Industrial Computer SMPS (TDA4605-3) Faulty

[kolbep]

Hi All.

I am looking at the SMPS section of an Industrial Computer's Interface Board.

Day 1, Unit Would not power on, until about an hour later
Day 2, Unit would not power on again when in the machine. If I disconnected and hooked it up to a mains socket, it worked.
 Eventually discovered that the Step Down transformer in the Machine converts to 220v, and the standard socket outlet was giving 240v.
 So the unit refused to start up on the 220v.
 Changed the transformer tap so that the machine was supplying 230v to the Industrial computer, and it ran fine the whole day.
Day 3. Unit would not power up any more (you could hear the spark when the plug was put in, when the capacitors were charging, but no output).

Taking the unit to the bench, The Switchmode PSU section of the control board uses a TDA4605-3 Switchmode Controller (and it looks like a small Voltage reference 8 pin chip on the rear).
The outputs are supposed +12,-12,+5v

I do not have a schematic for the board, so I am just working with the Appnotes for that chip, which are similar.
I found that the  R1 is open circuit (75Kohm is indicated in the appnote, but the resistor on the board, from what I could see, did not seem to have the correct colour code).
Replaced it with a 220Kohm that I had on hand as a test. Shortly after power on, the fuse blue, and the fet and TDA4605 both measure shorted.

I have ordered a replacement fet and TDA4605-3, so will be replacing those when it arrives, and will probably do the series lightbulb trick (just need to find an old style bulb),
But, Regarding the Resistor, the Appnote, and what are on the board seem to differ. How critical is R1's value, and would it be a problem if I get a 75k as per the appnote (or can somebody help identify the resistor I have pulled (pics included).

The controller also takes a divided voltage input to pin 3, which is basically a low-volt cutoff. It has a resistance Ratio of 540k:9.4k, I calculated if the cap voltage is 400v, then I should get 6.84v on that pin. But I cannot figure out from the datasheet, what the chip would find acceptable.

Thanks for any suggestions.

Peter

[SeanB]

75k is not too critical, though i would also change the capacitor C2, because it probably is leaky, and also check those SMD resistors have not drifted in value. Clean the board after putting in the new parts, though the resistors in the sense circuit can be checked in circuit, so long as they are not higher than the marked value they will be fine. The 75k resistor should be made with 2 lower value 0.5 W resistors in series, probably 39k and 33k, which will come in at 72k, close enough, or 2 39k in series, which will be 78k, both within 5% tolerance.

[kolbep]

Thanks Sean.

I have placed the order from Mantech (including the Resistors, Caps, etc), and my customer will be sending their driver up to collect tomorrow.
When got the quote for a replacement board (only R165000.00, and the wait for shipping from Germany), I thought, maybe it is worth a shot to spend a few hours and try salvage this board, even if it means replacing everything on the Primary of the Transformer.

P

[SeanB]

I had a similar quote from Eli Lilly, and the Chinese copy of the entire machine worked out to be cheaper to buy instead, and take the old machine, strip and send parts for machining. That total came to way less than just the ELANCO quote, and as a bonus no 3 month wait.

[kolbep]

So, Just a follow up :

I forgot to mention that the machine was a VF200 Vacuum Filler for a Butchery to stuff sausage, etc.


I went and Replaced the SMPS Controller Chip, FET, Electrolytic caps, etc, and the board came to life.
However the voltages were low, and the powersupply was pulsing.

Then Eskom Loadshedding happened. So I paused for the night.

The next day, it popped into my head that the Powersupply probably wants a minimum load, so I took the risk, and hooked up the CPU card to this Powersupply/IO card, and the unit came to life.

The R1(2 of them in series) was getting quite warm, So I made a serial-parallel chain of them, to keep the input to that Controller at 12V, and the current within the specs for the chip.

And, it has been a few days now, and the machine is still running, and the client is happy.

P

[inse]

I wonder why R1 is getting rather warm as it is only supplying the startup voltage for the TDA.
From the picture, I'd say its a 12k, 5% resistor (brown, red, orange, gold)
When operating, it is supplied by the auxilliary winding on the transformer.