but there they sit to this day.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Dan Knapp
Posts: 402
Joined: Wed Aug 06, 2008 8:34 am
Real name: Dan Knapp

Re: Power capability question

Post by Dan Knapp »

My supply is rated for up to 40 kV @ 320 mA or 13 kW drawing up to 44A three phase 208VAC input. My single phase power is 240 VAC, but out of an abundance of caution I’m dropping it to 208 with a bucking transformer. The single phase feed is also a 20A circuit, so I’m limited to half of whatever I could get operating single phase. I posted on this a few years ago generating considerable discussion, but we never broached the power issue.
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Power capability question

Post by Richard Hull »

Bring it up on a variac unloaded and see what voltage output you windup with.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
Dan Knapp
Posts: 402
Joined: Wed Aug 06, 2008 8:34 am
Real name: Dan Knapp

Re: Power capability question

Post by Dan Knapp »

Pretty sure I can get it up to the spec 40 kV unloaded. The question is how much current it could deliver.
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Power capability question

Post by Richard Hull »

I'll bet you will have plenty for fusion experiments. Load it in some fashion with an milliamp meter in the circuit. Preferably on the grounded side. If it is a decent supply it should have a voltmeter and ammeter on it.
Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Power capability question

Post by John Futter »

Dan
putting 208V stright in should not prove troublesome

Some supplies do not adjust well to bringing up slowly due to stability issues

I see no problem with hitting it with your 208 volts although this is a bit low due to cube root average RMS from the 3 phase input
a couple of fast HRC fuses downstream of the variac if you still want to use it would be wise

good luck
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Power capability question

Post by Rich Feldman »

Applause for buck transformers!

This has been discussed here before, within the last 5 years, perhaps in the thread you started.
I thought it was about VFD's for motors, instead of HV power supplies, but why would the mains-to-DC-bus circuits be different?
There are plenty of references on Internet about "VFD derating for single phase input".
All models are wrong; some models are useful. -- George Box
Dan Knapp
Posts: 402
Joined: Wed Aug 06, 2008 8:34 am
Real name: Dan Knapp

Re: Power capability question

Post by Dan Knapp »

I read some of the references on VFD debating, but still didn’t find a clear answer to the question of what fraction of the power of a three phase driven dc supply can you get driving the input three phase bridge rectifier with single phase power, I.e. just driving four of the six diodes in the three phase bridge. The references mention a factor of 1.73 for the power from three phase versus single phase in VFD’s, but I didn’t find an explanation for the figure. There are multiple things involved. In addition to just the difference in current delivered by four driven diodes versus six, there is also the difference in ripple in the raw DC (the rectified three phase power being smoother than single phase since the peaks are closer together).
I’m asking once again if anyone knows of a reference that derives the 1.73 factor.
User avatar
Bob Reite
Posts: 576
Joined: Sun Aug 25, 2013 9:03 pm
Real name: Bob Reite
Location: Wilkes Barre/Scranton area

Re: Power capability question

Post by Bob Reite »

The ripple frequency will be 360 Hz (assuming a 60 Hz) supply) for the three phase connection. Operating it from single phase will make the ripple frequency 120 Hz, and it will not be filtered as well. Don't know if that would be critical for this application or not. If so, the size of the filter capacitors would need to be tripled.
The more reactive the materials, the more spectacular the failures.
The testing isn't over until the prototype is destroyed.
JoeBallantyne
Posts: 313
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Power capability question

Post by JoeBallantyne »

Dan, why don't you just get a rotary phase converter? Even if you don't get one that can match the power output capacity of your power supply, you could get a 5 or 10 HP phase converter, and limit the current you pull from the supply to contol the max load it places on the converter.

Then you don't have to worry about how to derate 1 phase versus 3, and can just limit the current so that at max voltage you draw less power than what the phase converter can supply.

Joe.
Dan Knapp
Posts: 402
Joined: Wed Aug 06, 2008 8:34 am
Real name: Dan Knapp

Re: Power capability question

Post by Dan Knapp »

Joe, I could easily put together a three phase converter from a three phase motor, but I don’t need the full design power of the supply. I’m just trying to determine how much power is possible operating in single phase mode. I keep finding the figure of 57% (or 1.73; 1/1.73 = 0.57), but I’ve yet to find and explanation of how the figure was determined. I’m still searching.
JoeBallantyne
Posts: 313
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Power capability question

Post by JoeBallantyne »

Never really took a power systems class since even 30+ years ago when I was in college they had already deprecated that from the curriculum.

The 1.73 is a rounded version of the square root of 3.

208 3 phase power is 120 volts in each line with a phase separation of 120 degrees between the lines. It is called 208 because that is 120 mulitplied by the square root of 3.

When you calculate the power of a 3 phase 208 line with 30 amps of current you are supposed to multiply 208*30*1.73. Which gives you 10.8KVA. Which is the same as 120*30*3.

This was a link I found: https://www.raritan.com/landing/three-p ... -explained

Effectively the 1.73 is the square root of 3 applied to the RMS voltage and separately to the RMS current that flows in a single line of a balanced 3 phase circuit, so that when you calculate the power delivered in the circuit it is 3 times the power delivered by a single line. Since 1.73*1.73=2.9929

I suspect that is where this factor 1.73 you keep seeing comes from.

Joe.
JoeBallantyne
Posts: 313
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Power capability question

Post by JoeBallantyne »

It is also possible that when you calculate the RMS (real mean squared) voltage for 2 nominally 120V RMS lines phase shifted by 120 degrees that you get 208V as your output.

Question is what does the graph of cos(x)-cos(x-120) look like. If you integrate the square of that over one cycle what do you get.

After 30 plus years the math is rusty and so I will leave it as a basic exercise for the reader. :-) :-) :-) (I always DESPISED it when professors did stuff like that.)

Joe.
JoeBallantyne
Posts: 313
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Power capability question

Post by JoeBallantyne »

For your scenario where you are running 240V RMS through a transformer to get 208V RMS, the total power you are delivering is just the 208V times whatever RMS current you are supplying.

Note that the phase is different because your lines are 180 degrees out of phase instead of 120.

I suspect if you push your supply hard (ie: try to pull out as much power as the nominal amount of power you can supply on the input), it will NOT stay in spec for ripple, or regulation, because it is expecting to be delivered a more steady/continuous supply of power than it is getting.

If you need your supply to perform up to its specifications, you should build that rotary phase converter, and feed it 3 phase.

If that is not really important in your application, then I suspect if you never pull more than 1/2 the nominal power you can supply the input (say 208V * 20A = 4160VA) it will likely work reasonably well. So never pull more than 2KVA on the output if you are on a 20A 240V circuit.

My 2 cents. (Which are very possibly worth absolutely nothing, no FED intervention required at all.)

Joe.
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Power capability question

Post by Rich Feldman »

I was about to suggest derating to 2/3 of nominal power, based on average current in the input bridge rectifier.
Two of six diodes are idle when you run on single phase, so the high side diodes (for example) each conduct 1/2 instead of 1/3 of the DC bus current.

Under that condition, I bet the ripple current in DC bus capacitors will be greater, and true-RMS current in the 4 diodes that are working.


Could simulate it with a free SPICE program if not confident in one's analytic answer.
Even get fancy and simulate front end power-factor-correction circuit.
All models are wrong; some models are useful. -- George Box
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Power capability question

Post by John Futter »

That is why I said use the 240 volts natively multiplied by root of 2 gives 339 volt dc buss
208 x root of 3 phase gives 359 volts DC buss.
In fact you might need a boost transformer to bring it up a little ( 20 volts) but I reckon that it will go well until the DC buss caps complain about the high ripple current

check to see if it works if it locks out on under voltage you know what to do

i seriously doubt it will lock out on over voltage
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Power capability question

Post by Rich Feldman »

There's no place for sqrt(3) in conversion from sinusoid RMS to peak voltage.

Sqrt(3) is the ratio between 208 and 120 volts.
208 is a nominal phase to phase voltage for three-phase in 120V land.
Bridge rectifier will charge bus capacitors to 208 x sqrt(2) = 294 volts. In six pulses per cycle, or 2 pulses if one wire is not connected.

p.s. the next higher voltage for 3-phase here is 277 V phase-to-neutral, 480 V phase-to-phase.
AFAIK, 3-phase motor nameplates etc would say 480 V, while light fixtures for the same high-bay room would say 277 V.
All models are wrong; some models are useful. -- George Box
John Futter
Posts: 1848
Joined: Wed Apr 21, 2004 10:29 pm
Real name: John Futter
Contact:

Re: Power capability question

Post by John Futter »

okay I thought 3 phase rectifiers did around approx 2.5 times the rms over single phase root 2 times the rms
correcting my mistake above
average dc load voltage
Where: VS is equal to (VL(PEAK) ÷ √3) and where VL(PEAK) is the maximum line-to-line voltage (VL*1.414).

VDC = VS x (cube root of 3/pi) = 1.65 x Vs

208 x 1.4142/1.44 = VS = 204.3

VS x 1.65 = 337 VDC


240 volts single phase gives 240 x 1.4142 = 339 volts

close enough for rock and roll
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Power capability question

Post by Rich Feldman »

I just learned a new trick in LTspice, while preparing this demo.
Can interactively set differential voltage probes between, say, phase 1 and phase 2.
3phase_rect_sch.png
3phase_rect_sch.png (9.8 KiB) Viewed 4426 times
The three "sine" voltage sources each put out +/- 170V (for 120 V RMS).
Chart shows their voltages (with respect to neutral) in green, blue, and red. Phase 3 node is called n001 because my name tag missed the wire.
Three bigger sinusoids represent the instantaneous phase to phase voltages, which are +/- 294 V ( 208 V RMS ).
At the top we see DC bus voltage, the voltage on n_VP with respect to ground. It peaks at 294 volts minus 2 diode drops.
3phase_rect_tran2.png

Just for fun, I reran the sim after disconnecting phase 3. That's analogous to running on 208 volt single phase.
Three differential voltage probes remain on the diode junction that used to be phase 3, and now floats around, weakly tied by reverse bias behavior of the diode model.
Colors might be different. We could add charting of phase currents, diode currents, and capacitor currents to get insight mechanically.
3phase_rect_tran3.png
All models are wrong; some models are useful. -- George Box
User avatar
Rich Feldman
Posts: 1470
Joined: Mon Dec 21, 2009 6:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Power capability question

Post by Rich Feldman »

For reference, here's the view when two 120V sources have phase difference of 180 degrees. Phase to phase voltage is +/- 340 V (240 V RMS), and DC bus voltage reaches 340 V as we all agree.
3phase_rect_tran4.png
Now back to the 3 phase, with load resistor now 50 ohms so average bus current is almost 6 A.
I charted the phase-to-phase voltages on both sides of phase 2, and the current pulses from phase 2 (pink) and the other 2 phases. All peaking at about 30 A in this simplistic model.
3phase_rect_tran5.png
Lower picture shows phase 3 disconnected. Capacitor-charging pulses are 3x less frequent. Each surviving phase delivers 50% more charge than before, in 2 instead of 4 pulses per cycle, with current amplitudes approximately doubled.
All models are wrong; some models are useful. -- George Box
Post Reply

Return to “High Voltage - Fusor Input Power (& FAQs)”