Emcoturn 120 lathe retrofit

When you operate on single phase, the input current is shared between 4 input diodes; with 3-phase it's shared between 6 diodes. The dissipation in a rectifier diode is proportional to the average current, so in single phase mode, the diodes run ~50% hotter. So VFDs are often derated by something like 35% for single phase operation. Or conversely the diodes are often overspecified to begin with so they can run single phase without derating.

Unless the VFD has an active PFC front end (which these don't), it should work fine. Apparently some VFDs have a "missing phase" detection but almost certainly not these ones. The voltage ripple on the DC bus is slightly higher but that won't be an issue.

Unless you are planning to run the VFD at full load continuously, I doubt there would be a problem other than due to the quality of the VFD(!). The motor won't be rated for continuous operation anyway, so it's probably rather academic.

andypugh wrote:

LutzTD wrote:

I was doing more detail layout of the Automation Direct VFD and have hit a snag. The 5hp version of this drive is 3ph input only. Does anyone have experience with the Huanyang inverters on Ebay? Looks like the 5hp and the 10hp versions allow 1ph input

You can see if the AD drive works on single phase. It might well do. You should probably de-rate somewhat, but to be honest you are unlikely to hit 5hp during normal turning activities.

I have the 5hp HY VFD on my lathe, and it seems to work nicely, and was very easy to configure with (It's own special variant of) Modbus and a USB to Modbus dongle using the LinuxCNC hy_vfd component.

Hey Andy, did you have a 5hp or a 3hp in your lathe? Is the HY VFD you have simialr to the one pictured or is it another model?

Thanx

Scott

thanx, looks like I may go that route. Another unrelated question. I see you have the VFD close to the motor, is that because it is convenient mounting spot or is that to get it out of the control or some other electrical noise issue?

I am pushing my expected noisy components and AC stuff to the left as that is the shortest route out of the box. I planned to put the VFD at the lower left and run the motor wire up and out in the left wire track. Would you suggest I put it in the housing behind the lathe and just run ac power and the signal out to it there? I am not sure how the VFD works but I assume it to be a very noisy component like a switching PSU or such. I do have the 24vdc power supply on the right, which I also think is a noisy component but my intention was to possibly put the computer, the Mesa and the toolerator3000 in a walled area and run all of the lower power lines along the top and out about 4 inches to the right of the high power hole. What are your thoughts, can I do more, should I do more? I would like to get shielded wire for the motor, the wire in your picture looks like it has a braided jacket, do you have a source for that? Thanx again

andypugh wrote:

LutzTD wrote:

Would you suggest I put it in the housing behind the lathe and just run ac power and the signal out to it there?

I am no expert on electrical noise. But as far as I can tell the main problem with a VFD is the noise they send back down the power inoput to the other components, rather than radiated noise from the output side.
Many VFDs have an input filter, some don't. Input filters are surprisingly cheap, and also easy to find on eBay, just search eBay for "rasmi".

And for encoders and resolvers I have 50m of a cable like this:
www.rapidonline.com/lappkabel-00277131-6...m-4-x-0-14mm-63-6881
Except the 3x2 + 2 + 4 version:
picasaweb.google.com/1081645046564043805...#6217838146609796818
It's a super complicated cable, the 3x2 cores are inside the overall sheath and braid, and then they have their own sheath, shield, drain wire and kapton layer.
I am actually planning to sell my spare 40+ metres of it.

Although the current in the output side is reasonably smooth due to the inductance of the motor windings, the voltage is certainly not. Each phase voltage varies from -ve DC bus to +ve DC bus at the PWM switching frequency (usually in the 2-10kHz range), with switching times of the order of 1us or faster. And when connected to single phase, the bus itself is wobbling up and down by 340Vdc at mains frequency, so the final waveform is complicated. You can pick up the edges easily in adjacent wiring, particularly if you have some means of accidental capacitive coupling, eg the cable is close to something sensitive. Screened wires on the motor connection will eliminate that tendency but there will be an induced voltage in the screen that needs to be conducted to ground at the VFD end of the screen. The motor needs to be firmly grounded too but the advice is usually to leave the motor end of the screen floating to avoid a ground loop. If you connect it at the motor end too and are unlucky, you may find that the motor screen develops a nice resonance. Yes, I've spent many hours doing EMC testing on drives, motors and PSUs....

The input filtering is required to meet the various mandated EMC standards which can be very challenging. If the VFD is sold unfiltered, it's a good idea to fit one as Andy says. The reason some are sold without is that the actual filtering required depends very much on the particular installation. I've just bought a very nice Yaskawa drive for my Milling machine and the supplied filter is a large fraction of the size of the drive itself...

Andy - I have a shopping trolley open in Rapid Electronics at the moment, full of drag cables, boot ferrules etc. If you are interesting in selling some (or all) of the fancy cable, I'd be very interested.

I crammed all my electronics (incl VFD) into a relatively small box on the side of my machine in order to save space and get the machine through doorways. I was very careful to keep the VFD and it's wiring as far from the rest as I could. When I first got things running I had what appeared to be bad noise issues and spent a bunch of time trying to diagnose and correct it. I ended up installing a Rasmi input filter (Andy's suggestion) and large ferrite beads on the output cable to the motor. Neither of these things fixed the particular problem I was having. I borrowed a friends spectrum analyzer to see if I could trace the noise to tie it to some event and while I could see plenty of noise and where it occurred in the cabinet it didn't help to find how to eliminate it. Ultimately I found the problem as I began to re-route various wires to provide further isolation and I discovered that the USB keyboard I was using was unusually susceptible to noise on its input wire and was generating spurious characters. Simply replacing the keyboard solved the problem.

I am not sorry I put the filters on the VFD, it did reduce noise overall according to what I could see on the spectrum analyzer and is suggested practice in general and also specifically shown in the automation direct manual for the VFD. I recall that the Rasmi filter and ferrite beads I found only added about $50 total to the build. But in the end it was other random issue that bit me. Just be as careful as you can with routing wiring, shield where appropriate, and you may be fine.

My last resort was going to be to move the VFD out of the electronics cabinet but it didn't come to that and the lathe has been working great.

-Tom