smd resistors

[ewcc]

Hi,

I took out a board out from a netgear 16port switch trying to find where the fault lies.  The resistors
on this board is very small (about 1 mm in length).  Well, I assume they are resistors since they have
the R<#> next to them.  Anyway, there is a set of resistors that look the same though testing their
resistance gives me different answers.   While the majority of them are around 9k to 10k ohms,
there are two that are around 32 ohms.  Since I don't have a schematic, I have no idea if they
are supposed to be that value. 

Anyway, I'm not asking about how to repair this board (as this isn't the repair forum), I'm just
asking that given the small size of these resistors, how do I know what their values are?  Is
it even possible?  I mean, the larger resistors have the values on them (i.e. 1R3); but these
don't.  I have attached a picture of that part which has the resistors.  The black thing is a heat
sink on a chip.

Any clarifications appreciated,

Thanks


ewcc

[james_s]

Yes those are resistors, is there any printing on them? At least down to 603 size there is usually a 3 digit code that indicates the value. At any rate it would be extremely unusual for a resistor to fail such that it drops in value, usually if they go bad they go up or completely open. Also checking in circuit can be difficult because any other parts connected with them can interfere with the reading.

Whatever problem you're having with this switch I doubt it is related to resistors.

[TopQuark]

First, always make sure you are measuring the SMD resistor outside of the circuit, i.e. removing it from the board before measuring it. Measuring it in circuit can give you erroneous results as the rest of the circuit connected to the resistor terminals can mess with your readings.

For technique and tools, the most ideal and mess-free setup would be LCR tweezers + soldering tweezers + sharp stainless tweezers. You hold on to the resistor on it's side with tweezers, desolder it with hot tweezers, then measure it with LCR tweezers, then optionally solder it back with hot tweezers. It is very fast and reliable.

I personally use LCR research Pro1 Plus and Hakko fx-1003. Expensive, but worth it if you do it all the time for R&D and repairs. These tools and techniques will work for SMD inductors and capacitors as well.

If you want to use basic tools, you can use your soldering iron and a blob of solder to heat both side of the resistor to desolder it. Then just use your multimeter probes on both sides to get a reading. Not very elegant, solder blob may desolder other stuff nearby, multimeter probes may slip and shoot the SMD part into oblivion, YMMV.

[EPAIII]

It is obvious that you have very little experience with troubleshooting and repairing any electronics, much less modern PCBs with surface mounted devices.

The classic method of troubleshooting is to use an oscilloscope to tract the signal through the circuit. This requires a schematic so you know what that circuit is and a parts layout diagram so you can find each part. But a well labeled PCB can be substituted for the layout diagram. If the schematic has waveforms on it, that is a great help.

But I doubt that you have that. So what can you do?

1. Look for obvious problems: bad solder joints, parts that are burnt, other visible faults. A good magnifier or low powered microscope is a good tool here.

2. Look for bad electrolytic capacitors. They often have an end that is bulging. If you see any like that, replace them. In fact, if some electrolytic capacitors show this, replace all the electrolytic capacitors because the others will often fail soon anyway.

3. Look for other obvious faults. Bad switches, which usually can be tested in circuit. Burnt relays. Smoked inductors. Anything obvious.

4. If you haven't succeeded in finding the bad part(s) by this point and you do not have a schematic or a scope, then your best and easiest path is probably just replacing all the ICs and semiconductors (transistors, diodes, etc). Active devices fail much more often than the passive ones. I have had to do this in spite of years of experience and having all the above mentioned items at hand.

5. It is hard to measure resistor values or the parameters of other devices while they are in the circuit. This is because an Ohm meter or other tester will "see" not only the part you want to test, but every thing else that is connected to it. So, for instance, a high value resistor can give a very low resistance reading because of other paths through the circuit it is connected to. The rule of thumb is that all the leads, except for one, should be completely disconnected. With surface mounted parts, this usually means ALL the leads.

PS: Yes, resistors almost never fail by having their resistance fall to a lower value. Most bad resistors I have seen were obviously burnt up and showed either a very high resistance or an open circuit. And I seriously doubt that your problem is a bad resistor. That would be my very last guess.

PPS: If there aren't any bad electrolytic capacitors on the board, then that "black thing with the heat sync" is probably where the fault is. And, if that is so, then the board and the switch is trash - er, that is, not economically repairable.

Good luck!

[Andree Henkel]

The 10k and 33R are most probably ok, first are most likely pull-up or pull-down, while the later may well be series resistors in data lines.

[Siwastaja]

Such small resistors basically never fail. If you are repairing the thing, look for something else. If you want to reverse-engineer it for some reason, you should desolder the components and measure off-circuit.

[JustMeHere]

33 ohms is a standard resistor