How often do continuity testers damage boards

[IoTPanic]

Hey everyone,

So I was working with some ATMEL-SAMD21 boards I designed, and I had a board die on my while I was diagnosing another issue, and my boss, who is an MIT graduate and has spent many years in the industrial electronics industry, asked if I did a continuity test. I replied with a yep, and was given a speech on how the voltage from a continuity tester can kill sensitive electronics, and he told me to stop wasting my time on it now, given I tested if a data line to the MPU was held to the VCC rail. I was also doing other things to it. I honestly doubt it was my test, but I'd like some feedback on situations like this.

So my question to the community is, have you actually had a board die because of a continuity test, is what my boss talking about relevant/correct, and if this is an issue is there a dedicated piece of testing equipment that helps solve this issue?

Not looking to take up anyone's time, the board is long gone, just a question I have. 

[ataradov]

Was the board powered at a time? With the powered board there may be issues, but not likely. Unpowered board is very hard to kill using any function on the meter.

And if the board was powered, then normal DC voltage measurement will tell you where the pin is driven. Although I don't really see how you can kill anything this way.  Most likely you have inadvertently shorted something else on the board.

[vk6zgo]

Hey everyone,

So I was working with some ATMEL-SAMD21 boards I designed, and I had a board die on my while I was diagnosing another issue, and my boss, who is an MIT graduate and has spent many years in the industrial electronics industry, asked if I did a continuity test. I replied with a yep, and was given a speech on how the voltage from a continuity tester can kill sensitive electronics, and he told me to stop wasting my time on it now, given I tested if a data line to the MPU was held to the VCC rail. I was also doing other things to it. I honestly doubt it was my test, but I'd like some feedback on situations like this.

So my question to the community is, have you actually had a board die because of a continuity test, is what my boss talking about relevant/correct, and if this is an issue is there a dedicated piece of testing equipment that helps solve this issue?

Not looking to take up anyone's time, the board is long gone, just a question I have.

I'm a bit bemused by the habit people have of using the "continuity" function on boards.

I normally use the Ohms range on my old Fluke 77, as its output voltage is not high enough to turn semiconductor junctions on.(& because I was brought up on analog meters which had no continuity function)
The "continuity" & diode test functions are combined on that instrument, & are well & truly high enough voltage to turn junctions on.

That said, I doubt it would "kill" anything.

Maybe it was a bit of "received wisdom" from the old days when a continuity tester consisted of a battery &
electromagnetic buzzer.
Those things could definitely kill semiconductors!

[bugi]

As a side-question, were you using proper anti-static protection while doing those continuity (and other) measurements?

[SparkyFX]

Which device will get killed by what continuity tester is hard to answer, as these are unknown. People using it to check for errors might not be able to make the distinction because it is unknown if it was broken before.

The simple way to be sure would be to measure the output of the continuity test function with a second meter or oscilloscope and compare against maximum (reverse / IO pin) voltage of devices on the board. Modern testers do have a much higher impedance than the old 9V battery with buzzer, resulting in a somewhat lower risk of damaging something, OTOH logic devices became a lot more susceptible too (given their lower Vdd or low current output stages).

A relatively new meter (Peaktech 3440) states diode test as 3.2VDC/0.9mA maximum, continuity test as <0.35mA test current.
The 121GW lists 3V/1.4mA and 15V/7mA for diode and selectable 30/300 Ohm for continuity (no voltage given, 3.2V measured).

So even when taking point-to-point measurements on e.g. 3.3V boards: As you could connect the meter´s ground lead to supply (or an active high output) of a device, if the board is powered it means the voltage is increased by the voltage output of the (floating) meter and might exceed absolute maximum ratings on the other test lead. One safer way around that problem would be to connect the ground lead only to ground and measure voltages at certain test points (kind of representing the amount of junctions between these points).

[james_s]

I'm having trouble seeing how a continuity tester could damage semiconductors. I mean maybe if you were using an antique low end analog meter, or one of those testers that uses a light bulb?

[coromonadalix]

never damaged any boards, i do lots of tests ... anti static measures is a must, and be sure any board have power supply capacitors discharged

[IoTPanic]

As a side-question, were you using proper anti-static protection while doing those continuity (and other) measurements?

Not proper, but my office has wooden floors, I don't wear staticy clothing, and periodically throughout the day I ground myself on various computers I built and know the cases are grounded. 

Was the board powered at a time? With the powered board there may be issues, but not likely. Unpowered board is very hard to kill using any function on the meter.

And if the board was powered, then normal DC voltage measurement will tell you where the pin is driven. Although I don't really see how you can kill anything this way.  Most likely you have inadvertently shorted something else on the board.

It was not, I totally understand continuity puts a small current on the line, but I never had an issue with a continuity test before and was curious if modern electronics are just fine with it or not. I have no trouble believing old meters with old MPUs from the 80's this might be a big deal but it doesn't make too much sense to me with modern electronics.

Which device will get killed by what continuity tester is hard to answer, as these are unknown. People using it to check for errors might not be able to make the distinction because it is unknown if it was broken before.

The simple way to be sure would be to measure the output of the continuity test function with a second meter or oscilloscope and compare against maximum (reverse / IO pin) voltage of devices on the board. Modern testers do have a much higher impedance than the old 9V battery with buzzer, resulting in a somewhat lower risk of damaging something, OTOH logic devices became a lot more susceptible too (given their lower Vdd or low current output stages).

A relatively new meter (Peaktech 3440) states diode test as 3.2VDC/0.9mA maximum, continuity test as <0.35mA test current.
The 121GW lists 3V/1.4mA and 15V/7mA for diode and selectable 30/300 Ohm for continuity (no voltage given).

So even when taking point-to-point measurements on e.g. 3.3V boards: As you could connect the meter´s ground lead to supply (or an active high output) of a device, if the board is powered it means the voltage is increased by the voltage output of the (floating) meter and might exceed absolute maximum ratings on the other test lead. One safer way around that problem would be to connect the ground lead only to ground and measure voltages at certain test points (kind of representing the amount of junctions between these points).

Thanks for the info!

[IoTPanic]

never damaged any boards, i do lots of tests ... anti static measures is a must, and be sure any board have power supply capacitors discharged

I understand all of that, having a charged cap could have ruined my day I guess. I knew it, but I guess I haven't been consciously thinking about discharging caps. But honestly this was a prototype board and was a little sparse on the capacitors.

[pigrew]

The "sensitive" devices would likely be discrete transistors, which often do not have ESD protection diodes.  Some of my meters have outputs more than 10 volts in continuity mode.  The gate of a MOSFET can have max voltage limits of less than 10 volts and is high impedance so the college wouldn't be lowered during the measurement. Similarly, a reverse biased BE junction of a BJT often has quite low breakdown voltages.

Most digital ICs have ESD protection diodes which would shunt your current into the power supply rails. If your power supply rails are high enough impedance (very lightly loaded), this could also cause damage to the gates in there ICs input cells.

I normally do not use the continuity test mode, but when I do, I add an adapter I made with anti-parallel 1N4007 diodes to the input of the DMM to make a poor-man's "dry-ohms" mode. 1N4148 diodes would work equally well in this application. A 34401A reads 565 ohm with this adapter (1 mA test current), high enough to not trigger the beeping.

[IoTPanic]

The "sensitive" devices would likely be discrete transistors, which often do not have ESD protection diodes.  Some of my meters have outputs more than 10 volts in continuity mode.  The gate of a MOSFET can have max voltage limits of less than 10 volts and is high impedance so the college wouldn't be lowered during the measurement. Similarly, a reverse biased BE junction of a BJT often has quite low breakdown voltages.

Most digital ICs have ESD protection diodes which would shunt your current into the power supply rails. If your power supply rails are high enough impedance (very lightly loaded), this could also cause damage to the gates in there ICs input cells.

I normally do not use the continuity test mode, but when I do, I add an adapter I made with back-to-back 1N4007 diodes to the input of the DMM to make a poor-man's "dry-ohms" mode. 1N4148 diodes would work equally well in this application. A 34401A reads 565 ohm with this adapter (1 mA test current), high enough to not trigger the beeping.

Wow, I'm really glad I asked this, that trick with the diodes is very clever. This all is making more sense to me.

[bugi]

As a side-question, were you using proper anti-static protection while doing those continuity (and other) measurements?

Not proper, but my office has wooden floors, I don't wear staticy clothing, and periodically throughout the day I ground myself on various computers I built and know the cases are grounded. 

Grounding periodically throughout the day won't do a thing.. the static charge can develop in a fraction of a second. Just see how Dave does it by simply standing up from his chair (though his materials could be worse)


With a bit of bad luck, stand up/turn around/shuffle a paper/move a plastic box, pick up the DMM probes, touch a bad spot... *zap* without you even noticing anything.  Still, it might not have been the reason in your case this time, but in general with sensitive devices, using proper anti-static protection stuff is pretty cheap and easy way to eliminate one potential source of problems.

[IoTPanic]

As a side-question, were you using proper anti-static protection while doing those continuity (and other) measurements?

Not proper, but my office has wooden floors, I don't wear staticy clothing, and periodically throughout the day I ground myself on various computers I built and know the cases are grounded. 

Grounding periodically throughout the day won't do a thing.. the static charge can develop in a fraction of a second. Just see how Dave does it by simply standing up from his chair (though his materials could be worse)


With a bit of bad luck, stand up/turn around/shuffle a paper/move a plastic box, pick up the DMM probes, touch a bad spot... *zap* without you even noticing anything.  Still, it might not have been the reason in your case this time, but in general with sensitive devices, using proper anti-static protection stuff is pretty cheap and easy way to eliminate one potential source of problems.

I will make a equipment request for anti-static protective gear for my work station now I guess 

[hugo]

A Fluke 179 multimeter set on continuity mode will put 7.25 V, 1mA through the DUT and that would be enought to turn on some MOSFETs.

[Neilm]

If the continuity tester is designed to meet IEC61557 it can output >200 mA with an open circuit voltage of >4 V. This could damage a sensitive component (for instance if you put it on a 0.9 V power rail). I have not ever had direct experiance of this.

[HKJ]

On most meters continuity is less than 3½ volt and below 1mA shorted, the current is feed from a resistor not a constant current generator. Very few components can be damaged by this level of voltage and current.

[james_s]

>200mA continuity test strikes me as the sort of thing that would be used on mains and automotive power wiring, not troubleshooting electronic equipment. Certainly I wouldn't expect it on an ordinary multimeter, more likely a more specialized tool, like the old continuity test pens that contained a battery and a small incandescent lamp.

[Neilm]

>200mA continuity test strikes me as the sort of thing that would be used on mains and automotive power wiring, not troubleshooting electronic equipment. Certainly I wouldn't expect it on an ordinary multimeter, more likely a more specialized tool, like the old continuity test pens that contained a battery and a small incandescent lamp.

It is designed for mains wiring - however I have seen "multi-meter" that do attemt to output something similar to the requirements. Some older meters output 5 V with quite a few mA behind it.

[stj]

interesting subject,
i have often wondered how dangerous it could be using a modern meter with 2.5-3v output to test a board with 1.8v logic on it.