Another dead MSD

[b.shaw]

I burned up another MSD 6AL (6420) recently and since it was out of warranty I started looking at repairing it. There is not much information out there about these units but I did find this posted by a gentleman in Indonesia:

Gabriel Espinoza from Jakarta, DKI Jaya, Indonesia blogged about it in 2009 –

---

i noticed that type of msd many time having problem or burned component inside, then i research the problems and find thats all problems of msd damaging process begins from 2 little condensators ( tantalum 2.2uf ). If buy new msd better you change this two C's with unipolar 2,2uf 50v electrolyte.
Anyway one of them have wrong position / polarity if compared from datasheet of ic TC4627

when condensators fails then voltage that drives mosfets begins to drop and mosfets starts to overheat because they working regime falls to the linear region ( not switching anymore ).... and then IGBT BUP 213 fails and with the consequences of burning step up voltage mosfet FS70KM and sometimes damaging zener diode 1N5244 B, in rare cases high speed rectifier diodes can too be damaged.


99% after you changing condensators on pin 1,2 & pin 3,4 ( gnd ), then changing zener 1N5244 B near FS70KM ( 4-th semiconductor from left ), change FS70KM ( you can replace it with IRF 1010E but you must using mica isolator from cooling body ground / casing and silicon paste), and change BUP 213 ( you can change it with IRG4PH40K again IRF product. Using mica isolator and you must move it position to side of housing using extension wires)...

---

Looking at my dead A6L and the TS4627 data sheet I see he is partially correct - one of the polarized caps is installed backwards. It is marked on the silk screen backwards also, so it is a design error that goes all the way back to the MSD schematic. This cap can fail prematurely causing the cascade Gabriel described above.

My question for the group is "Does anyone have a dead MSD ignition unit?". I'm interested in different models, to see if this problem was copied to more than just this one product.

Thanks.

[b.shaw]

btw - the circuit seems to work OK with the cap installed either way - the chip is producing 15v either way. The MSDs run OK for a while - I am getting a year or so of use from them before they fail.

Bill
'84 928s Vortech/MS II