Another smart guy Electrical / Electronics problem (???) This one...????

originally posted by: M5xaz
a reply to: Flyingclaydisk

Well, if it's not an intermittent connection/wire/thermocouple junction, as others have suggested above, and not TWO broken meters ( seems unlikely), might want to consider electromagnetic interference.

Would there be a large transmission tower nearby or is there a magnetic source nearby like a running electric motor or generator ?

That's an interesting angle. We do have some 320kV lines south of us not too far. I'm genuinely hoping to not complicate this with outside variables such as EMI/EMF.

originally posted by: 1947boomer

originally posted by: Flyingclaydisk
Okay, so I'm testing out a thermocouple today. Wait, let me back up; earlier today I wanted to measure a temperature with one of my multimeters which has a temperature function. I attached the temp probe and got some really wonky readings. Something's wrong. My readings were all over the place. Some, just null.

I have two meters which do temps. My first thought was it was the meter I was using (which is new). So I tried the same probe in an older meter I have. Same weird stuff, but less weird than the first readings. Okay, so now I suspect it's the temperature probe, and not the meter. Time to test the temp probe.

The probe is a Type K temperature probe. The basic principle is two different metals (of wire) brought together at a sensor will create a voltage. In other words, they will convert heat to electrical energy. This is how most electronic temp probes work. Okay, so this should be easy enough to check...

I set one of my better meters to DC voltage and clamp the positive and negative leads. I'm reading mV. I've got a value; let's say it's 20mV. Okay, so I put the probe in some hot water and the voltage goes up. Good! Right? I take it out of the hot water and the voltage goes down. Not the probe, right? Well....
.Then I start watching the voltage with the probe just sitting on my bench The voltage goes all the way to almost zero. How can this be? There's ambient temperature in the room, so there must be voltage...right? Then I grab the probe cable and the mV readings jump up to about 76mV or so (not touching the probe, just the sleeve of the cable). Could be lots of things, so I put the probe and cable back down, watching more closely this time. The probe will suddenly jump up to about 70-80mV for no apparent reason, then when left alone will go all the way back down to near zero. WTH?? Move the probe with a insulated rod, the voltage jumps up. At first I'm thinking a short in the probe cable, but that's not it because a short would only reduce the reaction between the two metals. What's going on?

I checked this with three different meters and got similar reactions, but with different values. Now, on my most accurate meter I get a temp reading of zero. On my other meter I get a reading of 72 F. Well, the room is about 72 F, so maybe it's not the probe...BUT, when I unplug the probe from that meter I get a reading of...72 F! Now I'm really scratching my head! So I shut all the meters off and went to go do something else (more productive). I thought maybe the one meter was holding a residual temp. Nope! Turned that meter back on about an hour later and got the same reading, 72 F (without the probe plugged in).

My Klein meter still registers zero, like there's a fault. Checked the resistance and capacitance across the probe, with nothing weird.

Any ideas? Is it the probe or do I have two bad meters??

A thermocouple is a “couple” Two different conductors with different work functions; that’s what drives the voltage in one direction instead of the other. It sounds to me like the connection between the two conductors is intermittent. Usually, they are just twisted together at the ends. That’s what I would look for.

Good point. This one seems to be soldered/welded/bonded together at the probe end. There's like a little ball there which the two wires terminate into.

My first reaction is that it's a conductor issue also. I just wish I had another probe I knew was working which I could compare it to.

originally posted by: Flyingclaydisk
Well, that doesn't really make sense, because how then could such a probe ever measure ambient temperature?

If the input voltage is zero, the meter interprets that as the temperature at the probe end is the same as the temperature at the end that's plugged into the meter, I don't know why you think that's a problem. Zero volts is a valid measurement, if the probe plugged in and working properly. If the probe isn't working properly, or if it's not plugged in, then it's not a valid measurement of what's going on at the probe end of the thermocouple.

In all the working probes I've seen the first reading you get after putting the setup together is ambient temp (whether it's 70 F, or minus -30 F.) And, when you get done testing something, the meter returns to displaying ambient temp.

Makes sense.

If what you're suggesting is correct and there is no differential temp then ambient temperature would always be the same value.

No, because the multimeter or whatever you plug the probe into has to have its own thermometer to measure the temperature at that end. The thermocouple gives the difference in temp at the other end, then the probe adds that to whatever its internal thermometer measures, to give the total temperature reading computed for the probe end of the thermocouple.

So it won't always give the same ambient temp reading. If ambient temp is 72, the meter will display 72 plus 0 degrees for the ambient temp probe. If ambient temp is 62, the meter will display 62 plus 0 degrees for the ambient temp probe. It's explained here:

blog.wika.us...

Thermocouples measure temperature differentials, not absolute temperatures. Two wires, each made from a different metal, are joined at the tip. This is the measuring junction. At the other end, the wires are connected to a body of a known temperature, called the reference junction. A thermocouple works by taking the difference in voltage between the two junctions, explained by the Seebeck effect. The measured voltage is converted into a temperature unit, with the temperature reading displayed on a device or transmitted to a remote location.

So if the meter and the wire are all at ambient temperature, then the meter will display the ambient temperature the meter itself is measuring. This is the result you're seeing with no probe plugged into the meter giving you 72 degrees, and the result of zero volts you see with the probe assembly all at ambient, so I don't understand why you are so doubtful. If the meter and the wire were all at 62 degrees, you should get a reading of 62 degrees, it's not the same as 72 degrees, so I don't understand why you say " ambient temperature would always be the same value." 62 degrees and 72 degrees are not the same value.

a reply to: Flyingclaydisk

Oh, oh! I know, I know! You have to put aluminum FOiL around the t.v.'s rabbit ears to get better receprion....

"M.S.? Get your ass back in here!".....I'm gone.

Don't know, just don't stick any body protruding part in any wall sockets....

a reply to: Arbitrageur

Hmmmm....Gonna' have to think about that one for a moment.

ETA - If that's the case, then I must have at least one bad meter.

a reply to: [post=25800697]Flyingclaydisk[/spost]

reason i asked, when i got my first digital meters from my boss as a Christmas present along with our bonuses one year. i didn't realize that they had a memory function on them.

our old anlaog simpon and amprobes didn't have them, and it took me a month or so to remember to make sure that i didn't turn it on or off.

My FIRST post on ATS. Make and model of meters? I've been in the test equipment and standards calibration/repair business for a long time. 1984. I'll help however I can.

a reply to: calman787

Meter is a Klein MM700 meter. Temp probe is a Type K (came with the meter).

I just ran a different test (can't believe I didn't think of this first). I get continuity between the negative leg and the end of the probe, but no continuity between the positive leg and the end of the probe. I'm thinking there's just a broken / open conductor on the positive wire.

Thanks for any input you may have.

ETA = The other meter is an old Craftsman 82325. (This is the meter I was getting the 70 F reading on, with or without the probe attached).

originally posted by: Flyingclaydisk
a reply to: calman787
I just ran a different test (can't believe I didn't think of this first). I get continuity between the negative leg and the end of the probe, but no continuity between the positive leg and the end of the probe. I'm thinking there's just a broken / open conductor on the positive wire.

Always open wouldn't be consistent with what you wrote in the OP, would it?

" Okay, so I put the probe in some hot water and the voltage goes up. Good! Right? I take it out of the hot water and the voltage goes down. Not the probe, right?"

But the wire could be broken, but making "flaky" or intermittent contact which would be consistent with what you described as "wonky" readings if that means jumping around or something.

As for two meters behaving differently, first, could it just be because of some coincidence with the sometimes open, maybe sometimes not open broken wire, or could the meters have different designs in some way that you're misinterpreting?

But yeah checking the resistance across the probe terminals was the first thing I suggested, should be low like less than 1 ohm. The article I posted said even a reading of 100 ohms is too high and could indicate deterioration of the probe, but infinite resistance (open circuit) is even higher than 100 ohms!

Arbitrageur nailed it. The chromel alumel alloys which compromise a K-type thermocouple do not take a lot abuse from folding and bending as much as copper does. You most likely have an intermittent open. Best to get a new beaded probe. If you want to check basic accuracy with the new one check at ice point and steam point. Crushed ice in a slurry. Distilled water and ice preferably but tap water is OK. Stir. Should be approximately 0° C/32° C. If you have a teapot, boil water, preferably distilled. Steam at spout should be approximately 100° C/212° F at sea level. Good luck

a reply to: calman787

You beat me to it. I used to use thermocouples a lot in testing. We had a known heat source and we would double check them before running the test.

a reply to: calman787

I found (5) of them for the price of (1) online, so now I have five coming.

Yes, they are indeed fragile. Sad though, because this one was brand spanking new out of the box.

originally posted by: Flyingclaydisk

originally posted by: 1947boomer

originally posted by: Flyingclaydisk
Okay, so I'm testing out a thermocouple today. Wait, let me back up; earlier today I wanted to measure a temperature with one of my multimeters which has a temperature function. I attached the temp probe and got some really wonky readings. Something's wrong. My readings were all over the place. Some, just null.

I have two meters which do temps. My first thought was it was the meter I was using (which is new). So I tried the same probe in an older meter I have. Same weird stuff, but less weird than the first readings. Okay, so now I suspect it's the temperature probe, and not the meter. Time to test the temp probe.

The probe is a Type K temperature probe. The basic principle is two different metals (of wire) brought together at a sensor will create a voltage. In other words, they will convert heat to electrical energy. This is how most electronic temp probes work. Okay, so this should be easy enough to check...

I set one of my better meters to DC voltage and clamp the positive and negative leads. I'm reading mV. I've got a value; let's say it's 20mV. Okay, so I put the probe in some hot water and the voltage goes up. Good! Right? I take it out of the hot water and the voltage goes down. Not the probe, right? Well....
.Then I start watching the voltage with the probe just sitting on my bench The voltage goes all the way to almost zero. How can this be? There's ambient temperature in the room, so there must be voltage...right? Then I grab the probe cable and the mV readings jump up to about 76mV or so (not touching the probe, just the sleeve of the cable). Could be lots of things, so I put the probe and cable back down, watching more closely this time. The probe will suddenly jump up to about 70-80mV for no apparent reason, then when left alone will go all the way back down to near zero. WTH?? Move the probe with a insulated rod, the voltage jumps up. At first I'm thinking a short in the probe cable, but that's not it because a short would only reduce the reaction between the two metals. What's going on?

I checked this with three different meters and got similar reactions, but with different values. Now, on my most accurate meter I get a temp reading of zero. On my other meter I get a reading of 72 F. Well, the room is about 72 F, so maybe it's not the probe...BUT, when I unplug the probe from that meter I get a reading of...72 F! Now I'm really scratching my head! So I shut all the meters off and went to go do something else (more productive). I thought maybe the one meter was holding a residual temp. Nope! Turned that meter back on about an hour later and got the same reading, 72 F (without the probe plugged in).

My Klein meter still registers zero, like there's a fault. Checked the resistance and capacitance across the probe, with nothing weird.

Any ideas? Is it the probe or do I have two bad meters??

A thermocouple is a “couple” Two different conductors with different work functions; that’s what drives the voltage in one direction instead of the other. It sounds to me like the connection between the two conductors is intermittent. Usually, they are just twisted together at the ends. That’s what I would look for.

Good point. This one seems to be soldered/welded/bonded together at the probe end. There's like a little ball there which the two wires terminate into.

My first reaction is that it's a conductor issue also. I just wish I had another probe I knew was working which I could compare it to.

Those are garbage temp probes, I had a feeling that's the kind you were talking about.

If you want a good temp probe for your Klein, go with this
Cla mp probe

Still wondering what the other meter is

ETA - I assume the probe you are talking about looks like this?

a reply to: Vector99

The other meter and model is (3) posts above yours.

Yes, the probe looks similar, but there is an adapter which takes it to the banana plug configuration to plug it into the meter.

EPILOG / Follow-up - (if anyone is reading this)

Well, as I think many of us suspected, it was indeed a bad probe. The new probes showed up and the meter jumped into action with dead-on accuracy!

The intermittent voltage readings threw me off for a minute, especially when heat was applied, but in the end it was a bad probe. Out of box failure.

Happy camper now!

The End.

ETA - Thanks all who participated. Hopefully you had a little fun.

originally posted by: Flyingclaydisk
in the end it was a bad probe. Out of box failure.

Thanks for confirming that.

By the way it's apparently not an unusual failure mode for a brand new probe due to the difficulties of welding.
Of the three ways to connect the wires at the probe end, twisting, soldering, or welding, each has advantages and disadvantages. The welding, if done successfully, has the best resistance to high temperature, and the best longevity, but the heat from the weld can cause a break in the wire, intermittent or persistent, adjacent to the weld.

If that's the cause of your failure, the probe may not be a total loss if you want to try one of the other two methods of connecting the wires. Soldering may require a special type of solder for K-type wires, and then you have upper temperature limits lower than the weld because of the melting point of the solder. Stripping the ends of the wires and twisting them together also works short term, but you would have to constantly check the calibration in the icewater and steam since it's not such a corrosion resistant connection as the other two, so it may work at first but get out of calibration after some corrosion comes between the twisted wires.

It might be more fun to play with the old probe, than throwing it out. If it's under warranty you may want to make a warranty claim though, instead of experimenting with it. But if it's out of warranty, there's nothing to lose by cutting off the weld and seeing if you can get it to work properly using the twisting or soldering method. Or if you're brave you might even try the weld, but that's extremely difficult, even the factory screws that up, as you might be able to affirm.

Who won the last thread?

nobody tells me # these days.

a reply to: Arbitrageur!

Well, considering I bought (5) new probes for a total of $10, I'm not sure how much of a warranty claim I'd get off of Klein. Probably not much (even though they sell the same probe for $39 bucks...LOL!). Experimenting with the old probe sounds fun. Although, sending the old probe back for a warranty repair also sounds fun. Choices...choices! I think I'm leaning towards experimenting with the old probe.

I did examine the old probe though and it doesn't look like the issue is right at the tip. The tip appears to be intact. I noticed when testing the probe that the orientation of the braided cable had a lot to do with the readings I was getting. Sometimes I'd get nothing. Other times I'd get increasing DC voltage with heat as I would expect.

I certainly had more than $2 bucks worth of fun just playing with the thing! My biggest happiness is, the issue wasn't with the meter!

a reply to: Lysergic

I assume you mean "this" thread. If so, I think everyone pretty much won. It was all just for fun.

If you're referring to the other thread about the 3 phase power, well, thenail won that one hands down. He figured it out early. That one was a real head-scratcher in the moment with some serious time pressure.

Ironic that it was another Master Electrician who deciphered the issue at the time. I'll never forget his words..."It''s just simply NOT possible!! Well...the only possible way this could happen, and it's a million to one, is IF...."

And, that is exactly what happened, even though the odds of it are way beyond 1,000,000 to 1.