What's new
What's new

Is it possible to fool a 3ph input VFD into using 1ph input for household use?

superUnknown

Aluminum
Joined
Jun 9, 2013
Location
Canada
I just blew up a cheapo huyan yang 7.5kw 1phase input. I've had much better luck with the automation direct GS2 drives. BUT the biggest one they make for 1ph input is 3hp.

In the past, when I tried to hook up a larger GS2 to single phase, it (not surprisingly) shows an input phase error. I don't have a big GS2 to try it out on anymore.
VFD schematic.jpg

Since the DC bridge doesn't care where it gets power from, I'm wondering if it's possible to fool the drive by connecting two terminals to the same single phase "hot"?

(Yeah, yeah, "shouldn't", "not rated for amperage"... "gets hot"... "let the smoke out" :rolleyes5:)

Is it possible to "fool" the phase detect part of a VFD?:scratchchin:
 
Thats strange, all the drives I have ever used will also take single phase, and run fine as long as you size them right. I know for sure that the Hitachi drives will work just fine, they say if its a 3 phase drive, to double it for single phase input, so 10HP motor is a 20HP Drive, but most people I know go about 1/3 larger and they work fine. Also driveswarehouse lists larger drives, for 1 phase input too.

This of course works if you don't already have the one you are talking about.
 
That's good to know! Must just be a pain in the ass "feature" of the GS2. I'll start saving my pennies for a "real" drive...
 
I have a 10HP Mitsubishi drive, and a TECO Ma7200 10HP drives that both are powered via 1Ph input. I'm using both on 7.5HP machines. I have not had any problems.
I too blew up the 7.5kw huyan drive, and it was on an unloaded 4HP motor. They are junk, do not have the same programing features as drives that are 10 years old technology. eBay seller replaced it as failed same day I received.
I know there are drives that say they require 3 phase input.
Your diagram is confusing, relative to the 700VDC. 240VAC time sq. root of 2. (1.414) = 339 volts peak. This would be double with 440AC input.
Now I would like to know how to trick a 440 VAC drive to work on half that, to drive 240VAC motors, as there's tons of these on eBay.
If your diagram as shown on the input does not work, I don't know if a RC phase shift (series resistor capacitor, best you can do would be a 90 degree phase shift, but you can cascade two phase shifter) could be connected if the input phase detector was electronically more clever. This would be a series RC connected between the 2 AC power leads, and the new connection, to the 3rd phase input.
phase shifter.jpg

The time constant needs to be 1/60 of a second (60Hz system), so the product of RC needs to = 0.0166. If we use a 10 micro farad capacitor (0.000010 farads), the resistor needs to be 1666 ohms. You don't need exact values, but the capacitor needs to be 350v (working volts or more). If you scale the capacitor down, the resistor gets scaled up by the same multiplier (the product of R and C needs to be 0.0166). I state this as smaller capacitors are easier to get with higher voltage ratings.
If I'm wrong in this, and I can be, someone will point this out. And as usual, I may create more confusion.
 
Oh man, I gotta say, ignator, you are the man!

Huyan; yeah we should all know better... Mine ran for maybe 30 hours at steady 3.5amps. Then I ran it at 7.1amps for 4 minutes, I could smell it getting hot then ZAP! Input traces on the PCB went phase to phase, it must have delaminated off the PCB?!! Garbage!

huyanjunk.jpg

Very perceptive of you to notice the 700VDC bus. You're right, it's from a 440VAC input schematic. I didn't think anybody would notice. Sorry for the confusion.

Phase shift is a cool idea. I was trying to figure out the RC to put in there. But looking at the price difference between a good drive and the automation direct, it's better just to buy a Hitachi...

So, I checked the diodes, caps, and IGBT; ALL OK! Then I soldered a bodge wire in place of the PCB trace. Flashed it up, and the Chinese piece-of-shit lives again! So enjoyment, double happiness!
 
So, I checked the diodes, caps, and IGBT; ALL OK! Then I soldered a bodge wire in place of the PCB trace. Flashed it up, and the Chinese piece-of-shit lives again! So enjoyment, double happiness!

I've gotten the best prices from Dealers Electric.
Teco Westinghouse Phase Converter, AC Adjustable Speed Drive Systems, Distributor
I have 2 of the 10Hp MA7200 vector drives. I'm not pushing any of the machines to their full HP rating, so I'm not derating 50% as the above link indicates.

Lucky, normally a toasted board makes carbon, that forever will burn between voltage inputs. But this clearly indicates they are not designed per any code where wire size (and circuit traces are wire) must be sized for circuit breaker wire size protection. I betcha they didn't maintain the circular mils of copper required. My drive made a very load bang (assume this was DC bus capacitor). I'm going to keep it for a metal bandsaw conversion (4x6 is probably the best I can expect).
I've gotten lucky by keeping an eye on eBay, and getting BIN on very affordable VFDs. I typically stay away from any drive that the vendor does not keep the instruction manual online (allen bradley/rockwell automation is the worst for old documents, at least I've never found them, but see lot's of used ones with no manual).
 
My drive made a very load bang (assume this was DC bus capacitor).

I thought that too, made a "pop" just like a cap blowing. But hey, easy fix. Please take yours apart! We wanna see what blew!
 
My drive made a very load bang (assume this was DC bus capacitor).

I thought that too, made a "pop" just like a cap blowing. But hey, easy fix. Please take yours apart! We wanna see what blew!

I sent that drive back without breaking the seals. There was no magic smoke, so that was what I assumed the problem was. It was dead, no low volt power to drive LED display. Huyan eBay auction indicates it's based on the "Adopt Japan Mitsubish chip", so my assumption is they stole/got 10+ year old intellectual property.
 
Ahhh soooo. Most-ah famous mitsubish branding.

Ha! There wasn't any intellectual property in there! Just a couple of off brand caps and some counterfeit Fairchild Semiconductor IGBTs...
So when the replacement blows... you know what to do ;)
 
Your diagram is confusing, relative to the 700VDC. 240VAC time sq. root of 2. (1.414) = 339 volts peak. This would be double with 440AC input.
Now I would like to know how to trick a 440 VAC drive to work on half that, to drive 240VAC motors, as there's tons of these on eBay.

Capacitor-based voltage doublers were used in PC power supplies for ages (the little red 110/220 switch) - you can sometimes do the same with VFDs with a bit of tweaking.

Essentially, you need a drive with two capacitor (banks) in series. Connect the neutral (or one phase if it's a 240V centre-tap system) to the centre tap, and the (other) phase to the rectifier input. Each half-cycle will charge one half of the capacitor bank to peak voltage, instead of the whole capacitor bank.

I suspect more derating would be needed as the caps would see more ripple current. Alternatively, get a drive with DC+ and DC- terminals and connect extra caps externally.
 
Whether or not a VFD can accept single phase input is really just about how the VFD mfr designed their power supply for the electronics inside. Some mfrs chose to tap off of the DC bus with a DC-DC chopper power supply, so they don't care if the line side is 1 or 3 phase, others choose to use an AC to DC power supply on the line side, so they do. Some of those with the AC power supply then allow you to just jump two of the line terminals together so that it powers the unit, some do not (I don't know why).

But technically, the power electronics portion only "cares" about 2 things; AMPS going through the diode bridge, and DC bus ripple.

Amps on single phase will increase on the diodes by the difference between single phase line current and 3 phase line current, the square root of 3 (1.732x). So if a VFD is rated for 10A on the output, the diode bridge would be rated for 10A on the input too, but if you feed it with single phase, the INPUT current increases to 17.2A, hence the need to de-rate.

The DC bus ripple issue is that when you have 6 diodes from a 3 phase bridge rectifier feeding the DC bus (2 per phase), the DC bus ripple is never going to dip to Zero VDC in between "pulses" (how a diode conducts). But when you feed the drive with single phase, the DC bus ripple DOES drop to Zero VDC 120 times per second. DC bus ripple can cause damage to the transistors used to recreate the AC going to the motor, so there are systems in the DC link of the VFD to smooth it out. Larger more expensive VFDs use a combination of a DC link choke (reactor) plus capacitors, but smaller Asian made drives use capacitors only (because they make ALL capacitors, so they are cheaper for them). So to make sure the DC bus is smooth enough, the drive needs MORE capacitance, plus the capacitors will work harder and get hotter. That then affects the amount of de-rating you need.

De-rating rules-of-thumb then are that if the VFD has a DC bus choke, you must de-rate the VFD by at least 50% of it's rated capacity. So if you want to power a 7-1/2HP 11A motor, you need a 15HP 22A VFD (but always go by AMPS, not HP).

If you have a smaller cheaper Asian based VFD, you must follow the mfr's suggestions on de-rating. Some however will not give them to you in the manual, you must ask. There is a general trend to just use the 50% de-rate number, but that is not what some of them will tell you, because it is more complicated than that due to heat. So what you will find IN WRITING is often a much steeper de-rate of 65% (VFD de-rated TO 35% of its rated capacity). So if you have an 11A motor, you need a VFD rated for 31A, not 22A. Alternately, if you want to use the 50% de-rate rule, then you must de-rate the ambient temperature by 50% as well. So if you have a VFD rated for 40C operation, you must keep it at 20C,or 68F, which is all but impossible without an air conditioner. This rule is debated, a lot, but if you don't believe me, ask for a manufacturer's de-rate formula IN WRITING whereby they will back it up under their warranty.
 
Once in a while you may run into a newer VFD that has just enough bus capacitance to be stable. These may have NO electrolytic capacitors (unreliable, short lived, and temperature sensitive), they may just have one film capacitor (reliable, long-lived, stable) as a bus capacitor.

That works OK on 3 phase input, but for single phase (which goes to zero as noted by jraef), it will really not work at all.

Those will not have any single phase rating at all. They have to use three phase.

We did a series like that for a COTS MIL application, for instance (we needed reliable -40C operation). It is a newer concept that has apparently some presence in the industrial drive area as well as MIL.
 
Hey unknown, possibly cogent but maybe not. As I am sure you know, diodes only conduct when the voltage on their junctions is greater than what's downstream from them. As a result with a cap input filter they frequently supply current in big bursts over very short conduction periods

And as such when pushed hard (assuming cap banks) it's good to have a little impedance in the supply side, integrating those nasty high current peaks and damping voltage spikes somewhat.

Cheap, quick and dirty add a 3% line reactor on the incoming AC line... just make sure the reactor is up to the task in terms of amperage rating. Something more than the primary fusing is advised. This might buy you a little margin.. and it's just good practice,
 








 
Back
Top