How do you test diodes, resistors, and capacitors?

[giolicious]

I'm trying to fix a 128k Analog Board and would like to know if I can test the diodes and resistors and capacitors without removing them from the board? i have a digital multimeter.

[GnatGoSplat]

I'm not even remotely an expert, but here's how I check them:

Diodes - put multimeter in diode check mode. It should show the voltage drop when leads are attached in correct polarity (usually 0.3V to 0.6V) and nothing (open) with leads reversed. If open both directions, diode is bad (open). If 0V both directions, diode is bad (shorted). If you find a bad diode, pull one leg and retest to confirm it's the diode and not something else in the circuit.

Resistors - Put multimeter in Ohms mode. Reading should match resistor bands. However, resistors often measure lower ohms than their bands if they are placed parallel in a circuit, so just because the reading doesn't match, doesn't mean the resistor is bad. In my experience, they are seldom bad if not discolored/burned (although there is such a thing as a fusible resistor that can go open without any visible indications). I have never seen a shorted resistor in my life, so typically if the resistance measured is near what the bands say they should be or LOWER, the resistor is okay. If the resistance is significantly higher, the resistor is probably open. If you want to be 100% sure, you can desolder one leg of the resistor and test it that way.

Capacitors - Put multimeter in Ohms mode. Capacitor should start with low ohms and start to rise as the capacitor charges. It could be 0-ohms (shorted) or read nothing (open). A multimeter is not really a very good tool to use to check capacitors though, as the capacitor could be bad, but still exhibit normal looking behavior on a multimeter. Usually you want to use an ESR Meter to check capacitors. If electrolytics are swollen or leaking, then no reason to check it, it's bad.

[giolicious]

Thanks GnatGoSplat!

How about the two big diodes at C20 and C21, how do you test them? is it the same procedure as you mentioned above?

[Trash80toHP_Mini]

GREAT TOPIC!

I love acronyms . . .
Equivalent Series Resistance Meter
. . . but I really love WikiPedia Linkage!
ESR Meter HA! A WikiPedia link appears inside our URL code!

I once saw a HowTo on a HomeBrew version of the (very expensive) ESR Meter, has anyone else seen one?
This sounds like another great project for the 68kMLA Technical Unit boffins!

[Gorgonops]

I've personally had very little luck testing capacitors in-situ. I've found that a given capacitor (which is good) will test to its rating on the bench but when it's soldered into place it the fact that it's grounded into other circuitry will prevent it from giving a normal reading.

Maybe that's just my experience, but I wouldn't trust a multimeter reading to diagnose a bad cap.

[GnatGoSplat]

Thanks GnatGoSplat!

How about the two big diodes at C20 and C21, how do you test them? is it the same procedure as you mentioned above?

C20 and C21? Do you mean capacitors? Yes, it would be the same procedure as I mentioned above, but be aware as I mentioned before, a multimeter can only tell you 1 of 2 things about capacitors:
- It's definitely bad (shorted or open).
- It's maybe bad (looks normal on the multimeter).

A multimeter can't tell you that a capacitor is GOOD. Even if your multimeter has a capacitance range, an electrolytic capacitor can have its capacitance in spec, but its ESR way off.
Definitely need an ESR meter to tell if a cap is for sure good, and yes, you often have to remove them to test out of circuit even with an ESR meter.

This ESR meter is highly rated by people who have to deal with bad caps often: http://www.anatekcorp.com/blueesr.htm
You can also buy it here: http://www.flippers.com/catalog/index.php?cPath=3
I have its predecessor, definitely a valuable tool for repairing anything using electrolytic capacitors. Although if this is the only thing you plan to repair for awhile, rather than buying a meter, it might be cheaper to just replace the caps.

[giolicious]

Oops sorry, I mean the two big diodes at CR20 and CR21. The ones with heatsink. How do you check/test them?

[Dennis Nedry]

The only problem with testing components without removing them is that other components elsewhere on the board may, and often do, interact with the reading. Capacitors are even a little bit more tricky, because first of all, only more expensive testers can directly test capacitance. Second is that most testers that do test capacitance don't test them under load. It may look good on the tester, but once loaded down, might be a different story.

Usually when a capacitor goes bad, it won't test good with a capacitor tester. A lot of times caps short out, so you can use an ohm meter or diode tester across it. If it's shorted, it's bad. BUT if it's not shorted, that doesn't mean that it's good!

[Trash80toHP_Mini]

I've personally had very little luck testing capacitors in-situ . . .
Maybe that's just my experience, but I wouldn't trust a multimeter reading to diagnose a bad cap.

Have you used an ESR Meter at all? ISTR, 4sPhoto saying that he's had great success using one.

You can also buy it here: http://www.flippers.com/catalog/index.php?cPath=3
I have its predecessor, definitely a valuable tool for repairing anything using electrolytic capacitors. Although if this is the only thing you plan to repair for awhile, rather than buying a meter, it might be cheaper to just replace the caps.

Thanks for the links, I just downloaded the assembly manual, does it pay to spend the extra $40 to get it assembled . . .
. . . or is it a good way to earn a few credits in the Electronics School of Hard Knocks for a negative $40 tuition payment?

[GnatGoSplat]

Oops sorry, I mean the two big diodes at CR20 and CR21. The ones with heatsink. How do you check/test them?

If they're diodes, you can check them the way I described with the diode check function - meter will read voltage drop (usually 0.3-0.6V) in correct polarity, infinity (nothing) in reverse polarity. If either appears shorted, then you'll need to desolder one leg or remove it and test out of circuit to confirm it's bad.

Thanks for the links, I just downloaded the assembly manual, does it pay to spend the extra $40 to get it assembled . . .
. . . or is it a good way to earn a few credits in the Electronics School of Hard Knocks for a negative $40 tuition payment?

I assembled mine myself, but I enjoy both assembling electronic circuits and saving money so it was a no-brainer decision! If you're fairly adept with a soldering iron and enjoy spending time assembling things, I'd recommend getting the kit and assemble it yourself. It's fun!

[phreakout]

Boy, this makes me want to pull back out my old electronics text books I was given back in high school. I had taken the course for my junior and senior years through a vocational campus. Mostly taught about AC and DC circuits, component recognition, safety, series and parallel resonance circuits, schematic creation and diagnostics, soldering, repair and overall diagnostics of circuits. Now, 13 years later, the program is no longer offered except for the community college. Quite sad and disappointing, though.

73s de Phreakout.

[Bunsen]

Stickied, and moved to Hacks and Development

[Gorgonops]

I've personally had very little luck testing capacitors in-situ . . .
Maybe that's just my experience, but I wouldn't trust a multimeter reading to diagnose a bad cap.

Have you used an ESR Meter at all? ISTR, 4sPhoto saying that he's had great success using one.

No, just the capacitance setting on my cheap multimeter. And looking at the prices of an ESR meter I think I'll stick with that. ;^b

(I could replace every capacitor on all of my Commodore PETs several times over for the price of the cheapest one I see.)

As noted, if the capacitor is sitting loose the multimeter seems to give a reading close to the rated value of the capacitor, which really is good enough to me. (Gives me a quick double-check to verify I'm installing the right one, as the markings on tiny capacitors can be really difficult to read and interpret.) But in the most recent particular case: I was repairing a bad 555 Timer-based RESET circuit which involved three capacitors, and in-situ none of them registered anything close to a sensible reading on the meter. I tested one, got excited that it looked off, replaced it, and it turned out to not solve the problem. (And the one I'd just installed "tested bad" once it was soldered in there.) So I tested the two remaining ones, they both gave funky readings, so I just said "screw it" and replaced them both at the same time. Problem solved, but I destroyed the capacitors when I snipped them off so I don't know which one was the bad one and I couldn't test after the fact. I'd be miffed, but really all it cost me was an extra 60 seconds with the soldering iron and about 3 cents for the extra capacitor.

Maybe a better meter works moar better. I have a *fantastic* $20 China Special.

[Trash80toHP_Mini]

No, just the capacitance setting on my cheap multimeter. And looking at the prices of an ESR meter I think I'll stick with that. ;^b
(I could replace every capacitor on all of my Commodore PETs several times over for the price of the cheapest one I see.)

I'm a tool user/collector/maker, so I'd rather do the ESR Meter project than soldering all my caps en masse!

Maybe a better meter works moar better. I have a *fantastic* $20 China Special.

I've got half a dozen (at least, and I can never find the one I want ) ranging from a $5 Harbor Freight Digital up through a nice Compact. Beeping CrapShack Digital and a really nice Analog Meter from the PhoneShack . . .
. . . before they turned into CrapShack!

I still haven't found a SillyScope for under $100, but then again, I've got a nice Logic Probe from Radioshack that's mostly untried for . . .
. . . waaaay longer than I'd like to admit!

[techknight]

And.... a capacitor can have a great ESR and its capacitance be way off.... been there done that on multiple occasions.

And, in rare occasions, a capacitor can have its ESR good, and capacitance good, and STILL be bad. Why? leakage towards the working voltage. Been there done that as well, but again its rare.

Ideally you want your ESR as close as 0 as possible, but we are not in an ideal world and even the best of capacitors will still have an ESR slightly above 0. This also varies with capacitance, and the reason being is an ESR meter is a fixed-frequency AC ohm-meter. Adjust the frequency, and the whole table changes drastically.

[GnatGoSplat]

I still haven't found a SillyScope for under $100, but then again, I've got a nice Logic Probe from Radioshack that's mostly untried for . . .
. . . waaaay longer than I'd like to admit!

If you've got an old PC or laptop to spare, look out for an old Tektronix TWD120. It's a SCSI-based 100MHz scope, controlled by software on the PC. System requirements are very low, the software was designed for Windows 3.1, but it works in all newer 32-bit Windows that I've tried. The nice thing about a PC-based scope is you can capture images of your waveforms. I've seen them go for next to nothing on eBay in the past.

[Trash80toHP_Mini]

Thanks, I'll add that to my saved searches!

[onlyonemac]

I could replace every capacitor on all of my Commodore PETs several times over for the price of the cheapest one I see.

Err... meaningless if you don't have that many Commodore PETs .

[James1095]

An ESR meter is the *only* worthwhile way to test electrolytic capacitors. The capacitance function on a multimeter is useful for identifying unmarked capacitors but it won't tell you if the part is any good. I bought one about 15 years ago and use it all the time, but I do a lot of repair work. If you just want to fix a few things here and there, you might be better off using the shotgun approach and just replace all the electrolytic caps.

In terms of vintage macs and peripherals, it should go without saying that *all* surface mount electrolytics need to be replaced, and they should be replaced with dry tantalum or multilayer ceramic types now available, these will never leak and corrode your PCB. Every 68k logic board and NuBus card I've seen recently that uses these has had them start to leak, they were only rated 1,000-2,000 hours to begin with and the specs are lousy even when new. Most of the through-hole parts seem to hold up well, the exception being those used as output filters in switchmode power supply circuits including the horizontal scan and high voltage area in CRT monitors. 90% of the dead power supplies I encounter have bad capacitors and spring back to life with those replaced.

[krye]

With capacitors, it's true you can't measure capacitance in circuit. You can, however, find clues as to whether or not it's good with resistance and voltage measurements. You should see resistance across the leads if it's good, 0 resistance if it's a short. As far as voltage, you should see voltage rise out the cap as you apply power to the circuit. This might not necessarily tell you if a cap is bad or not, but if you replaced one, then you'll know if it's making good contact.

Just remember, you can take voltage and resistance measurements in circuit, in parallel, with a component, or in relation to ground. But not current. Current has to be measured in series. If you try to take a current measurement across a component in parallel, you can blow your meter.

Another thing to remember is that meters show a "difference in potential", not the voltage. If you have 100V leading into a resistor, and 95V coming out, placing your leads across the resistor will show you 5V. That is the difference in potential from voltage in to voltage out. If you touch it, you about to get zapped with 100V not 5V. That's why it's good practice to always take measurements in relation to ground. (black lead on ground, red lead where you're measuring.)

Safety first!