PT100 vs PT1000
-
@Corexy The advantage of the Pt sensors over thermistors for wide temperature ranges is that the temperature coefficient is very well defined, and fully standardized across all the sensors. Thermistors have varying beta values, and when used over a wide temperature range can give significantly different results between two nominally identical units. In my day job, I work in a ver highly temperature-controlled lab with 0.01C regulation, and logging to 0.001C. This is a US Government standards lab. For this, we use calibrated thermistors because of their very high sensitivity. They are perfect for narrow-range applications. However, to cover a few hundred C or more, they are far from ideal. Although a Pt sensor is less sensitive (0.3%/C at room temp vs. typically 6%/C for a thermistor), they are very linear and interchangeable. A thermistor has too big a coefficient for wide ranges; a 100k thermistor at 25C is < 100 ohms at 280C. That's a huge dynamic range to cover. One the other hand, a Pt1000 is 1k at 0C and 2k at 273C. This is an easy range to digitize, and the resistance is high enough that modest-length wires don't affect the value too much.
The Pt100 sensors are well liked for heavy industrial and high precision applications because they are fairly robust, and the very low impedance makes them relatively noise resistant. On the other hand, the signal levels are low, and you must use a four-wire Kelvin connection unless the leads are very short, since the wiring resistance will be a big contribution to the total resistance. For high-precision work you can't even correct for the wiring resistance in a 2-wire connection, since you don't know the temperature of the wires. -
I was gonna say thermistors are cheap and "good enough" for most printers as you can tune the settings to achieve good (enough) accuracy and precision. PT100 are accurate, precise, reliable and pricey - PT1000 somewhere in between, closer to PT100 than thermistors, as evident by the price.
But that guy
had abit more meat on the reasoning.For even higher temperatures, thermocouples are the way to go.
-
@bearer @dc42 @deckingman @mendenmh
At this point I'm pretty sure I'll go with this PT1000:
https://www.aliexpress.com/item/32859917151.html?spm=a2g0s.9042311.0.0.401b4c4duXbwYr
I like the fact that it's got a 1m cable length and only needs 2 wires, as I'd rather have it run all the way through to the crimps on the plug at the board end in one piece. I'll leave the resistors as is for now, and if I'm worried I'll get a temp probe for my multimeter and see how bad it is actually out. I've got a couple of fried boards here, so I can actually have a practice run on resoldering the resistors if it is an issue. It seems it should be comparable or better than the thermistor as is, so that will do for now.
I'll also set one up in the enclosure as well, for exhaust fan control.
Thanks for your advice, and for enabling my obsessive compulsive disorder. I really enjoy the pre planning stage, it's a part of the fun for me.
-
@Corexy Precision Piezo sells a PT1000 also and it has 200cm leads. Here is a link for the product on Filastruder.
I have these working well on a Duet 3 and Duet 2 Ethernet and the temperatures appear to be very close. When both printers use the same temperature and filament they have very similar results.
-
@mwolter said in PT100 vs PT1000:
@Corexy Precision Piezo sells a PT1000 also and it has 200cm leads. Here is a link for the product on Filastruder.
I have these working well on a Duet 3 and Duet 2 Ethernet and the temperatures appear to be very close. When both printers use the same temperature and filament they have very similar results.
Thanks mate, I've already ordered the one above and it has 1m of leads which should be enough as I'm thinking of mounting the board on the back of the machine rather than under it. If I do go with the board under the machine, I'll order one of these for the extra lead length.
Cheers for the tip!
-
I have tested both piezo and triangle labs in the past.
Piezo worked great at low temps, but at high temps the wiring insulation and potting compound start to degrade, it seems like some kind of epoxy-based(?) potting compound used. I heated it up to 400C and it all melted and the sensor came out of the block. This is not a product defect as it indicates on the product page it's only good to around 200C, but if you want to try high temps it's something to think about. Wires are very small and flexible, easy to manage.
The triangle labs performed much better, the potting compound is a better ceramic type material and it held up at 450 without a problem.
Precision on triangle labs was adequate, got 8 degree spread between 5 sensors all checked at 450C, which is less than +-1% from the median reading. Couldn't check accuracy as I don't have a calibrated temp probe.
Triangle labs wiring is a lot stiffer. It's stranded but the strands are pretty thick. I wouldn't feel comfortable running it in a constant flex with a smaller radius, but I haven't tested this. (I have a connector on mine near the hot end and run it through more flexible cables back to board)
Tested with about 5-6m of cable, standard 4.7k resistors on duet2 board. I had more than half an ohm resistance in the wires so I bumped it up to R4701.
-
@nhof
Thank you for the detailed info. I've ordered a few Trianglelabs PT1000 RTD's and will test them. -
@nhof said in PT100 vs PT1000:
I have tested both piezo and triangle labs in the past.
Piezo worked great at low temps, but at high temps the wiring insulation and potting compound start to degrade, it seems like some kind of epoxy-based(?) potting compound used. I heated it up to 400C and it all melted and the sensor came out of the block. This is not a product defect as it indicates on the product page it's only good to around 200C, but if you want to try high temps it's something to think about. Wires are very small and flexible, easy to manage.
The triangle labs performed much better, the potting compound is a better ceramic type material and it held up at 450 without a problem.
Precision on triangle labs was adequate, got 8 degree spread between 5 sensors all checked at 450C, which is less than +-1% from the median reading. Couldn't check accuracy as I don't have a calibrated temp probe.
Triangle labs wiring is a lot stiffer. It's stranded but the strands are pretty thick. I wouldn't feel comfortable running it in a constant flex with a smaller radius, but I haven't tested this. (I have a connector on mine near the hot end and run it through more flexible cables back to board)
Tested with about 5-6m of cable, standard 4.7k resistors on duet2 board. I had more than half an ohm resistance in the wires so I bumped it up to R4701.
Hmmm...I noted the description in the link above for the Piezo states they work over a wide range of temperatures and are rated for 400C in 3D printing applications. It seems they are referring to the wire about the 200C Supplied with 200cm high-temperature wire (rated for 200C), and fibrglass sleeving for use near the heater block.
I am hoping that's the case as I ordered a pair of these too, at reasonable expense shipped to Australia...have I wasted my money?
-
The reason I prefer PT100 and PT1000 sensors for wide temperature ranges is that a thermistor has great resolution and moderate accuracy around the centre of its temperature range, but poor resolution and accuracy at either end. There are two reasons for this:
The variation of resistance is just too great to measure accurately using a fixed series resistor. Accuracy and precision degrade when the ratio between thermistor resistance and series resistance gets too high (in either direction). A typical 100K thermistor varies between about 220K at 10C and 85 ohms at 300C. Using a 4K7 series resistor (as on Duet 2) this gives a ratio between 47:1 and 1:55. The Slice thermistor goes from 943K at 10C to 101 ohms at 500C. That's a ratio of 200:1 to 1:47.
The ADCs in microcontrollers are not as good a dedicated ADCs. They have good resolution (12 bits on Duets) but significant gain, offset and linearity errors. So the combination of a large ratio of thermistor resistance to series resistor and the gain error of the ADC makes it very hard to read the thermistor at low temperatures - even with the automatic ADC calibration that most Duets have.
PT1000 sensors have a resistance in the correct range to get the best from the ADC, especially wheh using a lower value series resistor (we use 2K2 on the Maestro and on Duet 3). The resolution isn't as good as for a thermistor in the middle of the temperature range, but it's fairly constant across the whole temperature range. So it beats a wide-range thermistor at high and low temperatures.
PT100 sensors use a daughter board that has its own precision ADC. So they provide high resolution and high accuracy, provided that you don't spoil the accuracy by using a 2-wire connection with significant resistance in the leads. BTW if anyone needs high accuracy and resolution but wants to use a PT1000 sensor, it's possible to convert the PT100 daughter board to read PT1000 instead. But for most users, the resolution of a PT1000 connected to a thermistor input is good enough.
@mendenmh made some additional good points in his post.
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42 said in PT100 vs PT1000:
The reason I prefer PT100 and PT1000 sensors for wide temperature ranges is that a thermistor has great resolution and moderate accuracy around the centre of its temperature range, but poor resolution and accuracy at either end. There are two reasons for this:
The variation of resistance is just too great to measure accurately using a fixed series resistor. Accuracy and precision degrade when the ratio between thermistor resistance and series resistance gets too high (in either direction). A typical 100K thermistor varies between about 220K at 10C and 85 ohms at 300C. Using a 4K7 series resistor (as on Duet 2) this gives a ratio between 47:1 and 1:55. The Slice thermistor goes from 943K at 10C to 101 ohms at 500C. That's a ratio of 200:1 to 1:47.
The ADCs in microcontrollers are not as good a dedicated ADCs. They have good resolution (12 bits on Duets) but significant gain, offset and linearity errors. So the combination of a large ratio of thermistor resistance to series resistor and the gain error of the ADC makes it very hard to read the thermistor at low temperatures - even with the automatic ADC calibration that most Duets have.
PT1000 sensors have a resistance in the correct range to get the best from the ADC, especially wheh using a lower value series resistor (we use 2K2 on the Maestro and on Duet 3). The resolution isn't as good as for a thermistor in the middle of the temperature range, but it's fairly constant across the whole temperature range. So it beats a wide-range thermistor at high and low temperatures.
PT100 sensors use a daughter board that has its own precision ADC. So they provide high resolution and high accuracy, provided that you don't spoil the accuracy by using a 2-wire connection with significant resistance in the leads. BTW if anyone needs high accuracy and resolution but wants to use a PT1000 sensor, it's possible to convert the PT100 daughter board to read PT1000 instead. But for most users, the resolution of a PT1000 connected to a thermistor input is good enough.
@mendenmh made some additional good points in his post.
Thank you David.
I would be interested in hearing how to mod the daughter board, just for interests sake. At least then if I bugger it up I can just plug straight into the Duet wifi board and live with the resolution I get. I'd be much more comfortable taking a chance with the little daughter board I wasn't going to use anyway.
So all future boards/models/developments will use the 2.2k resistor to be compatible with PT1000's, is that the way you are going?
-
@Corexy said in PT100 vs PT1000:
I would be interested in hearing how to mod the daughter board, just for interests sake.
The mod is to remove the 400 ohm reference resistor and replace it with a resistor of around 4000 ohms, e.g. 3900 or 4300 ohms.
-
The triangle lab sensors have woven fiber insulation all the way to the sensor, similar to the e3d thermistors / heaters if you're familiar with those.
I think the woven insulation sleeve near the block should help on the precision piezo, but I'm skeptical of whether you could run them at 400c without problems. Note that I ran it up to 450C, not 400 so I may have gone over-temp in any case. (i double checked my numbers vs my last post)
E3D recently released a PT1000 sensor which looks quite nice, but unfortunately they're out of stock everywhere.. I need to get my hands on one of those to test and compare..
Here's a pic of the piezo one. You can see the woven insulation held up fine, but the polymer insulation / potting compound burned up, and consequently the sensor itself (white bit) came out of the metal casing when I tried to remove it from the heat block.
-
Wow, that's brutal!
Never mind. I've got all manner of bits coming and an E3D PT100 here as well.
I feel that 350C would be as high as I'll be going possibly for PC.
