immersun repair no.2 - repost

[Marcus]

dug out the notes i did for the post on immersun repair on the old forum - as there seems to be interest ATM.

There's no schematic per se, but an overview on how it works and the repair of the particular fault on this immersun.

OK it’s worth explaining how the immersun works in broad strokes so if someone’s looking to repair one they don’t have to start from scratch. here’s a rough block diagram showing the path the a.c. takes from input to output. The choke/inductors are shown but I’ve ommitted the smaller capacitors between live and neutral and earth.



The Immersun sort of works like a step down switchmode converter with synchronous rectification, except that it works with a.c. rather than d.c. . To divert power to the immersion it turns on the high side switch with a variable duty cycle, passing current through the 2nd inductor. When the high side switch is off the low side switch is turned on (and vica versa of course) to allow the current to continue to flow through the inductor to the immersion. Each switch is two back-to-back IGBTs.

[Marcus]

a couple of pics to help follow the layout on the actual PCB(NB the topside pic showing the transformers caps and inductors is of the later version immersun with two 15v (small brown) transformers - the earlier version has just one.

[Marcus]

The fault:-

the fault with this immersun was all the IGBT’s (Q1-Q4) were blown, along with all the semiconductor components of the high side switch gate drive (U1, Q7,Q8, Q9, Q10, D2, D3), a couple of resistors (R13, R9), a zenner diode (D8) in the isolated 15v supply, and the 0.5va, 2x15v gate drive supply transformer (parts list to follow).

It’s often difficult when you have a trail of fried components to be certain which one was the 1st ‘domino’ to fall – or to know why. I do however have a theory:-

The gate drives of high side switch and the low side switch each have their own isolated 15v d.c. power supply – this is needed as one operates at the 230v a.c. live side voltage whilst the other operates at the neutral side voltage. In these V2.50.2 immersuns these 15v supplies are provided by the small brown 0.5va vigotronix transformer, which has two ‘isolated’ 15v output windings. I believe this is a design flaw because whilst the transformer has a split bobbin design to provide good isolation between the 230v primary and the 15v secondaries, there’s no such isolation between the two secondaries, just the enamel on the wire and possibly a bit of transformer tape.

My theory is that the insulation between the two 15v secondaries failed, creating a short between the live side gate drive cct and the neutral side gate drive - leating to catastrophic failure of the semiconductors.

This theory is borne out by the fact that the secondaries of the transformer were open circuit, and failed a 250v insulation test using the megger between them. The final damming evidence is that in the later V3.08.2 version of the immersun, there are two separate 0.5va transformers mounted on a small sub-pcb as a retrofit to the main pcb (which is still designed for a single transformer), providing two properly isolated 15v supplies. This suggests to me that immersun are aware that the single transformer design is flawed, and introduced this modification for that reason.

pic of the new two-transformer modification - it's a snug fit.

[Marcus]

Identifying components and ordering replacements:

For some of the larger devices like the IGBTs it’s simply a matter of reading the id off the front and looking it up with a component supplier. For some others, like the varistor, you have to find the manufacturers datasheet and cross reference the marks on the device to the ‘part ordering code’ on the datasheet, then take that to the part suppliers

For the small SMD components it’s a bit more of a problem: the small 6leg devices on the PCB (Q7, Q8, Q9 and Q10) were all toast, as were D2 and D3, but these small devices have simple codes that are not unique – even if you know the make there’s no guarantee the manufacturer won’t reuse an identifier for different devices. In the end you have to look for devices that are physically the same and have the same markings then check the datasheet to see if they are a device that will work in the circuit.

For the diodes D2 & D3 (marked RPW 3d) this was fairly easy to guess they were back-to-back zenners, and if you ended up with something slightly different it wouldn’t matter too much, but these devices seem to be the right ones, RP designating 20v, W = made in china, and 3d indicated manufacturing date.
http://www.farnell.com/datasheets/1508362.pdf

Q7 - Q10 (marked 66C and b4) were eventually identified as dual fets (one n-channel one p-channel) which would work in the circuit,  (DMG6602SVTQ) which according to the data sheet would be marked 66C and the B4 would indicate a manufacturing date of april 2014 - plausible.

http://www.farnell.com/datasheets/2814392.pdf

So the list:-

Part   mouser no.
U1     634-SI8232BD-D-IS         SI8232BD-D-IS  Silicon Labs Gate Drivers  

part             farnell no.   desc
Q1 -Q4      2781406    FGH40T65SQD-F155 FIELD STOP TRENCH IGBT, 650V/80A, TO-247    IGBT
Q7 -Q10    2061522    DMG6602SVT MOSFET, NP CH, 30V, 3.4 A TSOT23       dual n & p channel fet
D2, D3   2069380    BZB84-B20,215 DIODE, DUAL, ZENER, 0.3W, 20V, SOT23    back-to-back 20v zenner

D8 15v      zenner >=0.5w
R13       7.5ohm
R9       15 ohm

0.5va 15v transformer x2

[Deej01]

Hey Marcus

Found this, one of your earlier posts, really good explanation and some listings of the parts..

Any idea on the caps ??

[Marcus]

Hi deej01,

I'd forgotten I'd re-posted this.

I don't think i actually had to replace any of the small caps, only the big black 5uF pair (immersun repair #1), so i don't have any part no.s, but i do still have immersun #1 here, sitting disused on the wall, and could take a look at it - which caps were you interested in?

Marcus

[Deej01]

The caps on the low switch side C1 and C4, I think they are all the same on the switching side !

[Deej01]

Immersun repair #1, I haven't found that yet !

[Marcus]

I probably haven't re - posted #1.

P.s. I haven't forgotten i was going to look at my immersun's caps, but I've been rather busy with work, plus the plumbers finally came back and commissioned my new ASHP last Thursday, and despite the incompetence of the installers, it's been working and I've been distracted by playing with it (and badgering them to correct their mistakes).

[Deej01]

Thank you, I did try measuring them on the board but I am getting silly readings, clearly this is not possible without the correct equipment !!