Replacing unknown component

[eligri]

Hi!

A neighbour has asked me to help repair a broken pcb for a large cement blender. I have found three burnt out components; but the number on them is not visible and the company won't provide me a schematic.

Suggestions? What should I try doing?

Only have a few days to repair this.

https://imgur.com/a/brYjBCL



Cheers

 

[Pommie]

You need to post better pictures on this site.

Mike.

 

[Nigel Goodwin]

They are low value resistors (less than an ohm each, but that's all you can tell) - presumably they are used to monitor current?.

Do you have a low ohm-meter, so you could check them - they 'might' not be faulty, just over-loaded somewhat and discoloured.

There's also no guarantee replacing them would cure the problem, as it could be a fault elsewhere that's drawn too much current.

 

[Colin]

Since they are all less than one ohm, you can try something like three x 0.47 ohm one watt resistors (carbon-type) from your local shop. This is called "middle of the road" decision.

 

[schmitt trigger]

Resistors do not burn on its own. They burn because something else caused them to fail.

So I agree with Nigel. You could "fix", the problem, only for the resistors to burn up again if the original condition which caused the overload persists.

 

[Diver300]

It's not a good idea to guess the value. As Nigel said, they are likely to be used to monitor the current, so if you get the value wrong, the processor (U2 in the photo) will see the current as being too big, and turn off the motor (or other load) when it shouldn't, or not see the current as being as large as it is, and not turn off the motor when it should.

Resistances that low are difficult to measure, as the resistance of multimeter leads and contact resistance can make big errors. You should use a 4-terminal ohm meter. As those are expensive and rare, and the resistors may well be damaged anyhow, I would suggest that it might be easier to find a similar cement blender and get a photo of the PCB. The resistance values are easy to read from resistors that large if they aren't damaged.

 

[Nigel Goodwin]

Resistances that low are difficult to measure, as the resistance of multimeter leads and contact resistance can make big errors. You should use a 4-terminal ohm meter. As those are expensive and rare,

Interestingly a friend of mine was recently showing me his low-ohm meter that he built years ago - but never used as the four terminal probes were too expensive. However, he recently found cheap ones from China, and they work perfectly - he particularly wanted it for finding S/C SM decoupling capacitors on large (TV) PCB's. By measuring across the supply rails at different points you can find exactly which capacitor is S/C.

Here's a video of it:

 

[JimB]

I have a fancy 5 1/2 digit bench DMM which I have used on a couple of occasions to find the faulty (short circuit) decoupling capacitor on boards which have many such capacitors in parallel.
The way I did it was to inject some current from a bench PSU and follow to volt drops along the circuit board tracks using the DMM, which has a resolution of 1 uV.

However, your friends solution with his milli-ohm meter is a lot cheaper.

The Chinese Kelvin Clips look interesting.
I could use those with my fancy DMM which has connections for 4-wire resistance measurements.

JimB

 

[tomizett]

Indeed very likely that something else has failed and blown these resistors - so probably best try to find out what it was before investing too much time trying to identify the resistors.
What's opposite them on the other side of the board? There are some big pins coming through and what looks like a hole for screwing a tab to a heatsink - the pattern of the pins is unfamiliar though.
It's likely there is a big FET or IGBT blown in there somewhere - so probably best looking for that first.

 

[chemelec]

Or you could make my Low Ohm Meter.

 

[JimB]

Or you could make my Low Ohm Meter.

Nice simple circuit.

JimB

 

[Nigel Goodwin]

Nice simple circuit.

JimB

I presume it's similar to what Michael made, although his was mains powered rather than battery.

 

[chemelec]

I presume it's similar to what Michael made, although his was mains powered rather than battery.

Who is Michael and What did he make?

 

[Nigel Goodwin]

Sorry, post #7

 

[chemelec]

Sorry, post #7

In that Video, Milli Ohms is 1/Thousand, Not 1/Million

 

[Nigel Goodwin]

In that Video, Milli Ohms is 1/Thousand, Not 1/Million

I know

 

[unclejed613]

when working on Wyse terminals (the "green-screen" terminals many businesses used to have for everything from point-of-sale systems to managing customer information, etc...), it wasn't uncommon for one of the 40 or so TTL chips on the terminal PCB to short. the +5V rail was distributed in a grid fashion on the board, with no jumpers for isolating rows and columns of the grid. the fastest way to find the bad chip was to connect a current limited supply, and follow the voltage drops of the traces with an oscope on high gain... when you went across the top row, you would see progressively lower voltages until you passed the column with the bad chip, and the remainder of the columns would bee the same after that, then moving down the column until you pass the bad chip, where again the voltage drops would level off. it was also possible to turn up the current limit and feel for the hottest chip, but i didn't want to burn a hole in the board... so, in essence i was using an oscope and a power supply as a milliohm meter... there was also a nice hall effect sensor that could be used for following high current paths on boards to find shorted devices.

 

[eligri]

Sorry, been on vacation, and haven't checked this thread during that time!

I do have a low-ohm meter, but no replacement resistors. Probably not worth importing some, if the issue is likely to be elsewhere?

 

[debe]

In that video when he used a digital meter it showed the lowest resistance on the faulty diode.