Thermocouple probe with EEVblog BM-235 inaccurate. Options to correct?

[rbm]

I own one of the fabulous EEVblog-branded BM-235 multi-meters.  Great meter and it's quickly becoming my go-to meter for most of the measuring jobs I do.  I love that it came with a thermocouple probe to measure temperature but the readings are inaccurate, making me doubt that I can rely upon it.  I first began to doubt the accuracy when I tested the tip temperature on my soldering iron (set at 350C but read 250C) and then my Weller hot air rework station (set at 650F but read 525F).  I tested by not very scientifically measuring known temperatures - ice at 0C and boiling water at 100C.  The indicated readings were -2C and 97C respectively.  This might be attributable to my techniques but it might also be mismatch between the characteristics of the probe and the meter that gets worse at higher temperatures.

What are my options?  I guess I could try other thermocouples with the meter.  I could also go out and buy a more accurate temperature probe that can measure within the range I need (up to 500C).  Is there a way to calibrate the meter itself to correct for this mismatch and reduce the error?

[joeqsmith]

I own one of the fabulous EEVblog-branded BM-235 multi-meters.  Great meter and it's quickly becoming my go-to meter for most of the measuring jobs I do.  I love that it came with a thermocouple probe to measure temperature but the readings are inaccurate, making me doubt that I can rely upon it.  I first began to doubt the accuracy when I tested the tip temperature on my soldering iron (set at 350C but read 250C) and then my Weller hot air rework station (set at 650F but read 525F).  I tested by not very scientifically measuring known temperatures - ice at 0C and boiling water at 100C.  The indicated readings were -2C and 97C respectively.  This might be attributable to my techniques but it might also be mismatch between the characteristics of the probe and the meter that gets worse at higher temperatures.

What are my options?  I guess I could try other thermocouples with the meter.  I could also go out and buy a more accurate temperature probe that can measure within the range I need (up to 500C).  Is there a way to calibrate the meter itself to correct for this mismatch and reduce the error?

There was a person who had posted on one of my videos where they bought a brand new meter and was trying to use a wall wart to check the meter's accuracy.  The idea being they felt the wall wart had some specs about AC voltage and current and this is what they felt the meter should read.   They proceeded to measure line voltage with it as well.   Of course we all know that we can't use this sort of data as a reference and wall wart's outputs are highly dependent on the load.     

That said, I wonder why you feel the problem is with the meter?   Do you feel your solder iron is that accurate? I wonder how much of a variance you would see looking at various locations on the iron.

I have this meter along with the virgin probe it came with.  I'll see how mine behaves.  Maybe there is a problem with them.  I did not check mine outside of room temperature.

[meeder]

Thermocouples aren't the most accurate probes in the first place but you should take a couple of things into account. Depending on the construction it can take a very long time to stabilize after you plug the probe in the multimeter since the cold junction should be at a stable temperature when measuring.

[MosherIV]

I tested the tip temperature on my soldering iron (set at 350C but read 250C) and then my Weller hot air rework station (set at 650F but read 525F). 

Not exactly a very scientific test for temperature is it?

Bear in mind that thermocouple probes have a working range, most cheap ones will not got beyond 250C
I do not know what the specifications of the one for the BM235

If you want to verify the accuracy of the temp on the Bryman, try measuring melting ice (ahoudl be 0C) and boiling water (100C)

[Kleinstein]

The -2 C and 97 C reading for ice and boiling water point to an about 2 degree error in cold side compensation. But this would be an essentially constant error. The error is slightly on the high side, but temperature measurement is often not that accurate.


For the higher temperature you might need a more accurate reference point. The tip temperature of a soldering station can very well be 50 C off.  One possibility would be the freezing of a known solder alloy (e.g. 178 C for typical leaded). You might need protection of the tip, as solder may alloy with the probe and thus slightly change the tip. 

An other option would be to provide a small external voltage (e.g. 10 mV range thus corresponding to about 270 C) and check the reading.

[IanB]

To give you a feel for what to expect, here is a snapshot of four different instruments all reading the same temperature (all four thermocouples close together in an insulated box and the readings allowed to stabilize).

One normally might expect an accuracy of ±1 degree F at best with such an instrument. (The BM235 manual quotes 1% ± 2 deg F.)

I also noticed when setting up this test that there is a lot of noise and the readings fluctuate a lot. The thermocouple tips themselves are also very sensitive and the tiniest change in surroundings or air currents will cause the reading to move.

So overall I would not be so hasty to doubt the instrument. Granted there might be sample variations, but I would guess the meter has been calibrated fairly well.

Note that water will only boil at 100°C if it is distilled water at standard atmospheric pressure. You are unlikely to meet those conditions outside a laboratory. Similarly, getting melting ice to achieve exactly 0°C also takes some care.

[rob77]

-2 and 97 celsius for 0 and 100 is actually not a bad result...
the difference when measuring the soldering iron was most likely caused by poor thermal coupling - your thermocouple was touching the iron in 1 point  instead of being "surrounded by the temperature it measured".
the hot air station is a different story - the the further are you from the nozzle the lower the temperature is.. and the gradient is pretty steep, so few centimeters might mean tens of degrees difference.

[joeqsmith]

My BM235, BM869s and my Mastech meter using various probes....

[billfernandez]

I live at 5,000 feet above sea level.  Water boils at about 95 degrees C (203 degrees F).  So knowing atmospheric pressure, and using distilled water, and determining at what temperature it SHOULD boil in your situation, are all important to get the best results.

Even so, as a rule of thumb, don't expect any one temperature measurement to be more accurate than around 3 to 5 degrees F away from the truth.

[Jester]

I have a couple of reasonably accurate digital thermometers (NIST calibrated). If I place the probes very close to each other, the displays typically agree to 0.1C

At times they will temporarily creep away from each other because the meters have different filter time constants. That being said if I separate the probes by say 6" the displayed temperatures are not nearly as well matched, my hunch is that air currents have a noticeable effect on localized temperature.

[rbm]

Thanks for the replies.

I certainly believe the meter and thermocouple are reasonably matched and the +/-2% errors seen between 0C and 100C are within expectations.  But, 100C difference between expectations and actuals lead me to question the measuring device as I couldn't believe that the measured devices themselves were both so extensively out of calibration..

If I assume from my non-scientific test that I should expect a +/-2% or even +/- 5% error in the readings given the variability of the probe, then I would expect  that same error to be maintained to the high end of the scale.  It means my soldering iron and hot air station require calibration.  I couldn't believe at first that both workstations were in error but maybe that is the case.

I'm seeing a 125C difference in the set point to the indicated temperature on the hot air station which concerns me.  I made the measurement by introducing the probe directly into the opening of the nozzle. So, it should be within a few degrees of the set point.  The air volume was set at 4L/m so that the velocity of the air wasn't too fast to cool off the air stream.  The Welller station does not have any analogue calibration mechanism so I have to figure out why its so cold.

I'm seeing a 100C difference in the setpoint to reading on the soldering station.  Even if it did measure in a way as to surround the probe with heat, it should not be that far off. The soldering station is a cheap Atten clone of a Hakko (I expect it to be off) but it has a pot to adjust tip temperature, so it means I can bring it quickly into spec.

In short, I don't have any reason to doubt the meter, so I need to trust the readings and get on with the job.

[Macbeth]

There are devices specifically made to test soldering tip temperatures like Hakko FG-100

Plenty of knock-offs available too for very cheap.

https://www.youtube.com/watch?v=iqxEMFCvGtc

[IanB]

I'm seeing a 125C difference in the set point to the indicated temperature on the hot air station which concerns me.  I made the measurement by introducing the probe directly into the opening of the nozzle. So, it should be within a few degrees of the set point.  The air volume was set at 4L/m so that the velocity of the air wasn't too fast to cool off the air stream.  The Welller station does not have any analogue calibration mechanism so I have to figure out why its so cold.

Did you watch Joe's video above? He did some tests at higher temperatures.

However, you should observe two limitations. Firstly the meter electronics are only specified to indicate temperatures up to 400°C / 750°F (see page 25 of the manual).

Secondly, a simple probe such as the one supplied with the meter is only good to measure up to about 250°C (see here for example, or here).

If you heat the probe up above the maximum temperature indicated there is a possibility you could damage it, which would cause it to read inaccurately afterwards. You might like to obtain a new probe for comparison if you already heated up your existing probe above 250°C.

[rbm]

Thanks Ian.  I was not aware of the video Joe posted prior to him posting but have since reviewed it, as well as reading other sources for information on termocouples.  I'll see if I can locate another K-type thermocouple to check the one I have.

[MacMeter]

My question would be, assuming you are within the temp tolerances of the probe, would it be a bad idea to use it to test solder? How would you remove all of it from the tip sensor. This is the first DMM temp sensor I've owned, and I'm sure there must be some things one would NOT want to use the probe on, that could either ruin it or effect its accuracy the next time?

[joeqsmith]

The solder may not wick to the wire but I don't how accurate that would be.  Then again, I can't think of a reason why you would need to do this.   

When I made the above video, I showed the upper limit for the 235.  The probes jacket will melt at fairly low temps.  It's not Teflon.  The meters normally come with fairly low cost probes, so I would not be too concerned about damaging one. 

[MacMeter]

The solder may not wick to the wire but I don't how accurate that would be.  Then again, I can't think of a reason why you would need to do this.   

When I made the above video, I showed the upper limit for the 235.  The probes jacket will melt at fairly low temps.  It's not Teflon.  The meters normally come with fairly low cost probes, so I would not be too concerned about damaging one.

Thanks, good advice. I'll get a 2nd one for any situations where it may get damaged. The reason I mentioned the hot solder, was from another thread where it was mentioned as one way to test temp accuracy from that set on unit, to the actual tip temp. Testing boiling water and other such things, so I got curious how robust the temp probes and wires might actually be. I watched the video once, will watch it again, as your videos are well done, especially when you BLOW THINGS UP!

[wraper]

I first began to doubt the accuracy when I tested the tip temperature on my soldering iron (set at 350C but read 250C) and then my Weller hot air rework station (set at 650F but read 525F).

And this if how some clueless people check their soldering iron temperature, think that it's off and then screw up the calibration. They just touch the dry tip with thermocouple and think it's the way to do  . Thermocouple, unless ultra thin type specially made for soldering iron calibration, will always be much colder that soldering iron. Don't forget there is heatsink (wires) on the other end of the joint, which will draw away the heat.
EDIT: Also don't forget that thermocouple joint isn't flat but round, so just barely touches the tip, unless is put into a solder blob. Therefore heat transfer is not good at all.

[wraper]

Secondly, a simple probe such as the one supplied with the meter is only good to measure up to about 250°C (see here for example, or here).

If you heat the probe up above the maximum temperature indicated there is a possibility you could damage it, which would cause it to read inaccurately afterwards. You might like to obtain a new probe for comparison if you already heated up your existing probe above 250°C.

K type thermocouples are good to 1100^oC (and about 1250^oC short term) Only insulation / sheath is what limits the temperature to lower values. So unless you get a short somewhere, they are able to measure temperatures higher than rated completely fine.
http://www.pyromation.com/Downloads/Data/emfk_c.pdf

[Toolio]

I measured my Amtex CS18 solering iron with the Brymen 829s multimeter (Says in the manual it can read upto 1000°C ).
Used the thermocouple that came with it and put the tip of the iron and the tip of the thermocoulpe a blob of solder tin.
It ran upto 365°C. Would that be a correct reading?

The datasheet says it produces 390°C

https://www.antex.co.uk/products/precision-range-soldering-irons/cs18/

[wraper]

I measured my Amtex CS18 solering iron with the Brymen 829s multimeter (Says in the manual it can read upto 1000°C ).
Used the thermocouple that came with it and put the tip of the iron and the tip of the thermocoulpe a blob of solder tin.
It ran upto 365°C. Would that be a correct reading?

The datasheet says it produces 390°C

https://www.antex.co.uk/products/precision-range-soldering-irons/cs18/

It's uncontrolled firestick. Temperature can be whatever depending on mains voltage, room temperature, air flow around it...

[Brumby]

I always take temperature readings like this with a large grain of salt.

Before you even start to think about instrument accuracy you have to have the thermal coupling issue nailed down.  For any confidence that the sensor is reading the actual temperature of the target, I would want the target, the sensor and the immediate area around the two to be in thermal equilibrium.  If not, then any turbulence or temperature gradients will cause errors.

[Toolio]

My dealer advised against measuring a soldering iron tip with a multimeter and thermocouple combi. The multimeter can read tepms upto 1000°C but the thermocouple that came with it can only deal with temps upto 250°C. Higher temps could damage it. They advised me to get a FG-100 for this purpose. https://www.eleshop.nl/fg-100-soldering-iron-tip-thermometer.html

They also said that it is not that important to know the exact temperature, as long as it is hot enough to melt the leadfree tin and not set too hot. But i want to calibrate the solder station because it i saw on youtube that the readings can be very wrong (more than 20°C differnce).

[paulbt]

Hello everyone!

I have a question/curiosity that is a little bit offtopic, but not too much. I'm planning to buy a Brymen 235 and I was watching some review videos on youtube and I noticed that when you switch it to mV scale, it defaults to temperature. For me this behaviour is a little weird, so can you please confirm or deny this fact? Maybe the lastest ones have updated firmware and it defaults to mV as all the other dmms do?

[PA0PBZ]

It defaults to the setting it was on when switched off, either DC, AC, C or F, whatever was chosen with the [Select] button.