PDA

View Full Version : OT, RTD's (resistance temperature detector)



topct
07-14-2008, 12:11 PM
Would anyone know how to check or test an RTD? I think it might be the problem we are having with a heat press. The kind used to press designs onto t-shirts etc.

Also, would it be possible to fool the micro-processor that it is connected to, to determine if the board is the culprit?

I have disconnected it (the RTD) and read it's resistance and get about 50 ohms at room temp. The press goes through all the functions it is supposed to do when you turn it on. It's readout says that it is up to temp, but it is not. It does get warm and says it is ready to press, but it is not hot enough to press with. At that temperature the RTD reads about 65 ohms.

What I am trying to do is save a person over $400 for a new control board, RTD and the work involved to replace it.

What got me to thinking it's the RTD is that I found a broken connection on it one time when it was acting the same way. Fixed the connection and it worked fine. The connections are okay this time, but no joy.

Evan
07-14-2008, 12:33 PM
Otherwise known as a "thermistor". They do go bad. Take out the part if you can and see if it has a type number on it. Unfortunately there are many different kinds so it isn't possible to offer a generic "use this" suggestion. A Tee shirt press runs pretty warm, about the same as a clothes iron, so you need something reliable and with the same sort of control response as the original.

BTW, since the press says it is warm when it isn't I suspect the thermistor is positive temperature coefficient and the unit is open, or "mostly" open. A positive temp coefficient means the resistance goes up with temperature.

Mac1
07-14-2008, 12:34 PM
If you do a Google search you will find several RTD tables. None of them I looked at had anywhere near 50 ohm resistance at room temperature. If your meter is right, it sounds like the RTD is bad. I would check my meter first against something known.

Evan
07-14-2008, 12:49 PM
You can do a check on the temperature sensor by dipping it in boiling water. The resistance should go up quickly and significantly.

Swarf&Sparks
07-14-2008, 12:52 PM
that may be an LM335
data sheet here
http://www.national.com/mpf/LM/LM335.html

millwrong
07-14-2008, 01:27 PM
2-wire? 3-wire? 4-wire? Look at Omega electronics technical information on line. That might narrow down the answer you're looking for.

Swarf&Sparks
07-14-2008, 01:30 PM
Hell, just bypass it and use an electric frypan thermostat!

topct
07-14-2008, 02:19 PM
Thanks everyone for the suggestions. I need to get to the press a perform a couple of experiments.

Evan, when the press is cold I get the 50 ohm reading (suspicious). The press does heat up, but only to a point. When I measure the RTD then, it's 65 ohms. But like Mac1 says that 50 ohms doesn't match what the charts say for these things. They quote 100 ohms at what I assume to be room temp. So the RTD may be out of spec? If that were the case then I will assume that the controller is shutting down to prevent use.

The RTD is a two wire. There are no markings of any kind on it. The company that makes the press, Hotronics, says they were assembled in house but cannot or will not provide any specs. They will not sell a replacement. They say the new ones are different, yet the current wiring diagram matches the twenty year old press that I am trying to fix.

kmccubbin
07-14-2008, 02:20 PM
RTD's and Thermistors are totally different critters. An RTD is very difficult to read. Here's the pertinent info at Omega:
http://www.omega.com/temperature/Z/ResistanceElements.html

Kerry

Swarf&Sparks
07-14-2008, 02:39 PM
Sure that aint a K type thermocouple?
The thermistor (or etc) may just be cold junction compensation.
If It's just a T-shirt press, I'm serious, use a frypan t-stat.

Evan
07-14-2008, 02:41 PM
RTD's and Thermistors are totally different critters.

No, they are exactly the same "critters". Omega say so too.



Resistance elements come in many types conforming to different standards, capable of different temperature ranges, with various sizes and accuracies available. But they all function in the same manner:
...
Resistance Temperature Detector is a general term for any device that senses temperature by measuring the change in resistance of a material.


That is the description of a thermistor. I worked with machines that had a hot section running at the same temperature as a tee shirt press and performing the same exact function for over two decades. They all had some form of temperature detection and nearly all used some sort of thermistor. We didn't call them RTDs. That is most likely a simplification for the benefit of non-technical owners and operators of the machines.


The RTD is a two wire. There are no markings of any kind on it. The company that makes the press, Hotronics, says they were assembled in house but cannot or will not provide any specs. They will not sell a replacement. They say the new ones are different, yet the current wiring diagram matches the twenty year old press that I am trying to fix.

Mark them off the list of nice people to deal with. What a rip off. The company I used to work for was the same way. Unless you had the part number for something they wouldn't sell you the part. The parts didn't have the part number on them either. Even though the parts guy would most likely know the number for most items without looking it up he wasn't allowed to sell the part until you supplied the number. He was allowed to sell you a service manual for $100 so you could look up the number and order it.

In this case in order to proceed you need to know two things. What is the correct operating temperature of the press and is there a calibration adjustment on the controller board?

topct
07-14-2008, 02:58 PM
Thanks for that link Kerry. They do make RTD's that are 50 ohm at 0c.

It is an RTD. When this thing is working it's a sweetheart. The platen is constantly off and on the work and the temperature remains very constant. There is built in timer that counts down and signals when to lift the platen. The material and the heat setting glue they use does not like getting to hot.

I'm wondering now if the RTD is ok but the triac they use to turn the elements off and on might be the culprit. The press is about twenty years old. The triac is mounted directly above the heated platen (330 degrees). It is switching 1750 watts.

Swarf&Sparks
07-14-2008, 03:08 PM
NTC :cool:
http://www.designinfo.com/cornerstone/ref/negtemp.html

kmccubbin
07-14-2008, 05:39 PM
Thermistors act similarly to RTD's, but they do differ greatly in few respects. A thermistor is ceramic or polymer, an RTD is pure metal. The resistance change of a thermistor is very large and very non-linear. An RTD is very linear, but has only a very slight resistance change. For the purposes of blind troubleshooting, the amount of resistance change is important. Your resistances seem plausible for an RTD, but I'd expect more delta from a thermistor, along the lines of 200 to 300 % (or more???). An automotive temp sensor (thermistor) will be ~ 100 Ohms at room temp and several K Ohms at boiling (from my shaky memory). With an RTD, there's not too much that can go wrong, since it's just a chunk of metal with leads attached. I think you're on the right track looking at the switching side. If you find that the problem is in the controller, Watlow makes some reasonably priced units.

Kerry

topct
07-14-2008, 06:04 PM
Yes, one thing I noticed is that the RTD is very sensitive. You don't even want to breath on it or it will display a change.

You cannot touch it either, or it will show a difference in resistance. And the change is, as far as I'm concerned, is instant.

What I think I might try is a light bulb on the on the power connections to the elements. I can observe the on off. If the triac tended to fail open (it would be best to make them do that), that would be much easier and cheaper to replace than the RTD. It has a number on it anyways.

macona
07-14-2008, 06:42 PM
Might be simpler in the long run to get a PID temp control off ebay and retrofit the press to use it. The controls go pretty cheap.

kmccubbin
07-14-2008, 08:16 PM
Heating elements? Plural? Is there a possibility that one of them has failed, and that the temp is correct at the sensor, but too cool elsewhere?

Kerry

Evan
07-14-2008, 08:54 PM
I'm wondering now if the RTD is ok but the triac they use to turn the elements off and on might be the culprit.

If the triac failed the controller would not say the system is warmed up. It would either wait forever or if slightly more sophisticated declare a fault after a time out.


It's readout says that it is up to temp, but it is not. It does get warm and says it is ready to press, but it is not hot enough to press with. At that temperature the RTD reads about 65 ohms.

The first bet is still the sensor, second is the controller that receives input from the sensor. Of course the wiring between the sensor and the controller must be checked.

lwalker
07-14-2008, 10:18 PM
Technically, Evan is correct: a thermistor is a type of RTD. However, when most people in industry use the term RTD, they're usually talking about a platinum RTD which has a nominal resistance of 100 ohms, not a thermistor. This thing is almost certainly not that. Platinum is expensive and only used in high precision measurements; it's almost certainly a thermistor. Problem is knowing what the curve is.

It is unlikely to be a thermocouple; the resistance is too high.

J Tiers
07-14-2008, 11:02 PM
Just so happens we were recently in the midst of RTDs and thermistors at work.

A 50 ohm at 0C (or 25C) RTD is obviously not impossible, but is certainly not typical.

It appears that if it were a platinum RTD, which it may not be, it might have about a 4.5 ohm resistance change from 0 to room temp (25C).

The final temp of 65 ohms suggests that if a platinum RTD, it might be reaching only around 80 degrees C, which seems too low, but I suppose might be OK. I don't know much about T-shirt presses......

If it is a copper RTD, nickel, balco, or tungsten unit, the coefficient is higher, and the final temp is thus probably LOWER than 80. Since you measure 50 at 25C, and copper etc may be specified with the resistance at 25C, instead of 0 deg C, it very likely is a non-platinum type. Not to mention that they are probably cheaper and t-shirt presses are not lab equipment.......

http://www.omega.com/temperature/Z/pdf/z252-254.pdf

It sounds as if the controller could be bad in some way, and the RTD is correctly reporting that the temp does not get up there.

If there is a triac, or two SCRs, maybe one "direction" is not being correctly triggered, and the heater is receiving only half-waves.

If you had a clamp-on current probe and scope, or even a clamp-on meter, you could discover what current (and with scope what waveform) was going into the heater.

EVAN:

While both RTDs and thermistors have similar behavior, and are "technically" the same ilk of device.....

But when they are used practically, they are different animals, hence the distinction in names.

Most "thermistors" are non-linear in a significant way, linear only over a relatively small temp range, while most RTDs are, by contrast, pretty linear over a wide range (they DO have a correction if you need fractional degree accuracy).

Thermistors can also be NTC pr PTC, meaning they can be made to go up OR down with temp, which the usual RTD does not.

RTDs are generally metal wire, platinum for instance, while most thermistors of the general varieties are "semiconductor" materials.

RTDs tend to be 100 ohm or sometimes 1000 ohm, while thermistors are often higher, 10K is a typical base resistance.

Thermistors are often a LOT cheaper than RTDs, but you can still pay hundreds of $$ for extra good calibrated types, just as with RTDs.

Evan
07-15-2008, 01:24 AM
I merely pointed out that the web site given above explains that all the devices work in the same manner. It makes a RTD and a thermistor a distinction without a difference. We don't consider different types of transistors to be Not a transistor just because the construction and materials vary. A transistor is defined by what it does, as is a thermistor, not what it is made from. A thermocouple, BTW is an entirely different device and doesn't depend on resistance change to operate. Another example is the LED. LEDs use a variety of different materials to achieve the same result, photonic emission. They are all LEDs regardless of what they are made from or what wavelength they emit.

The same applies to the resistor, the capacitor, diodes, lasers etc. Electronics parts are categorized by function first.

Tee shirt presses operate at around 350 to 400 F.

darryl
07-15-2008, 05:15 AM
There's often a calibration control or adjustment pot of some kind in this type of sensing circuit. All it would take is a bad solder connection here, or an intermittent on this pot to affect things. If you fnd such an adjustment pot, carefully mark its current position, then tweak it back and forth a bit, then leave it at the original position. Test the unit again to see if there's any change in how it's working. Another way you might be able to test this pot is by measuring the sensor resistance while it's hooked into the circuit (power disconnected first please). If you get an erratic resistance reading while tweaking on the adjustment pot, it may need cleaning, resoldering, or replacing.

You could also simulate the operation of the sensor by wiring in a variable resistor in place of the sensor. You're talking about a range of 50 to 65 ohms where the unit either heats or thinks it has reached operating temperature. You can simulate this range by hooking up two resistors in series, say two 47 ohm resistors, wire that in place of the sensor leads, then alternately short then unshort one of the resistors. You should be able to definitively test the power circuit and the readout circuit, as the control circuitry will think it's either too cold or past operating temperature, and you can measure whether voltage is being sent to the heater or not.

Another trick I've used a lot is to have a small light bulb plugged into the same socket as the unit to be tested. With the wattage rating of the heater in this unit, you'll be easily able to see if the heater is drawing current by the slight dimming of the bulb. By simulating the sensor with the two resistors, you'll be able to either get a direct and repeatable heater function, or it won't follow. This will tell you if there's a problem in the power circuit or not.

I have no idea what the resistance of the sensor should be, but if it has changed, then it's not likely to remain stable. On the same hand, if it gives repeatable readings after several measurements at some temperature extremes, like say ice water and boiling water, then it's unlikely that it's bad. If it's a typical thermistor of either ntc or ptc (temperature coefficient), and it's cracked internally, the resistance readings would be higher than normal. It sounds like the sensor is a ptc, thus the resistance goes up with increasing temperature, and a cracked disc would definitely cause the control circuit to think the temperature is high enough when it's not.

J Tiers
07-15-2008, 08:09 AM
On the other hand, a cracked disc (for a "thermistor") would probably not give a 50 ohm reading.

The situationseems to be that there is an RTD, which is an odd value, but which seems to be giving information that concurs with the observed performance....... The press is not hot enough to work, and the RTD is reporting a temperature that is below what would be expected to work.

So, the most likely problem is that the sensor is OK, but the controller is not.

If the controller "triac" is bad, which is naturally possible, you would see that the controller never would report a good temp, it would always be calling for more heat, but would never get it.

So, since I see on re-reading the OP that the readout actually reports a good temp, meaning that the triac (etc) is probably OK, there are a couple possibles.

1) the RTD is fine and the controller interface to the RTD has a problem that makes it falsely read a higher temp than is actually present.

Since the readout is normal, it is unlikely that anything after the point that the readout comes from is bad. The problem is apparently that the controller interprets the RTD wrongly.

2) The RTD is functional, but has had a "shift" that makes it read a higher resistance than normal. That would make the controller see conditions for a higher than actual temperature.

The only way to distinguish these two is to either find the problem in the controller, or get information on the RTD.

Since the RTD info is apparently NOT going to be provided, possibly because the manufacturer does not know anymore, you are stuck with locating a problem in the controller.

Simple solution........ may even work.

Get a potentiometer, one with a total resistance of 100 or 200 ohms.

Wire it in as if it were the RTD, and vary it. See what it takes to get a "good" reading, and verify that the reading changes as expected. if you can get any clue what sort of RTD it is, that will obviously help, since you can use the generic curve for it to identify approximate resistances, and can set a value that should read correct, and slightly above and below that, checking to see what is actually read.

A really low resistance thermistor is less common, but certainly quite possible.

Another way to detect what it is and test it is to check it against a thermometer, and see if the curve is reasonably linear from ice water to boiling. if so it is probably an RTD, and you can calibrate it, then see if the display reading corresponds. While it may not read at ice water, the display should cover the boiling water area...... see if the temp is right.

if the RTD seems to behave right, but the display is wrong, good chance the display IS wrong, and problem is in the interface

topct
07-15-2008, 08:13 AM
O-boy, this is telling me I really need to get this thing on my bench. The press is across town and the person is reluctant to have me move it because it is so heavy. I think that armed with all this information I can get them to let me take it to my shop.

The company that makes the press calls the sensor an RTD. As to what exact type, who knows? It does change it's resistance very consistently. I can remove it from the warm press and it instantly goes back very close to where it reads when cold. When I put it back in it goes right back to the reading I got before removing it. I did this twice and the reaction was the same.

I am going to print out all the suggestions, and as soon as I get it here, I will attack it. The controller board have a couple of trimmers on it to twiddle with. I have an amprobe. I have light bulbs. I have hammers.:D With its cover off everything is exposed for poking and prodding.

This press has a lot of features built in that are mostly not needed. It can turn itself on and off at preset times, there are several preprogrammed heat and timer settings, it has a pressure reading on it (although there is no pressure sensor?). All not needed. And accept for the platen timer they are not used.

So if it is a worst case, an after market controller would be just fine. And certainly much cheaper than their $400 replacement.

Thanks for all the trouble shooting ideas. I had feeling this would be a good place to ask for help. The company that made this thing provided none.

topct
08-12-2008, 03:44 PM
I have worked on this machine before. Once to replace a mechanical part, and a couple of other times to repair the connections to that sensor. I don't know why, but it seems to deteriorate over time. A bad connection has always been visible. This time it was not. I know I checked it visually and this time it really did seem secure. I could measure continuity between the connection and where it plugs into the controller board, and I could measure the resistance of the sensor itself, at the sensor, but the connections must have been bad.

I cleaned and resoldered the wires and the crazy thing came back to life. :D

Thanks everyone for the suggestions for trouble shooting and I have them filed away just in case next time it might not be so simple.

I don't know why these connections go bad. It seems like some kind of corrosion occurs between the wire or wires and a new solder joint fixes it. Even though it can look and feel solid, it breaks down.

Evan
08-12-2008, 06:14 PM
It's a very common problem on boards that heat cycle. It especially affects heavier leads from power connectors and similar but can affect any solder joint. The connectors have a different coefficient of thermal expansion than the boards and the solder so eventually microcracks develop that surround the connector pins. They are nearly impossible to see unless it has been arcing in which case it turns black. You need a microscope to see the crack if it is clean. You can't really feel it by tugging on wires or connectors.

I used to take the power supply boards out of various machines and resolder all the heavy pins as a preventive measure. It saved me a lot of service calls over the years as could be determined by my much lower usage of various power supplies than most techs. Copiers were especially bad for that because of the hot section in the fuser which ran at the same temperature as your press and heated up the entire machine. Similar circumstances, similar problems.

[edit] A lot of the problem is originally caused by poor wave soldering on the heavy pins. It's hard to get a good joint unless you preheat the boards which is an extra time consuming step that reduces throughput at the factory.

wmgeorge
08-12-2008, 09:37 PM
I have worked on this machine before. Once to replace a mechanical part, and a couple of other times to repair the connections to that sensor. I don't know why, but it seems to deteriorate over time. A bad connection has always been visible. This time it was not. I know I checked it visually and this time it really did seem secure. I could measure continuity between the connection and where it plugs into the controller board, and I could measure the resistance of the sensor itself, at the sensor, but the connections must have been bad.

I cleaned and resoldered the wires and the crazy thing came back to life. :D

Thanks everyone for the suggestions for trouble shooting and I have them filed away just in case next time it might not be so simple.

I don't know why these connections go bad. It seems like some kind of corrosion occurs between the wire or wires and a new solder joint fixes it. Even though it can look and feel solid, it breaks down.

I worked on temperature control systems for a number of years, RTD's rarely go bad. Someone else on here hit it on the head the difference between an RTD and Thermistor. Good job, I'd say bad connection was the problem.