# Three phase converter - capacitor switching

I've started building a single to three phase converter. My research started with the likes of this paper. I've set a bank of 40uf/440v capacitors up and a potential relay. The relay is a Supco SUPR. That relay was ditched early on. Even though it was rated for a 5hp motor, and I'm using a 3hp, the contacts on the relay would often weld together and the starter caps weren't taken out of the circuit. The Supco got bad reviews about this. Also I found that the capacitors really should be switched as the load demands. I'd just gotten an old lathe, so two tools in the shop are three phase. And yes, I know a modern solution is to use a VFD. But that means I'd need one of these for each piece of machinery. I'd have to hack if it required a 3 phase input. And they are rather pricey. I would rather have an old school 3 phase bus that everything can run from.

I have searched and I can not find an application of switching capacitors with triacs for a phase converter. But it just makes sense to me to do so rather than using relays. So what I'm imagining from here is to switch each capacitor as demand requires. I've already done some testing with this mess. I still plan on keeping a 3hp motor without a load on the buss to aid in the rotary conversion so this is not just a capacitor converter.

And it looks very doable. When I started I mistakenly started using an MOC3010. I thought I had a zero crossing isolator on the shelf and didn't check the spec. That through me for a loop as I started switching capacitors with large potentials at the terminals. Well of course I destroyed a couple of triacs. Using a 2.7 ohm 25 watt surge resistor in series with the triac I was able to keep testing until the MOC3041 come in the mail. I have no experience working with high current, low frequency, phase shifting. Right now I know enough to be dangerous.

So, after that long winded introduction, my question. Using a microprocessor that can read voltages and switch triacs, what would be the most appropriate software method to keep phase and current demand as ideal as possible. Am I on the right track? I can do fine with the micro when I have an objective.

And it looks like this is what I'm trying to build.

Update:

The lathe, a Sheldon 15, has a two speed motor. 5hp@3600rpm and 2.5hp@1800rpm. The mill has a 2hp motor. The idler motor is 3hp. I've just recently upgraded the mill to LinuxCNC over ethernet. My main is 200 amp. I'll run from a 50 amp, that should be more than enough.

Update 2: I know I could just buy a rotary phase converter over the counter. In fact, I already have it. All I would have to do is get a better potential relay to finish this off. I've run the lathe, both speeds, on what I have so far. In fact, I've started the idler by hand with the triac. What I'm looking for is a better mouse trap for me and others that would want to go there. Like I did this a couple of weeks ago.

• I see that buying one is a bit expensive \$650 USD, but do you really think your saving any money by trying to build something fairly complex and needing specific expensive parts? Save your pennies and stop trying to re-invent the wheel. VTC as too broad in scope. – Sparky256 Jan 19 '18 at 21:03
• If the phase angles of L1, L2 are not 120 deg apart, then it may be pointless to generate L3 and then the triac opens as a ZCS. – Tony Stewart Sunnyskyguy EE75 Jan 19 '18 at 21:08
• Please edit your question to include the specific power input requirement of the motors you wish to drive, AND the power that is actually available. It may be that there is a better solution you have not thought of yet. – mkeith Jan 19 '18 at 21:19
• Hi @Tony Stewart. EE since '75 I don't understand how you can have a voltage phase angle with just two wires. It doesn't exist, (or just considered zero), from what I know. By generating voltage on a third lag now there are phase angles. – lakeweb Jan 20 '18 at 0:13
• I gave you the answer, consult an industrial electrician. It probably already can connect to what you have, since most 120V three phase equipment can be wired to 240 single phase without any additional parts. – drtechno Jan 20 '18 at 0:43

The capacitor bank is switched on the condition of keeping the current at L2 between half and equal to the current at L3.