FIMER X 1600 Resistor help?

  1. eddie49 Member

    A MOSFET should show conduction like a diode in one direction from Source to Drain. A shorted one will be a short in both directions, and this can be measured in-circuit. However, if there are transistors in parallel both/all will appear to be shorted.
    The positive output of the bridge rectifier will connect to the positive terminals of both 680uF capacitors, and then go on to the inverter section. If you cut the track just after the capacitors, that may allow you to power-up and check for the 325v DC.
     
    DavidL likes this.
  2. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Thanks Eddie. My vac pump thingy came yesterday so I managed to get bits removed. So far Ive removed 2 of the heat sinks with some IGBT's and rectifiers and diodes. I have 4 dead IGBT's (some are good to prove procedure luckily) and a dead diode. One of the IGBT's actually got vey hot and burnt the surface of the 22R. So if overheating caused by fan not working caused the components to fail would this cause the shorting which blows the resistor?. I will carry on removing all of the bigger stuff to check out of circuit now im this far and clean as I go. IMG_20200829_164307.jpg IMG_20200829_152712.jpg IMG_20200829_164313.jpg
     
    DavidL likes this.
  3. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Eddie what tells you the DC is 325v? I didnt spot that anywhere.

    Cheers

    Tony
     
  4. DavidL

    DavidL Forum Supporter

    Messages:
    180
    Location:
    nr Wigan, Lancs - UK
    I don't know for sure that this explains it, but remember that 230V AC is the RMS over a sine wave - i.e. it peaks at 325V ish

    (As it's a sine wave we can work peak from RMS, so - 230 * sqrt(2) = 325 ish)

    This might help a bit on the AC voltage measurement and relationship to peak voltage, certainly more so than anything I could put together :)
    https://spark.iop.org/explaining-rms-voltage-and-current#gref

    If full-wave rectified with little loss and sufficiently chunky/appropriately designed smoothing capacitors 230V AC in rectifies to 325V DC.
     
    Hopefuldave likes this.
  5. eddie49 Member

    Regarding the expected 325v DC, the reason is as stated by DavidL. When measuring the AC mains voltage with a typical meter, the value seen is the Root Mean Square ( RMS ) voltage, also known as the "effective" or "heating equivalent" value of the sine wave. It is about the central 2/3rds of the waveform, omitting the peaks. To get the peak value, multiply the 230 RMS value by 1.414 ( which is the square root of 2 ), i.e. 325.
    After full-wave rectification, 230v RMS AC would show maybe 260v DC, but with smoothing capacitors you would expect to see the peak of 325v DC ( although with some ripple ).
    If the welder was running at high power levels without a fan, then yes, overheating could cause those components to fail short-circuit, although I would have expected there to be an overheat protection sensor on the heatsinks. Alternatively, maybe IGBT 1 failed short, and then caused the other three to fail.
    When powering-up with these shorts, the soft-start resistor is acting like a fuse in line with a short of the mains.
     
    DavidL likes this.
  6. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    David, that was quite interesting to read cheers. So apart from some losses, rectified voltage is about the same as the voltage was at AC?.

    Cheers

    Tony
     
  7. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Eddie, your reply arrived as I was typing the previous. Re overheat yes the heat sinks have what I suspected is a temp sensor too. 1 of the IGBT's was definately shorted while the others (so far) have been open or just not working. At each collection on the sinks there is also a rectifier (fesf type) and these are checking ok so far. There are also 2 smaller voltage regs (7812) for 12v, but what would be the likely input voltage for these? They are on the board rather than the sinks.
    The shorted IGBT has given me some hope now, thanks again :)
    Tony
     
  8. DavidL

    DavidL Forum Supporter

    Messages:
    180
    Location:
    nr Wigan, Lancs - UK
    If you're unfamiliar, it's worth having a look at the wiki article on rectification, which is pretty well put together:
    https://en.m.wikipedia.org/wiki/Rectifier
    (Just the single phase stuff and ignore the centre tap for now - in fact, just look at the stuff after the transformer)

    The rectifier is either "blocking" out the sub-zero half the wave (half wave) or full wave using a bridge, where the opposing diode pairs are reverse biased on either the positive or negative wave - giving two positive peaks per one of the input AC.

    More commonly, we'd think about a rectifier in a power supply incorporating a transformer to bring the voltage closer to our chosen working voltage, but as we've not been through a transformer, we've done nothing to change that.
    As such, that voltage is still the same, though as Eddie says, RMS average voltage is now up as we've now (ripply!!!) DC with the peak of the ripple at 2* the input AC frequency.

    Put really crudely, the smoothing capacitors charge at the peaks and and discharge to smooth out the troughs. The effectiveness and variance of the smoothing effect obviously depends on the load characteristics and the capacitor design/choice.

    Back to your welder, I absolutely think you're right to be hopeful having found the damaged IGBTs - have you managed to ID them/price them up?
     
  9. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Thanks David, that helps clear a few things up :). I fed 12v to the relay to the dc side this morning and the relay clicked and the power led lit so that was nice to see. Yes the IGBT's are G15N 60 I have started to look for them but the brands are various. Cant see Vishay or known brands yet. Are there some brands to avoid (ie cheaply made and unreliable)?

    Cheers

    Tony
     
    DavidL likes this.
  10. DavidL

    DavidL Forum Supporter

    Messages:
    180
    Location:
    nr Wigan, Lancs - UK
  11. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    I cant remember the brand but it was a proper logo, I'll check tomorrow. I'll take a look at that link thanks :)

    Tony
     
  12. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Are on semiconductor ok too?Dont mind waiting for half the price :)

    Tony
     
  13. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Its ok I realised the ON semi... is a 220 package so not what I need (clamp issues maybe)

    The Mouser site sells Vishay for £2 each so if I order £33 pounds worth of stock its free delivery, so I might stock up on some caps and resistors too. (I have an original Space Invaders that uses lots of bits :) )
     
    Hopefuldave likes this.
  14. DavidL

    DavidL Forum Supporter

    Messages:
    180
    Location:
    nr Wigan, Lancs - UK
    Good spot on the packaging!!

    Dumb question coming up....
    However, are they definitely the right ones?

    Disclaimer; I'm notoriously bad at navigating mouser, but at a quick glance the only vishay 15N60 ones I could find were MOSFETs (i.e. source, gate, drain) and I think the Infineon was an IGBT. Though I've just looked at 5 datasheets back to back so might be confused!
     
  15. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Ooh you put doubt in my mind now. It seems when searching some are described as diodes, IGBT or mosfet. I think there is confusion over the labelling of the pins. I might be wrong but drain and source is the "flow" once activated and the same for collector/emitter. The gate in both cases turns them on? This ebay link says IGBT transistor with g15n? https://www.ebay.co.uk/itm/G15N60-I...191752876109?_trksid=p2385738.m4383.l4275.c10
     
  16. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    On the circuit board its marked IGBT 1 , 2 etc and the parts that came out are G15N60 with GAD744 written under. Got me confused now :)
     
  17. eddie49 Member

    A 7812 needs an input voltage that is higher than the expected 12v output. The maximum input is 35v. Like a transformer-based MIG, I suppose that an inverter arc welder needs an auxiliary power source to control the high-power welding circuit. Something has to drive the 12v soft-start relay coil after the initial power-on delay. I'm not sure if this is also a switched-mode PSU running off the AC mains input, or stepped down from the 325v DC as the caps charge.
    Regarding the IGBTs, here's 5 pcs for 6 quid:
    https://www.ebay.co.uk/itm/233420129597
     
  18. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Thanks Eddie, are you happy that the part number G15...... is what I need as Ive not seen them described as IBGT and now David mentioned it and Im in doubt. ps if I get it running I'll check the input V on the 7812's.

    Cheers

    Tony
     
  19. DavidL

    DavidL Forum Supporter

    Messages:
    180
    Location:
    nr Wigan, Lancs - UK
    I'm honestly not sure if there's a meaningful difference in this application (specs appear very similar!), but I'm concerned about the manufacturer datasheets listing them as IGBT or MOSFETs, along with the price differential between the Infineon SGW15N60 (listed as an IGBT, but still marked G15N60) and the alternatives.
    Though to be fair, mouser lists the Infineon as obsolete with a possible supersession.

    Sorry, short answer from me is I'm not sure :)

    Though in fairness the ones Eddie found are nice price wise, so no great loss - time aside - if they go belly up or are not fit.

    Can you snap a pic of what you took out and see if any of us can recognise the logo?
     
  20. Tonyg64

    Tonyg64 Member

    Messages:
    42
    Location:
    UK Warks
    Ah I forgot to get a pic but it reminded me of a familiar logo. Its ST . I think I'll order the mouser ones labelled G15N.. I have other stuff to get anyway so if it goes bang we'll know they were wrong :). I wouldnt mind cheap if they were easy to get to but one of the set is hidden and the heat sink is soldered into the board with quite thick posts. Ive got about 15 tabs open from mouser so I'll get the shopping done and take a chance :)
    Cheers

    Tony
     
    DavidL likes this.
Advertisements