Lets go into my induction stove

[Rui Almeida]

Ok Maxwell, any help from you is much appreciated.
I think you have to be patient with me, because I don't have an electronics course. The FET you are talking about, is it the power switching diode D34(D39), or is it something else? I have doubts on how I should measure the base to collector current on these small SMDs, but I think it's this current I measured flowing through the R11(R12) resistor. The way I did this was, I measured with my multimeter the voltage drop across R11(11.7V) and used Ohm's law to calculate the current going through to Q3(Q4) . The Q3 components I bought were these:, so I think they can handle that current, and the D34 components were these.
I don't know what all these components do in the circuit, there are so many, and the technical figure above doesn't help much.
Alternatively directly measuring the collector to base current would imply desoldering them, connecting them through wires to the PCB so that I can insert my multimeter in the circuit, right?
Since the stove has two of these PCBs and the left one does work(most of the time) I can take them both out and compare in-circuit multimeter measurements of resistances and diode readings of other components in the circuit of these burned parts. Thinking about desoldering a hundred or so, tiny SMD components to test them is indeed discouraging, but if must be...

 

[Maxwell]

I wouldn't worry about the microelectronics frying just yet, this could be a secondary issue.

What i would focus on, is turning the cooker ON reliably. You say it does not turn on half the time, you need to figure out why.

The main powerFET's will be large 3 pin devices (underneath the heatsink), if those are "ON" then the inductor should work, if not you need to take measurements from those FET's, what power are they putting out,, if they're not "ON" and they should be, check their base, see what's happening there. If the output isn't right, check what coming in, check the 5ùf cap (fails a lot on these!).

Check the major components first, and if those check out, you can make your way down to the control circuitry, the tiny stuff.

Most of the time, something major will not be turning on, or not be turning on reliably, and if the component checks out, it will be something in the control circuitry, often a shorted transistor (switch) or a shorted capacitor (reference to ground on both poles to see if they are shorted).

Also, you should check every single electrolytic cap on the board, check their ESR value and capacitance, when in doubt, replace.

Also check D12 and D15 for reverse leakage, I've had these leak on me before, causing all kinds of mayhem (including hole through board).

Resistors will be either open, smoked or cracked, if they are neither of those they are more than likely fine, a quick run over the with a high Ohm setting on your MM should bring clarity.

All the tiny crap on the back of the board is mostly just to amp up and filter the signal coming from the main IC and some power components on the top right (back of the transformer). The area where the big traces are is where the omph is being made, and all that oomph is then switched on by the two relays on the bottom left and dumped into the coils.

The large components are easier to check and more likely to fail. there's less of them, they're easier to desolder and easy to check. The tiny stuff you should only be concerned with once you've check all the larger components and made sure they were ok. A lot can be done in circuit, but some you'll need to desolder.

 

[Rui Almeida]

Thank you! I will do what you suggest, and will start by checking the specific components you mention and then the other big ones on the top of the PCB.
The FETs you mentioned, I suppose these are the Q1 and Q2 underneath the heatsink, I will first try to detect some significant difference between the two boards doing in-circuit measurements.
I think it will consume the better part of my weekend. When I'm through testing I will report.

 

[Rui Almeida]

I thought I just right now check Q1 and Q2, and there definitely seems to be comething wrong, here are the readings:
With the MM in the diode measuring mode:

COMPONENT :: connection :: LEFT BOARD :: RIGHT BOARD
Q1 :: C-E+ :: 403 :: 411
Q1 :: C+E- :: open :: open
Q1 :: C-B+ :: 501 :: 513
Q1 :: C+B- :: open :: open
Q2 :: C-E+ :: 406 :: 258
Q2 :: C+E- :: open :: 775
Q2 :: C-B+ :: 490 :: 258
Q2 :: C+B- :: open :: 775

I measured a shortcircuit on Q2 between B and E.
I'm sure I have to buy one or two of these components, do you know the specification of this component? Also, I once tryed to take off the heatsink but I think I have to desolder these FETs right? Or are there more things I have to desolder, do you know?
Do you think I should still go ahead and check all the rest?

 

[Maxwell]

You have to desolder the components under the sink to get it off.

The short is probably in another component, but you'll have to take out components one by one until the short clears. The "open" readings might indicate a broken resistor somewhere.

So desolder the FET's, take your readings again, see if you still have the short, if so, see what else is connected to those pins and find the most likely suspect, take readings, desolder, read again etc.

 

[Rui Almeida]

RIGHT BOARD:
I include a photo of the things under the heatsink.
So out of circuit the measurements in diode reading mode are as below
COMPONENT :: connection :: RIGHT BOARD
Q1 :: C-E+ :: 515
Q1 :: C+E- :: open
Q1 :: C-B+ :: open
Q1 :: C+B- :: open
Q2 :: C-E+ :: 258
Q2 :: C+E- :: 783
Q2 :: C-B+ :: 258
Q2 :: C+B- :: 783
On Q2 the outer terminals (B and E) are indeed shorted, while they are open on Q1.
The difference from in-circuit is that now on Q1 C-B+ is open, is this proper behaviour?
Component Q2 must be defect, all other measurements I've done seem ok, I found no leackage on D12 and D15 or other diodes, the values remained stable.
Should I go ahead and look for these FET components, or should I cary out more tests?

LEFT BOARD:
Update:
Last wednesday one of the relays on the board melted... fortunately in one of the last component order I got two of these so I could substitute it.
Till today, three days after, I haven't had the problem of one of the zones just turning off , lets see if this holds.

 

[Maxwell]

Hey

Did you replace that FET?

Measurements depend on what type of FET's they are, i can't recall now. If you are getting a short from the base to any of the other pins that means the FET is shorted and needs to be replaced. Relays melting are an indication of too much current passing through them, more than likely the shorted FET is the cause, also the cause of the zone turning off (overheat protection).

Q1 :: C-E+ :: 515 <- depending on the type this could be normal, or it might not.
Q1 :: C+E- :: open <- OK
Q1 :: C-B+ :: open <- OK
Q1 :: C+B- :: open <- OK
Q2 :: C-E+ :: 258 <- too low
Q2 :: C+E- :: 783 <- Either it's NPN or PNP, can't be both so this should be OL
Q2 :: C-B+ :: 258 <- Means it's always on, ie shorted
Q2 :: C+B- :: 783 <- See C+E-

 

[Rui Almeida]

Ok, thanks. I ordered FETs (BUTW92) (four of them just in case).
Because of your comment I will check the FETs on the LEFT board also.
When I have the new FET intalled on the RIGHT board I will let you know.

 

[Maxwell]

Loud bangs are not what we want

Sound like a short in the circuit that is blowing up the BUTW92 as soon as it is switching power, are you sure the coil is ok? Didn't you also mention one of the relays was blown before? This all seems to indicate that the zone is drawing too much power.

 

[Rui Almeida]

Hi Maxwell, thank you for your quick reply to my last post which I did delete, just not quick enough...
I got to see why there was a loud bang... some (unseen) solder joined two legs of the FET and on the filter board the security strip just melted when I turned that zone on. I soldered the filter board so as to make a new connection and on the main board waited for the element and substituted again the FET. And voila everything worked... but only for some days.
It's the right zone again it doesn't detect the pans
I have to open it all up again and see what the hell it is now.
unfortunately: "I will be back" with more details.

 

[Maxwell]

Off the top of my head i can think of only two reasons why that BJT would fail again so quickly... well aside from having two pins bridged ofc ;-).

1) Too much current is going through it cause it to melt internally. Measure amps from the mains with the right zone on and compare it to the left. If that checks out...

2) Bad snubber, assuming there is one. Traditionally it's a diode in series with a parallel resistor and capacitor, the capacitor will have a very high V rating, 1kv or there abouts, it will not an electrolytic, rather a foil or ceramic cap. The resistor may be open circuit, or the cap is dead, or the diode. If any of those components fail you'll get a voltage spike on the BJT every time it switches off, and this will kill the BJT in short order. Check the components on the snubber circuit, or if you happen to have access to a scope you can hook it up to the BJT and check the signal for V spikes.

Edit: another thing I wanted to point out is that you should actually be able to measure a diode across the B-E and B-C junction in the diode mode on a DMM, however I have found that sometimes BJT's will read OL even if they are good. Either way, whatever you are getting on the B-E junction should be identical to the B-C junction and you should always be getting OL on C-B, E-B, C-E and E-C for an NPN BJT.

Edit 2: I've done some MS-Paint on the image you uploaded some time ago. I traced around some components which I suspect might be part of a snubber circuit for the BJT, although the caps don't seem like your typical snubber cap, but check how the circuit is built, it's possible another cap is used for it that i can't see on the board right now.

Also the two caps on the right near the transformer seem bloated to me. You can see the plastic cover seems streched, that's usually not a good sign. I would just go ahead and replace all those blue caps to the right for starters, while you're at it you might aswel replace all the filter caps, since they're cheap and readily available. The small cap to the right of the transformer also looks suspicious, but it's hard to tell from the picture. Check them, and if you can't check them just replace them to be sure.

 

[davenn]

Ok, thanks. I ordered FETs (BUTW92) (four of them just in case).

those are not FET's, they are high current NPN transistors

 

[Rui Almeida]

Wow that is great Maxwell, that you have drawn it out for me! I will do the procedures you suggest in the coming days and will get back.

 

[Maxwell]

those are not FET's, they are high current NPN transistors

You're right, that's my fault.

 

[Rui Almeida]

Hi Maxwell!
Here are the updates on the RIGHT Zone, the left zone seems to still be working after I substituted that melted relay.
RIGHT Zone:
I took out for the x-th time both pcbs out.
I checked the "high current NPN transistor" that I substituted under the heatsink and it checked out fine.
About those suspicious looking caps of the photo, they actually don't at all seem bloated in any way and very much the same to those of the left zones pcb, they just come with a very slim upper plastic border. I also measured their values and compared the readings with the same elements on the left board and on both boards I got the same readings, also of the snubber circuit under the heat sink. I made in-circuit readings.
What I did easily spot was that again the Q3 and Q4 transistors are burned.
Maxwell could you please explain what it is exactly that you mean with your last comment:
"1) Too much current is going through it cause it to melt internally. Measure amps from the mains with the right zone on and compare it to the left. "
Can I do this with my MM, how do I do this?

 

[Rui Almeida]

RIGHT Zone:
Substituted Q3 and Q4, and since D34 and D39 gave strange readings, I substituted them too. Also substituted practically all capacitors(C10,C11,C19,C20,C33,C34,C36,C37). With out-of-board readings, only one of them seemed to be faulty. Substituted also resistor R32 which instead of 68kOhm was reading 96kOhm.
As a result of all this... nothing, the right zone still can't detect pans, the display just starts blinking.
Please help.

 

[Maxwell]

To measure amps you need to either put your meter in series with the mains line, or use a clamp meter. If you connect your meter in series make sure it can handle the load, this is tricky especially when you have possible shorts. You should be able to single out the zones at the connector terminals.

WIsh i had more info for you, but at the moment I don't have the time to go into it, will check back when i have some more time.

 

[Rui Almeida]

Thank you Maxwell.
Here are the latest news on the RIGHT zone:
I again checked the current across one of the R11 or R12 and it was still like what I got on the 3rd of november last year when I wrote:
"RIGHT-side:
Update on having put the new Q3,Q4,D34 and D39 components.
In measuring the current across them.
Well there is certainly a problem because, even with the right zones off, there is a current across these components, the measured voltage drop across the R11 resistor was 11.7V, so that means a constant current of (11.7V/48Ω=)244mA.
When turned on, the display starts to flash when putting power up on the zones as if they had no pans were on them."
And again the Q3,Q4.D34,D39 seem to have visibly overheated. It certainly wasn't smart of me to replace the originally burnt R11 and R12 0,5Watt 48Ohm resistors, by 1Watt(!!!) 48Ohm resistors, now everything(Q3,Q4,D34,D39) gets burned down the line.
I suppose that this current comes from the relay and that there is/are components that trigger the relay to always be on(?). Which component(s) is/are these, is it that ST IC chip: SG 3524 W990A9806?