Checking capacitors in placce

[Dave A]

Is it possible to check capacitors without unsoldering one end?

[Peter P.]

I would say no, because the rest of the circuit may have some capacitance that your meter may see as capacitance in parallel with the capacitor under test.

 

However, you could easily test this: Most capacitors have their value printed on the outside. All you need to know is the manufacturer's tolerance for that capacitor.

If memory serves me, 10% is typical. Tighter tolerances are available, and if no tolerance is specified I believe it's understood to be 20%.

 

So measure the capacitor in the circuit and if you're within 10% of nominal, you can safely measure in-circuit and assume the capacitor is good. If not, you'll have to lift one end to verify.

[Dave A]

That is what I intend to try this coming week. I had asked the question because I might get direction on which LCR meter to get. For some reason I thought I had read somewhere that you could do this. I have ordered a meter now so it will be what it will be.

[Griffen]

You can't check capacitance without removing one end, but you can check the resistance. It should be above about 10-15 kohms or above. If it is not that high, it could be bad. At this point, remove one end and then check for the proper capacitance value.

[boom3]

There is a way to check components in-circuit. It has been used for many years by manufacturers and techs. I call the Lissajous method, but it is probably called by many names. In essence, a signal is introduced to a known good board. An oscilloscope with the X Y channels (horizontal and vertical) separated is used to read the signal across each component. The Lissajous pattern waveforms are stored, or if the circuit is simple enough, remembered. Then the suspect board or Unit Under Test (UUT) as we techs call it, is hooked up. The waveforms across each component are observed. The bad component's waveform will be quite different than that of a known good one. 

The best-known commercial version is the Huntron Tracker. In its automated form, it is a tremendous time saver. Disclaimer: I have no connection to Huntron.

http://www.huntron.com

[Davecv41]

Don't waste yer money on a Huntron. You can make one with a 2 channel O'scope, a couple resistors, and a 6volt transformer. $5 in parts vs. a $1500 (20yrs ago) dedicated tool.

 

Check the schematic before measuring any component, otherwise you can end up measuring other stuff in the circuit.

[boom3]

Yes, it's true you can make a simple lissajous outfit with what Davecv41 describes. However.. "Check the schematic before measuring any component, otherwise you can end up measuring other stuff in the circuit" is not my experience. We are not measuring anything here. We are comparing the lissajous pattern of a known good component in a known good circuit with a  suspect one. The lissajous method reveals components at variance with known good ones in an identical circuit. Measuring still requires taking the component out of circuit by breaking (lifting) one of the connections. With point to point wiring (like the old Klipsch crossovers) this is easy. With a circuit board, one has to have requisite experience not to fatally damage the component or the board in lifting one end.

[djk]

" It has been used for many years by manufacturers and techs. "

 

Good way to damage transistors.

 

Small signal transistors become noisy after their base-emitter junctions are avalanched (reverse biased).

 

You can buy an ESR meter on ebay for $20 or so, it's not the value of the cap that's the problem, it's the ESR.

 

https://www.ebay.com/sch/i.html?_from=R40&_sacat=0&_nkw=ESR+meter&_sop=15

[Dave A]

Got my B&K 885 in and the answer is no they cant be measured in place. Going farther than this with analytical equipment wont happen here so it looks like cut , measure and resolder.