Friday, May 15, 2015

Sunkko 788+ Welder failure and repair

I have a Sunkko 788+ Spot Welder.  It failed on me rather quickly, either from a faulty triac, or from overheating the triac due to an aggressive duty cycle (I'd say I exceeded the rated duty cycle, but that would imply there was enough documentation to determine this, which... not so much).

The quality customer support on eBay was at least nice enough to suggest I replace the triac.  They did imply that this was about the only support I was going to get, so... well, in I go.

This might be part of the problem...



Failure Behavior

I was welding away at a battery pack, happily popping nickel strip onto batteries.  I stepped away for a bit to get the next part of the pack assembled, and when I came back, the unit wouldn't power on.  Well, I'm smart enough to know that if a unit comes with spare fuses in the box, blowing a fuse is somewhat common.  Annoying, but common enough.  I removed the fuse holder, dumped out the shattered remains of a fuse (it apparently crumbled when the holder was turned), and replaced the fuse.  Excellent.

I plug it back in, hit the power switch, and... BANG.  Something inside the unit fails, hard.

The unit still powers on, but it won't weld anymore.

https://www.youtube.com/watch?v=Pt8RJ3pCrlM shows it's behavior after the failure, if anyone is curious.

Well.  Time to dig in. Read on for the photos.


Having been told that my options from the seller were "Repair it," I decided to repair it.  Also, I was planning to repair it anyway, because I'm curious about things like this.

The case comes off easily enough, to show the transformer and the backside of the circuit board.


The front comes off easily enough.  An inspection showed some pulled traces - this is, as expected, truly a thrown together piece of work.


A whole lot of messing around later, I had the board out.  The other side had what appeared to be the actual problem: A large hole blown in the main switching triac.


The plastic front of the unit where this thing blew also gave evidence of the failure.


Well.  I don't know if this is going to fix the problem or not, but it's sure worth trying!  I mean, I just wanted to rebuild a battery pack, but a spot welder is almost the same, right?

A bit of research, an eBay order, and a week or so of waiting around provided me with new triacs.


While I was waiting for the parts to arrive, I discussed the failure with some people online.  One of the things that came up in conversation was the location of the triac.  It has a heatsink, but the triac and heatsink are in the very small amount of air between the circuit board and the front of the welder.  Further, it can't even get any good convection going because of the power switches right above it.

There's a giant amount of air space behind the circuit board - it would be nice to relocate the triac into that space, so it could cool better.

I sat down with the unit and figured out that, in fact, I could do just this!  The triac is mounted with through-board holes, and if I just flip the triac around, I can hang it into the open air.  This can't cool any worse than up where it was, and while it might not be strong enough to ship the unit that way, I'm not planning to ship it!  I'm planning to use it.

The Repair and Re-engineering


As far as repairs goes, this one is pretty simple.  Remove the old triac (I just cut the legs to make it easier), move the heatsink onto the new one, resolder it out the backside of the board.



I was genuinely surprised to find thermal compound between the triac and the heatsink.  I was expecting it to be dry, or at best, some thermal tape.


You can see some of the high quality soldering around the board.  At least they thickened up the traces for the 110 current... I do suspect this may be the only change from the 220v version to the 110v version other than a transformer winding difference.


The new triac was thermal compounded and attached to the heatsink before being mounted, facing the other way, very careful to ensure the pins remained where they were supposed to be.



The final step was to reattach the voltage sense lines (?) from the transformer before putting things back together and testing.





A bit more assembly, and success is mine!  The welder works, and the weld is solid!



Analysis

So, what happened?  Obviously the triac blew up.  What caused that, I can only speculate about.

It's important to recognize that this welder seems to be well enough liked in China and other countries.  So I don't think there's a fundamental design flaw.  However, this welder normally comes in a 220v version.  There are 110v versions, and I suspect the difference is mostly or entirely in the transformer winding.  I don't have a 220v version to pull apart and test, but as I've not seen any dual mode versions and the wall current goes straight to the transformer, I'd expect a factor of two difference on one of the sides.

However, this means that the 110v version is taking ~2x the current on the input side for the same output power.  And, all the input current goes right through the triac.

Given the lack of documentation, it's entirely possible I blew right past whatever duty cycle rating there was, overheated the triac, and caused it to blow up.

Or it might have just been a bad triac.

In any case, I've got better cooling on it now (I may add a fan), I'll take it easy on it, and I have a good understanding of what's in here.  And, plenty of spare triacs.

46 comments:

  1. Nicely done. i've seen this sort of 220v -> 120v issue in computer PSUs from the early 00's when auto-switching models became more commonplace. half expected you to install a massive heat sink. how is the chassis ventilated?

    ReplyDelete
  2. Poorly. It has a few slits cut in each side and no forced airflow at all. I'll try to update the other blog post with a side shot.

    ReplyDelete
  3. Good stuff. I'm encouraged that my single completed weld 709A might be salvageable. They are finally sending a new PCB. It completed one weld and POOF. Sadly I'm an electronics n00b with no local playmates interested in my dilemma. I remain hopeful. Kinda of an expensive footstool otherwise. Thanks for posting the story!

    ReplyDelete
    Replies
    1. I'm just assuming that if one plays in the realm of cheap Chinese equipment, one must be prepared to fix cheap Chinese equipment. It's an endless rabbit hole.

      Delete
  4. Just curious if you think adding a pc fan in the unit would help much in keeping the unit running cooler?

    ReplyDelete
    Replies
    1. I think it would, and I've considered it, but I simply haven't needed to so far. I go easy on the unit now, and let it cool - I'm not running the welds as fast as I can anymore. If the triac were getting hot, airflow would help, but just having it in open air instead of crammed on the front with no circulation seems to be enough.

      Delete
  5. Anyone need a 709A and new PCB? Mine quit. They replaced the board but the repair is beyond my scope. 110V version.

    ReplyDelete
    Replies
    1. Get in touch with me - rgraves@sevarg.net works. If you're in the US and willing to ship, I'll take it.

      Delete
    2. i will take it and pay shipping ?

      Delete
  6. This comment has been removed by the author.

    ReplyDelete
  7. Russell, My 788 quit working tonight, although the symptoms are a little different than your experience. I have symptoms such as there is no LCD display showing current and pulses. I haven't opened up the front part of the case yet, but it might be more than just a triac. Any suggestions on trouble shoot the 788 when there is no display, but the LED lights mounted on the welding arms work, fuse is okay, and the battery charger side works fine!

    I deleted the original comment because it's late at night and after re-reading the original comment it was on the incoherent side. Sorry about that.

    ReplyDelete
  8. My 787A start trouble some times. LCD start blink and then off. Need power on /off that come back.
    If this happened Its not work. Can some have this problem ?

    ReplyDelete
    Replies
    1. Hi.
      My 787A+ acted like yours.
      Please, let me know of your procedures in order to solve it.
      robinfosus@gmail.com

      Delete
  9. do u think we can replace the Transformer with one from a 1100W microwave for more current ? for thicker nickel strip? i have a 220V version ( australia) basically make use of the timer function to power a bigger transformer ... what do u think ?

    ReplyDelete
    Replies
    1. If you're going to do this, separate out the current flow to the transformer. I don't think the PCB will take the increased current flow for long, and you might want a more robust triac.

      Delete
  10. Holy Cow Batman you have just ordered this crap, so why are you reading this, are you a bloody masochist.

    So I am in the UK looked for Triacs on eBay to no avail any help please for the BCR30AM12L.

    Also any problems from UK users or am I being nervous and inventing problems, hopefully being blessed with this delivery Friday, as you guy's seem to think 110v's the problem (along with very short duty cycle).

    And Chinglish or a lack of instructions.

    Regards David

    ReplyDelete
    Replies
    1. Any of the BRC30AM series triacs should be fine. You may have a different model - I doubt they're consistent for long on the scrap triacs they throw in.

      The only real problem I've heard about with the 220v units is that they're really rough on circuit breakers as far as surge currents go. You may have to put in a slower tripping breaker if it pops the breaker as soon as you power it on.

      Delete
  11. Thanks Russell, as it happens....
    I am now working in Electrical retail (I loved Teachibg but hate kids (17YO's, Mine are fine it's the others)) so no problem with MCB B.C, ? curve whatever.
    As for the triac I will keep looking, would you like any shots of the inside of the UK one when it arrives.

    PS we work on a 30amp ring main here, but we are in an old Victorian house so mains probably to be desired and will be a spur off the ring.

    Sorry to go on Ta regards David

    ReplyDelete
  12. Also What voltage and amperage (600v 40A) for the Triac.

    BRC30AM = 30amps 400-600volts.

    I found this:- http://www.farnell.com/datasheets/1689729.pdf

    TOP2 is what it looks like you have but the RD91 (insulated) at twice the price if gates correct or will do.

    Interesting in the data sheet T=25Deg C, cooling fan mod very important.

    Your thoughts please Russell.

    Kind regards David

    ReplyDelete
    Replies
    1. I really don't know. I found some identical replacements.

      I may have a few spares somewhere if you can't find one.

      Delete
  13. Thank you Russell I have sourced them on the Bay from China at a tenner for five.
    But for now, I have on order (from up north) a 47 Ohm Resistor and a SPST with Missile cover Switch, Ohm's Law @ 240 volts.

    You got that Scope yet, or the time...

    I really think it will work via a 13 amp fused UK plug with the suggested Switch Modification (5 amps).

    Kind regards David

    ReplyDelete
    Replies
    1. I have a scope, but it's an old analog one.

      Time? Hah. Sadly, no. Shed building and such has been consuming all my time lately.

      Delete
    2. Hello David,

      my unit kept triggering the circuit breakers (on 220V) and recently blew the triac as well when turning on.

      Just wondering, could you please elaborate what modification you're planning with the resistor and the switch?

      Thanks!

      Clem

      Delete
  14. Ditto on Scope x2, both need switches/Pots cleaning/replacing. Ditto Time.
    Not ordered the Triacs yet, I have ordered Copper Rod though if you need any PM me, hopefully 400mm lengths x 4. I do have a Lathe/Mill again not working/no home for it, (Any Masochists in Kent UK).

    Thank you, regards David

    ReplyDelete
  15. Dear John (Unknown), I am about to start and I will first add a switch with a resistor bridging it with a missile cover for added security (SPST, 47R Axial 5W Ceramic Resistor 5%).

    Hope that helps, PM me an address so I can post you a resistor.

    I hope to start today but it is time, I also have a PCU heat sink and fan how I go about that is Top Secret (Buggered if I know but I will when done).

    Kind regards David

    ReplyDelete
    Replies
    1. David -

      I'd be very interested in some photos of your improvements, and if you're willing to share them, I'd be happy to put up a post about your tweaks!

      Delete
  16. Well trying to fix my 788+. I powered it on. It sat for a few minutes not being used when out of the blue it buzzed as though it was welding, then blew the fuse. I took it apart and I don't see anything obvious like a smoked component. I put an ohm meter on the triac to look for a short and found none. Tried a fresh fuse and the unit buzzed and blew the fuse as soon as the welder was powered up. Any suggestions? John johnmduprey@gmail.com

    ReplyDelete
    Replies
    1. Check continuity through the foot pedal plug. I understand they can go bad and cause problems.

      Delete
    2. Ordered a couple components to try replacing: both the BCR30AM triac and the MIP0225SY MOS/IC. Since is was blowing the fuse I figure it's probably a component that carries some current and those are the ones that stand out due to attached heat sinks.

      Delete
    3. This comment has been removed by the author.

      Delete
    4. Hi i am using 788 220v . It keep tripping my home circuit breaker. I tried on 20a circuit breaker as this is the max for home. Anyway to tweak this to make it works without tempered or change my home circuit breaker. As this change are not allowed for home

      Delete
    5. Jim Newson,

      where did you find the 220v version?

      Delete
  17. This comment has been removed by the author.

    ReplyDelete
  18. I blew the heck out of my 788H a couple of hours ago. Hence I'm now on google and I found this page. My 220V 788H tripped breakers. So I bought a microwave oven transformer (welling brand), and replaced the secondary winding with 2 loops of four-gage wire. I got only 1.6V AC from the 2 loop winding. The power into the spot weld was weak.... too weak to weld strips together. It was heating, but very very weak. Then, I turned off both switches (weld switch and power switch. But later, I turned on the main power (red switch) and didn't give enough time for the LCD panel to start up properly before switching on the green weld power switch. When this happened, the LCD panel lights went off.... and stayed off. I then turned both switches off. Then pressed the red main switch to turn the unit on.... but still no lights at all. No life or activity. Then, as I was switching the red switch to the OFF position (yes... switching to OFF), big flash of light came out from the board (as I had the top lid opened for testing). The breaker outside the house switched off. I see black soot on the board near a small IC somewhere below the molex-like white connector (with the power wires and the thermal fuse wires). And I also see spray/splay marks (on the internal floor of the 788H case) that look like dirty water lines, but it's dry...... could have been ejection of plastic or whatever. I'll see if I can fix whatever is broken on the board. I decided to order a second 788H from ebay (in case the one I just blew is truly history). My next modification will be more turns on the secondary, such as 4 turns. This is to increase the AC voltage to at least 4 Volt AC. One or two turns on the secondary will certainly not work. Waste of time. However, in order to get more turns, it will be necessary to either use lower gage number wire, or to use 'flat' (strap-like) wire. Otherwise, it's impossible to get more than a couple of winding turns into the transformer if large diameter wire is used. The 'nice' thing is, when I was testing the 788H with my new modified transformer, the breakers did not trip at all. I suspect that the tripping of 220V 788H models is due to the original (stock) transformer. I suspect this ... because I have previously dismantled the stock 788H (E-I lamination) transformer, in order to reduce the number of secondary turns to '1'. Same issue --- current too weak. But, just turning on weld power every 5 or 6 times led to breaker trip. And this is only starting up the machine in preparation for welding. Just turning the green light switch to 'on' would sometimes trip the breaker. However, I believe that I killed the board because I turned on the green switch too hastily without waiting for the circuit board to properly start up. So, if I do fix the machine, then I'll make sure to wait 10 seconds (after turning on the red switch) before turning on the green switch.

    ReplyDelete
  19. This comment has been removed by the author.

    ReplyDelete
  20. Back again! After a frustrating time with removing the 788H circuit board from the front panel, I found two confirmed destroyed components. One of the components that exploded apart (with a piece that got ejected off during the flash) was what-appears to be a surface mount bridge rectifier IC (4 legs. This IC is near the molex-like 4-pin power (and thermal fuse) plug. The other component that got destroyed was on the other side of the board, and that component is a thermistor, NTC 5D something. I think I need to wait for my new 788H to arrive in order to see if I can get some make/model numbers of these destroyed components. Even if I replace them, obviously no guarantee that the board could come back to a working condition. I'll see how it goes.

    Now, as for the 788H circuit board, removing the board from the front panel was a pain in the neck. This is because the sockets/grommets etc for the battery charging terminals appear to be permanently fixed. Regardless of whether those grommets were permanently fixed or not, I simply used a pair of cutters to cut out the grommets. Doing this allowed the circuit board to become separated from the front panel. In my opinion, the 788H is problematic. But I'm sure everything will be ok in the end, once all the required modifications are made.

    ReplyDelete
  21. Update : just noticed the label "DB207" on the circuit board. This matches the broken (exploded/shattered) surface mount bridge rectifier part. The other component, that blew apart is a 5D-9 thermistor. I've now ordered some of these, and will see what happens if I replace the blown parts on my circuit board. Fingers crossed.

    ReplyDelete
  22. Correcting an earlier error. Earlier, I wrote "it will be necessary to either use lower gage number wire", but I should have meant 'higher gage number wire', since higher gage means thinner cable diameter. And I need thinner cable in order to get more turns into the limited cross sectional area of the secondary winding slot of the transformer.

    Today, I managed to cram 6 turns of 8 AWG wire into my Welling brand microwave oven transformer. That provided 5V AC on the secondary side. Once my new 788H arrives, I will put this transformer into the new machine. My blown 788H seems to have 'missing' surface mount resistors. I'm thinking that they got vaporized during the major failure event.

    ReplyDelete
  23. Nice Blog Post,!!!!.... Thanks For Sharing Info... Time for a new Battery or maybe your not sure? AC Auto can use our specialised testing equipment and give a printout of the condition of your battery.We have many options 7 Brands available for all budgets. Call us today (03) 338 4653 and Click Link For More Info.. Batteries repair service

    ReplyDelete
  24. mine was brand new, same effect was seen however in my case it was NOT the triac it was the primary of the transformer. Solder joints to the board were very weak , then resoldered wires direcly and that solved the problem.

    ReplyDelete
  25. Porras,

    got pics? Where exactly are these joints? did you have to take the circuit board apart from the front cover?

    ReplyDelete
  26. Hello to all. I'm new here. I bought the Sunkko 709A that has failed for up to 3 months. In practice it turns on but not firm. Someone has already solved this problem. Do you know if the electric diagram is available?

    ReplyDelete
    Replies
    1. is it not welding? the op of this thread has found the fix. is your triac working?

      Delete
  27. Russell,
    What settings do you use on the 778+

    ReplyDelete
    Replies
    1. My settings won't be useful to you because I am compensating for some inverter sag.

      On grid power, I believed I used "2 70" with the current set knob turned down pretty far and it worked fine, but you'll have to play with your system - things change for different strip thicknesses and different alloys.

      Delete
  28. IC 7 ist defekt
    Bezeichnung?

    ReplyDelete