Mitsubishi compressor lock

Came to a unit today a Mitsubishi PUHZ P125VHA switched on at the controller but no cooling. Checked the filters wich were blocked but still no cooling. Looked at the controller again and i had a UF fault ( compressor lock)
Checked the condensor, the condensor fans would come on for 10 seconds then go off for 10 seconds then come on for 10 seconds then go offf again etc etc. Found the compressor stone cold. Disconnected the terminals to the compressor and checked the compressor for down to earth and also the resistance across the windings- both came back ok.
Decided to use my Daikin inverter checker. I connected the compressor leads to the checker then switched the outdoor unit on forced cooling. On the inverter checker 1,4 and 6 lit up for a few seconds then all 6 lit up for a second or so then1,4 and 6 lit up again then 3,4,5 and 6 lit up then finally all 6 lit up. I spoke to daikin about this and they said they dont know what the lights on the checker mean on a mitsubishi unit as they can only speak for daikin. So i rang mitsubishi up and they said they havent used that sort of inverter checker and couldnt advise what the lights mean???

I ended up measuring the voltage to each phase as it goes onto the compressor but with it disconnected and got 80volts for each phase, which would be enough to start the compressor as its an inverter but the compressor wont start.

I thought that because its been cold the liquid may have migrated back to the compressor and caused the compressor to lock due to liquid inside. I left power to the condensor but switched the controller off, which should give the compressor around 20v on each terminal and this acts as a crank case heater but i was getting 80v constantly instead of around 20 even with the controller off and the forced cooling dip switch back to normal.
Any ideas guys?

Could it be the compressor is locked as the fault code indicates? I mean the scroll is welded together!?

Hi marc.Did you get a clamp meter on it?,I have a daikin inverter checker and not found a mitsi system that it didnt work on. The red leds just visualises the output of the ipm.I believe as long as all 3 lights iluminate the same after start up the inverter pcb should be good,although once you hook up a comp to give it load the pcb can still be faulty!, sadly.Please correct me if im wrong though.

Quote:

---

Originally Posted by Sanderh

Could it be the compressor is locked as the fault code indicates? I mean the scroll is welded together!?

---

Thats a possibility but how would i know that? the windings arent down to earth and the resistance across them is fine.

Quote:

---

Originally Posted by stuartwking

Hi marc.Did you get a clamp meter on it?,I have a daikin inverter checker and not found a mitsi system that it didnt work on. The red leds just visualises the output of the ipm.I believe as long as all 3 lights iluminate the same after start up the inverter pcb should be good,although once you hook up a comp to give it load the pcb can still be faulty!, sadly.Please correct me if im wrong though.

---

I managed to get a clamp meter on it but didnt get any amps at all on any of the phases. My inverter checkers got 6 lights but im too well up on how it works, thats why i never use it.:confused:

Change the compressor everything is indicating that the scroll is mechanically locked. Have had a few of these and use the same tester. Change the compressor and they fire up no probs.

Is there anyway to test if a compressor is locked. Dont fancy ordering a compressor and find out its not that.

If the IPM is good and the compressor don't start then the compressor its stuffed. It's pretty common, I did three compressor changes a week from 2003-2005 when they first came out. Now I do all the ones that didn't fail back then, so only one per week, the first few vintages weren't very good, so I'm used to seeing failed compressors.

If the IPM is good and the compressor don't start then the compressor its stuffed.

I take it that this IPM technology is present on all modern Mitsubishi inverter systems. If so I would be grateful if someone could explain to me how you go about checking it works properly.

Also, what is a daikin inverter checker?

Quote:

---

Originally Posted by expat

If the IPM is good and the compressor don't start then the compressor its stuffed.

I take it that this IPM technology is present on all modern Mitsubishi inverter systems. If so I would be grateful if someone could explain to me how you go about checking it works properly.

Also, what is a daikin inverter checker?

---

Cant comment on the IPM but the inverter checker is connected to the leads on the compressor U,V,W and if the inverter board is ok then the lights will all flash together faster and faster until it looks like they are no longer flashing. If any of them dont light up then there is a problem with the inverter board.

Quote:

---

Originally Posted by Sanderh

Could it be the compressor is locked as the fault code indicates? I mean the scroll is welded together!?

---

What would cause the scroll to be locked/ welded together? Liquid slugging due to migration when its cold and the unit is switched off?

Reading between the lines it sounds like the gizzmo has six diodes on it that reacte with the voltage input by flashing faster and faster as theDC tension increases.

Am I getting it?

Without the experience that Paul H has I could only guess. You said liquid slugging, well if you imagine enough of that happening, and then the bearings holding the scrolls nice and tight becoming loose, the parts may not be running on their set courses. Mechanical parts running out of line will start by making rattling and nocking noises, but will ultimtely run into each other and the Waltz turns into a Break Dance.

Quote:

---

Originally Posted by expat

Reading between the lines it sounds like the gizzmo has six diodes on it that reacte with the voltage input by flashing faster and faster as theDC tension increases.

Am I getting it?

---

Yes thats about the jist of it.

if your gut tells you it's the compressor it usually is , from my experience inverters have an inbuilt safety device to protect themselves and shut down, thereby protecting themselves from any further damage. I have rarely found the inverter to be at fault , although this can be the case sometimes.

Thanks for the correction;)

Available from EDC International. Same thing, different label. They are very helpful, offered to ship one to Cyprus for me, but I actually got one when I was in UK recently.

Regards

Quote:

---

Originally Posted by marc5180

Cant comment on the IPM but the inverter checker is connected to the leads on the compressor U,V,W and if the inverter board is ok then the lights will all flash together faster and faster until it looks like they are no longer flashing. If any of them dont light up then there is a problem with the inverter board.

---

The IPM is what you are testing with the inverter checker. 380V DC goes into the IPM, three seperate phases of DC come out of it and into the compressor.
So the IPM does a lot of work and is the actual thing that drives the compressor. If diodes short out, (tested by mulitmeter on diode test, or which those edc/daikin testers), you lose a phase and the compressor won't run.
What causes a compressor to seize up? faulty manufacture or lack of oil. There's no oil separators on small inverters, most faulty compressors have the accumulator full of oil, so a lot is not returning (edit, so maybe it's not returning to the compressor due to low gas velocities when the compressor is running slowly).
It's either poor design or installation, (haven't made my mind up :) )
Maybe the designers haven't considered long term usage or done enough tests, maybe they should be designed to cycle off sooner, ie reach setpoint faster and switch off or not idle so long before hand.
Or maybe it's just poor installation, meaning the compressor can't run full speed, like it's cutting back on overheat (discharge temp, coil temps etc), so is always forced to run slow.

I have just uploading a short video to youtube of a daikin inverter checker being used on machines with Mitsubishi IPM modules, both working and not working. The 1st part of the video is an IPM from a working machine. The second part of the video is from a machine which had just had the compressor replaced as it was shaking and rattling and showing a rotor lock/position detection error. It turned out the compressor was OK and it was the IPM module that was not switching properly. Note how the LED's stay on at the end of the switching sequence.

http://youtube.com/watch?v=IN18g80RCX4

The connections onto the Daikin module were not quite colour for colour but it still works well. Mitsubishi output Red (U), White (V), Black (W) to Daikin connection point Blue, Yellow, Red.
The EDC one has prettier lights tho!!!

Quote:

---

Originally Posted by ozairman

I have just uploading a short video to youtube of a daikin inverter checker being used on machines with Mitsubishi IPM modules, both working and not working. The 1st part of the video is an IPM from a working machine. The second part of the video is from a machine which had just had the compressor replaced as it was shaking and rattling and showing a rotor lock/position detection error. It turned out the compressor was OK and it was the IPM module that was not switching properly. Note how the LED's stay on at the end of the switching sequence.

http://youtube.com/watch?v=IN18g80RCX4

The connections onto the Daikin module were not quite colour for colour but it still works well. Mitsubishi output Red (U), White (V), Black (W) to Daikin connection point Blue, Yellow, Red.
The EDC one has prettier lights tho!!!

---

Thats exectly the same inveter checker as mine and its the same sequence of lights as the one at the end of the clip ( the one thats showing fault). So do i get a new compressor, Inverter board or both to be on the safe side?
BTW what does IPM stand for?

Quote:

---

Originally Posted by marc5180

Thats exectly the same inveter checker as mine and its the same sequence of lights as the one at the end of the clip ( the one thats showing fault). So do i get a new compressor, Inverter board or both to be on the safe side?
BTW what does IPM stand for?

---

Intelligent Power Module.
So you get the flashing lights of incorrect order and they are staying on for a while before slowly dying out?
edit: also ozairman, on the unit that failed the LED tester, were all the diode readings OK both ways?
edit2: And was the IPM faulty, or the control PCB that times the IPM, or both?

What I'd like to know is why they sometimes call the inverter compressors "DC compressors" - when you're feeding them 3 live feeds, U/V/W - surely that just means that they're 3ph compressors?

I once had a 3ph motor and a single phase to 3ph inverter that I was going to utilise as a lathe motor but never got round to actually doing it, so sold the motor/inverter package on eBay.

Would it not be possible to use a single phase to 3ph inverter to test run the 3ph / "DC Compressors" in the field? (by-passing the a/c unit's driver altogether)

Just a thought :D

Quote:

---

Originally Posted by paul_h

Intelligent Power Module.
So you get the flashing lights of incorrect order and they are staying on for a while before slowly dying out??

---

I get ligts 1,4 and 6 come on thenall 6 come on then lights 1,4 and 6 come on again and then all 6 come on then and they stay on. The lights never go on and iv left it on for well over 2 mins.
If the IPM was faulty though surely i would have different voltage down each phase to the compressor but iv not, they are all 80volts.

That's what I was thinking in my edit, not the IPM faulty, as all that does is supply voltage, but the control board that tells the IPM when to pulse is out of whack in the example ozairman posted. Still he sees a lot and has only come across one which did that fault. And I personally have diagnosed heaps without that tester and haven't come across it.

I do agree that the compressor must be locked as per the fault code "UF" because the compressor is getting 80v on each phase but not starting. Maybe i was doing something wrong with the inverter checker but i thought it was supposed to be an easy way to check the inverter board. Maybe i should just put it back in its box again for another year:(

Edit: also i turned the controller off (as i said in a previous post) so the power to the compressor should have been around 20v to each phase and acting as a crank case heater and i was getting a constant 80v which it shouldnt have been getting which does lead me to beleive the main pcb that feeds the IPM is faulty?

No, I think the opposite, maybe you've come across the situation that ozairman posted, an IPM firing badly and the inverter checker picked it up because that can diagnose a faulty control signal, where normal offline multimeter checking can't.
I'm waiting for ozairmans response on this, since he seems to know these systems. I'm always open to new methods, and never come across what he's shown (compressor doesn't start, IPM looks OK = not compressor fault). Also the inverter checkers aren't common at all here, so I'm interested in iif they pay for themselves by doing something you can't do for yourself with a mulitmeter, megger and experience :)

Sort of replying to all above,
paul_h
The IPM module in the second part was definately at fault, on that machine the controller pcb is a separate board and as an experiment I tried using a controller pcb from a working machine to drive that IPM with the same defective switching pattern. Before the compressor had been replaced the IPM had been static checked and no open/or shorted transistors found. The "Christmas tree decoration" is a handy way of checking an output quickly when you have a compressor rotor location/lock issue and want to eliminate the IPM or Inverter PCB. But if the fault is some sort of powerup/lockout or breaker tripping fault then the static test of the IPM using diode check is the quickest way of checking for shorts and then possibly a compressor with a short to earth. Also a lot of the machines these days have the IPM and controller PCB as one big board so you end up changing both anyway. For fun I did try and unsolder a large IPM off a board a while back but even with a good quality soldering station it was near impossible due to the size of the pins and they are all using lead free solder these days which is a b!tch to rework.

electro
Why are they three phase DC compressors, because they are wound as three phase motors (as far as I know always in star). Inside the rotor of the compressor there are permanent magnets buried in the rotor instead of just stacked laminations with aluminium cast through them to create the poles like in a normal induction motor. This is supposed to give better efficiency and torque as there is less slip as the opposing magnetic force of the rotor is already there, rather than have to be created by induction.
A normal inverter for driving 3 phase AC motors would not work as most IPM's are matched to the compressors they are driving. The windings resistances of DC inverter compressors are also ULTRALOW. I looked up some info on this Mitsi unit (google is your friend!) and they are specced at 0.266 of an ohm. I think a normal inverter for driving a standard 3 phase motor would have apoplexy connected to a load like that!

Marc5180
How did you have the box hooked up which colour compressor wire to which terminal? Judgement on that basis means you have to have connected the box the same way. I have the daikin box connected Black, White, Red, top to bottom with the wires on the left side as you look at it. Not sure which is which but the leds down one side (eg 1,3,5) indicate an connection is being switched to the HV DC positive (P) and the leds on the other side (2,4,6) show a connection is being made to to the DC Ground (N). There should be one light on on EACH side at any one time as one leg of the compressor has to be connected to (P) and one to (N). This all happens rather fast (which is why if you measure across phases it reads as AC) so I normally look for the pattern of one side of leds flashing in one direction and the other side flashing in the opposite direction.

Also did the compressor make any noise at all? normally you will get some sort of noise out them whether it's a clunk or a shake or a rattling sound even if they are seized. Can often be described a "morse code" or "typewriter noise"

I would suspect an inverter fault from what you have said, when you say the compressor resistance were OK what do you mean? from above I think the winding resistances are 0.266 ohm for this compressor unless you have a mill-iohm meter you cannot read them accurately enough to detect winding imbalance.

Have you tried to do a static test of the IPM module with the power off and the connections unplugged? Set you multimeter to Diode Test then check between P-> U, V, W in forward and reverse direction and then N-> U, V, W in forward and reverse direction, you should get 3 sets of readings something like 0.35-0.5v/OL, 0.35-0.5v/OL, 0.35-0.5v/OL from the P-> UVW test and the same from the N->UVW test.

Quote:

---

Originally Posted by ozairman

Marc5180
How did you have the box hooked up which colour compressor wire to which terminal? Judgement on that basis means you have to have connected the box the same way. I have the daikin box connected Black, White, Red, top to bottom with the wires on the left side as you look at it. Not sure which is which but the leds down one side (eg 1,3,5) indicate an connection is being switched to the HV DC positive (P) and the leds on the other side (2,4,6) show a connection is being made to to the DC Ground (N). There should be one light on on EACH side at any one time as one leg of the compressor has to be connected to (P) and one to (N). This all happens rather fast (which is why if you measure across phases it reads as AC) so I normally look for the pattern of one side of leds flashing in one direction and the other side flashing in the opposite direction.

Also did the compressor make any noise at all? normally you will get some sort of noise out them whether it's a clunk or a shake or a rattling sound even if they are seized. Can often be described a "morse code" or "typewriter noise"

I would suspect an inverter fault from what you have said, when you say the compressor resistance were OK what do you mean? from above I think the winding resistances are 0.266 ohm for this compressor unless you have a mill-iohm meter you cannot read them accurately enough to detect winding imbalance.

Have you tried to do a static test of the IPM module with the power off and the connections unplugged? Set you multimeter to Diode Test then check between P-> U, V, W in forward and reverse direction and then N-> U, V, W in forward and reverse direction, you should get 3 sets of readings something like 0.35-0.5v/OL, 0.35-0.5v/OL, 0.35-0.5v/OL from the P-> UVW test and the same from the N->UVW test.

---

I had the inverter checker connected from left to right starting with the red then white then black.
The compressor didnt make any noise at all or give an indication that it was trying to start.
I havent done a static test because i didnt know how to do it when you say "P-> U, V, W in forward and reverse" what does the P mean and also, forward and reverse??

Quote:

---

Originally Posted by marc5180

I had the inverter checker connected from left to right starting with the red then white then black.
The compressor didnt make any noise at all or give an indication that it was trying to start.
I havent done a static test because i didnt know how to do it when you say "P-> U, V, W in forward and reverse" what does the P mean and also, forward and reverse??

---

Your connections were different to mine, that would make a difference to the flash sequence you are seeing.

As for the static test I will post a little drawing tonight and a couple of IPM pics, but here is a 'brief' explanation. You should see the P N UVW markings on the silk screen of the pcb near IPM module, Or on the circuit board diagrams if you can get them. The IPM normally has two screws in it holding it down about 1" to 1.5" apart and will be mounted to a large heatsink.

A transistor is like a diode as it will pass current in one direction not the other so can be roughly tested using diode check on your multimeter. The IPM is just a big module with 6 ultra fast switching IGBT transistors inside it that connect High voltage DC (P) to UVW and DC Ground (N) to the UVW plus some other fancy electronics to monitor current and temperature and the compressor rotation feedback signal.

With the machine isolated and the IPM disconnected set your multimeter on diode test mode. Do the following put your Red lead to the N terminal and then with your Black lead check to U, V, W you should get a reading normally between 0.3-0.5v on your meter at each point so note them down. Then reverse the connections start with Black lead on N terminal and check to UVW with your Red lead, you should see an 'open' or 'ol' reading indicating an open circuit. So if there were no shorts or opens should now have three sets of readings say 0.42/ol, 0.41/ol, 0.42/ol. That has just tested the three transistors connecting the N terminal to the UVW.

To complete the test you do the same thing from the P terminal testing the three transistors connecting P to the UVW terminals again doing it in "both directions" starting with the Black lead to P and Red to UVW then your Red lead to P and Black lead to UVW. You should get the same set of three readings from this side as well say say 0.42/ol, 0.41/ol, 0.42/ol. Any readings like 0.1/0.1 or the continuity 'beep' in both directions indicate a short. An open (ol)/(ol) is also not a good sign.

Note if the IPM is soldered to the main board with the filter capacitors on it then it may take a while to show the "ol" reading as the capacitors charge up.

Have a look at this page for some info on transistor and diode testing
http://www.elexp.com/t_test.htm
Understanding how and where to use it is a very helpful tool for checking out inverters, as it can be used to test diode bridges as well and even DC fan motors which are becoming pretty common these days.

Just an update guys, i didnt get chance to go back to the unit and figure out the problem because as it was quite a new install it was under warranty and therefore the installers where called back to fix it:(