Run capacitor for a DIY 600V rotary phase converter

[jim rozen]

3 hp load motor, with a 5 hp idler, will probably need tuning via capacitors to get it to peform well. In my case the idler is 5 hp, and the largest load motor is 1 hp. This is why I can get away with zero capacitors in the system. All you need to do is look at digikey to see 600 working voltage capacitors. They're not that expensive.

 

[TRC]

3 hp load motor, with a 5 hp idler, will probably need tuning via capacitors to get it to peform well. In my case the idler is 5 hp, and the largest load motor is 1 hp. This is why I can get away with zero capacitors in the system. All you need to do is look at digikey to see 600 working voltage capacitors. They're not that expensive.

Thank you. I'll look into that.

 

[Bob-J-H]

If you take 370 volt caps and wire them in series, they will become 740 volt caps but the mfd will be one half of the stated value.

 

[jim rozen]

That works as well but a high value resistive voltage divider should be used to be sure one capacitor does not 'hog' all the voltage.

 

[JST]

Also derate a bit. Different capacitor values (within tolerance) will cause the AC voltage to divide differently than expected.

I'd de-rate 15-20 percent. So 2x 370VAC which should be 740V, probably would be better limited to 630VAC. Still likely OK for the "600V" level.

Be aware that 600V is the point at which UL drops out. They only deal with voltages below 600V.

 

[TRC]

Be aware that 600V is the point at which UL drops out. They only deal with voltages below 600V.

I didn't learned electricity related stuff in english, could you define what is this UL?

 

[jim rozen]

UL (safety organization) - Wikipedia

en.wikipedia.org

 

[TRC]

Also derate a bit. Different capacitor values (within tolerance) will cause the AC voltage to divide differently than expected.

I'd de-rate 15-20 percent. So 2x 370VAC which should be 740V, probably would be better limited to 630VAC. Still likely OK for the "600V" level.

Be aware that 600V is the point at which UL drops out. They only deal with voltages below 600V.

Well, my circuit is running at 620V for now, but it's supposed to be 575V, so technically, I should be under 600V.

UL (safety organization) - Wikipedia

en.wikipedia.org

Thank you!

 

[rklopp]

I power my 575-V Canadian Aciera F-5 from 240-V single phase using a pair of 240-575 transformers in an open delta configuration AFTER the 240-V 1-phase to 3-phase converter. It works just fine either using an RPC or PhasePerfect. The transformers were about $150 apiece, one being NOS and the other used.

 

[JST]

Well, my circuit is running at 620V for now, but it's supposed to be 575V, so technically, I should be under 600V.


,,,,,,,,,,,,,,,

Voltage can be 10% different (high or low) from nominal and still be fine. From 575, that would be right at 632V. Probably OK.

"Balance" capacitors can cause some really high voltages with no load. That's where the problem may occur.

 

[TRC]

Voltage can be 10% different (high or low) from nominal and still be fine. From 575, that would be right at 632V. Probably OK.

"Balance" capacitors can cause some really high voltages with no load. That's where the problem may occur.

By no load, are we talking while idling, with no motor connected, or are we talking with a motor on the circuit without any load?

 

[JST]

Idling, RPC on, no motor connected.

There still may be connected control circuits.

And, it includes also cases where the load is very light.... Spindle not on, but maybe coolant pump is on, or some other small load..... That sort of case. The small loads won't put enough load on to pull down a high voltage due to balance caps being too aggressive.

 

[TRC]

Idling, RPC on, no motor connected.

There still may be connected control circuits.

And, it includes also cases where the load is very light.... Spindle not on, but maybe coolant pump is on, or some other small load..... That sort of case. The small loads won't put enough load on to pull down a high voltage due to balance caps being too aggressive.

Is this high load going to cause damage if not taken into account? If yes, do you know a quick way to take it into consideration?

 

[JST]

The high voltage "may" cause damage to something, it just depends what is connected.

The "correction" is not to permanently connect more capacitors than will boost the no load voltage to 10% high.

If more are needed for some loads, then connect those capacitors only when the loads are "on". After the "start" contactors in the machine (but before the motor overload protector).

 

[TRC]

The high voltage "may" cause damage to something, it just depends what is connected.

The "correction" is not to permanently connect more capacitors than will boost the no load voltage to 10% high.

If more are needed for some loads, then connect those capacitors only when the loads are "on". After the "start" contactors in the machine (but before the motor overload protector).

Since this is for a mill, does it mean that I'll need have some sort of cap bank that I'll need to turn on and off depending of the load?

 

[JST]

No.

The "balance" capacitors are there only to raise the voltage of the generated leg a bit. Everything will run with NO "run" or "balance" capacitors. It will just be a little less efficient.

Given that, your best move is to apply loads, determine how low the generated leg drops, and see if it is enough to cause issues (odd noises, excessive motor temp, etc).

You can then decide if you want the balance capacitors, and how much. Or, you can just figure you need some for best operation, and put in enough to raise the unloaded output on the generated leg to 8% or so over the incoming voltage.

Run things that way, check some voltage under load, look for issues of noise or overheating. Those can be handled for loads that need them the way I suggested above.


Since the object is to raise voltage, a transformer can be used to do that instead. You can boost the voltage going into the RPC idler (not the powered loads), or boost the generated leg coming out.

I would not obsess. This is not a critical parameter, it is merely "helpful" in most cases. It may help rectified loads like CNC machines (or not).

 

[jim rozen]

Goal here is to get the line-to-line voltages all the same within about 5 or 10 percent, while your load motor is running.

 

[rons]

View attachment 377020

Your labels are unconventional:
- Incoming AC1 and AC2 should be labeled L1 and L2, or at least connect to them. You use L2 and L3.
- On the output side you re-label L2 to L3 and L3 to L2.
- You use L1 as the generated phase. It should be labeled L3.

 

[scsmith42]

I already have the motor, I won't go and get another big motor in my small shop. I will consider this as a B plan, but I'll stick to 575V at the moment.

I know that and I purposefully choose not to do it because it would be more expensive than purchasing a few capacitors, a switch and get an electric box from the scrapyard.

Gentlemen, thank you for these recommendations. I'm glad that you are taking your time to give me advice.

However, I'd like to keep the subject of this conversation to the question asked or other information related to building a RPC from scratch. This post is to gather information about whether it's feasible or not, not to get told not to do it because it's not the easiest way.

I'll make a decision afterward, don't go think I'm 100% going this way.

With respect to your request above, here is my input. For a RPC you want a high rpm donor motor. 1175 is fairly low and it won’t allow you to start loads that you could start with a higher rpm motor.

I once made a 40 hp RPC from a 1075 rpm motor. It would not start a 15hp edger or a 10hp air compressor (that had an unloaded valve on the compressor head) - unlike my 30hp 3450 rpm RPC. When I looked into the reason why, I learned that higher rpm is necessary to overcome current inrush on high load motor startups.

Perhaps in your case the current inrush won’t be a problem. For my use, it was and I ended up selling the 40hp RPC at a loss.

 

[TRC]

With respect to your request above, here is my input. For a RPC you want a high rpm donor motor. 1175 is fairly low and it won’t allow you to start loads that you could start with a higher rpm motor.

I once made a 40 hp RPC from a 1075 rpm motor. It would not start a 15hp edger or a 10hp air compressor (that had an unloaded valve on the compressor head) - unlike my 30hp 3450 rpm RPC. When I looked into the reason why, I learned that higher rpm is necessary to overcome current inrush on high load motor startups.

Perhaps in your case the current inrush won’t be a problem. For my use, it was and I ended up selling the 40hp RPC at a loss.

Fortunately, I also have a 3k rpm 575v 5hp motor lying around. Good to know such an information, thank you for it!