I started to “upgrade” my MPCNC to a low rider v3 this summer to start cutting really big sheets of plywood. Unfortunately, I have not been able to cut out the strudplates yet, as always something else comes up - now it is the vfd of my water cooled china spindle - “Model A2-8015M”,
I did not do any changes to the spindle set-up - though I have not cut anything, I had the spindle spinning at 24000 rpm for some aircuts - another issue.
next I see an error on the vnd display when starting it up - did not manage to note it down - and as I was not sure if the spindle was turning, I hit the stop button on the vfd control unit and tried again. This time the fuse of the powerplug went of (but not the one of the vfd)
because I did forget to run the water cooling for a short period of time a few sessions before where the spindle was getting quite hot (but no smoke) I thought I need to check it before doing anything else. (but I had it running for a few minutes with cooling on during the “cutting air” session.)
when disassembeling, I thought it did not turn freely and was fearing a bearing issue, but later when I had it in hand there was no issue with it being turned by hand and I could hear no unnormal sound when shaking it.
measuring the resistance of the coils against each other showed a uniform picture, something like 3,x ohm. So I think it is fine
So I thought, lets give it a shot again - once I switched the power on, the vfd smoked a little and this yellow tube burned off - obviously I switched it off again.
The vfd is ~70 € - so I simply ordered a new one - but I still like to understand what might have cause the issue - and be sure the next one is not also going up in smoke. And I want to try to fix it…
I am having the machine set-up in a small garden shed without any heating and when the incident occured, it was around 0° C (32 F). Earlier I also found some earwigs in the cooling water - so the vfd might have had some interaction with them as well - though I did not find any traces indicating this.
Anyone any thoughts on the root cause of this and what part it is that burned of so that I can try to replace it?
Never seen the insides of the chinese VFD:s before. Based on the size & shape looks like a pre- or discharge resistor of some kind.
Be careful with these, since unplugged doesn’t mean safe due to the large DC-link and other capacitors. There should be discharge function built in, but you never know.
What is the function of the terminal the blown part is connected to? Is it the spindle power supply line? Its kind of strange that the fuse did not blow first so maybe the fuse itself is out of spec and should be replaced.
I would remove the blown part and see if there is any writing on it. Also there should be an indicator of what type of part it is on the silkscreen, such as R20.
The three terminals are labelled R, S and T, they are the 3 phase input terminals, when using single phase you just connect live and neutral to R and T. The yellow device is a 75R high wattage resistor labelled R39. The right side of the blown resistor is connected to the left of the fuse mounted to its right however I cannot find anything that is connected to the left side of the blown resistor, not even the seriously sized capacitors on the reverse of the board. I am not willing to dis-assemble mine further as there is a bank of heavy mosfets (I think) on the underside and I don’t want to disturb them.
The specs have an environment temperature range of -10 to +40 degrees C… but nothing about invading insects
My 1.5kW spindle has 2R5 windings, so within error tolerances of your findings, of course you know you didn’t do it any favours running it without cooling, the manual does go to pains to make out the damage caused to the bearings when run too hot.
Mine is not the ‘M’ variant, but I think that refers to whether or not the control panel has a wire connector or a plug and socket.
If it’s the precharge resistor, it would be dissipating that heat only for a very short period of time and then get bypassed by the big relay in behind it in the picture. But even then the 160W sound like a lot.
Thanks guys for your thoughts - especially to you Mike to actually open yours up to take a look. I don’t know much about electronics and I don’t have the ambition to understand how my vfd works in detail - but I will try to order a replacement resistor and see if soldering it there will make it work again - and give you an update.
But if it was some sort of start-up resistor that burned off - it must have been over-loaded - so that would mean there must be something else wrong, besides the fuse not going off. Could that have been caused by a short cuircuit in the spindle motor - I suppose when I am measuring the resitance of the windings - I am probably using a much lower voltage then when it is connected to the VFD - so maybe the insolation is not completly fried off but worn to a degree?
I guess I will figure this out latest when the next VFD failes also on start-up…but what do you think?
As temperature increases, the resistance of the motor windings also increases. One possibility is there was some type of thermal runaway condition that was caused by the controller continuing to increase the motor current to the point where the resistor failed but not enough current to blow the fuse.
Likewise with stepper motors, thermal resistance in the stepper driver and/or motor windings could also be a cause of losing steps. The increased resistance causes a voltage drop to the motor which reduces torque. I have noted the phenomenon where the stepper motor is functioning normally for about 15 seconds and then starts losing steps. After adding a large blower fan to cool the drivers, the motors will either function normally throughout the job, or lose steps immediately at the beginning of the job if overloaded.
Did you go through the VFD settings to make sure everything will work with your spindle before operating it?
VFD drives have a fair share of important configuration settings that generally have to be tailored to each spindle. This includes stuff like spindle frequency, current limits, brake/coast setup, accelleration times, and more. So in general, defaults aren’t going to work that well, and worst case can damage parts. Even if you buy the spindle and drive from the same place in a combo kit, a lot of the overseas vendors won’t preconfigure the drive to match the spindle. It is an AC drive, so things can get very inefficient/wrong if the incorrect configs are used, and often problems will physically mask themselves to an inexperienced user (not obvious something is off if it spins and speeds up/slows down on command, but that’s not all that is involved with properly driving a VFD). Caveat emptor… always verify VFD settings before the first spin up!
My Huanyang VFD came with a comprehensive manual describing the sorta large menu of config settings available. I got a spindle/driver setup that others have gone through and tuned up before I purchased. So I mostly followed their setup advice. To give an idea… here is one of the pages I read before spin up:
As you can see there’s a lot of PDXXX settings to go through, and a few of them could result in a fried driver and/or spindle if you have them setup wrong!
Yes - I have not verified the VFD settings - but I already bought the spindle + the VFD about 3 years ago and used it since with the pre-set settings - cutting aluminum and destroying lots of bits - but not super frequently - nevertheless, I will remember to do this going forward :).
short update as promised. I replaced the burned part with a 75 ohm resistor with no effect but the sound of something else failing when plugging in the VFD - I did not investigate further.
With the new VFD I checked the settings and found that I did not need to change anything really. Turned out the error before the failure of the first VFD persisted - so I thought it must be the spindel - might have had some negative effect that I once forgot to plug in the cooling - so I ordered a new spindle.
Once the new spindle arrived, the error persisted - and once measuring the contacting cables from the VFD to the spindle I found that one phase had no connection - the soldering in the plug on the spindle top was bad - honestly, I made a huge mess out of soldering all 4 cables into the connectors of the plug - but once I reattached it - pretty much as agly soldering as before - the error vanished.
So why did the VFD fail? I think I was starting to hammer on the stop/start button of the VFD when it showed me the OC-u error during speed up of the spindle - “overcurrent during acceleration” - so that I started the spindle again before it came to a complete stand still - I assume that caused the resistor to burn - without having done any more investigation on it.
The last issue I found - did I tell you I was cutting thin air? - was that my z-axis lead screws were having the wrong pitch - 2 instead of 8 - so my z-movement distance was always divided by 4 - I took the easy way out and simply ordered two new lead screws with pitch 8 mm… that solved it! - before I did that I actually changed the settings using the display control to change the steps per mm ratio - but I think 3200 steps per mm were simply causing my z-motors to loose steps as I could see that the z-axis did not clear the surface of the work piece everytime completly when cutting the strudplates leaving marks - setting the steps per mm back to 800 (default) and having the new lead screws solved it for me - so almost done now!
- another issue.