EEVblog Electronics Community Forum
Electronics => Metrology => Topic started by: View[+]Finder on January 31, 2021, 06:20:25 pm
-
It is not hard to build a makeshift oven styrofoam box (few resistors, simple temperature controller with relay) and put a ref in there. Total cost could be as low as $5-$10 and suitable to keep temperature stable.
Next step is active thermostat box with some ready-to-go controller, such as Arroyo (example (https://xdevs.com/article/tecpak585/)) or ILX 5910B (one was sold for $80 (https://www.ebay.com/itm/ILX-Lightwave-LDT-5910B-99-to-150-C-Auto-Tune-Temperature-Controller/333848804885)). Add some cheap chinese 40W TEC, heatsink, RTD from Digikey, cooler for hot side and getting stability better than 0.01°C become trivial for small device like A9 3458A reference.
Excellent suggestion!
Coincidentally, I'm a participant in Mark Rober's 'Monthly Classroom' online course and have a project underway for just what TIN described: http://mthly.co/p/gXkRm2 (http://mthly.co/p/gXkRm2)
There’s a photo attached of a little (150mm) aluminum box for the Thermal Stability Fixture now sporting a spiffy backpack that many will recognize as a CPU cooler salvaged from a long departed PC. The fin part is mounted on a thick aluminum plate (5mm) and between the fins and the plate is the Peltier junction. That clever device can get cold on one side and hot on the other--you can see where this is going--the other photos are of the inside warmer than the fins on the outside.
Next will be to add the thermal conductive layers around the Peltier to even out any surface imperfections and run a temperature test.
-
Metal box? Metal is rather poor thermal insulator choice. ^-^
Otherwise good start, just don't forget that Peltier device is NOT a cooler but a heatpump with poor efficiency.
P.s. I still see no point in unneccessary informational noise "Making Do with What You Got" in every thread title. ;)
-
You cant realiably measure the surface of unpainted aluminium via IR. This is due to IR reflections of surrounding things from the metal surface.
You may also need to adjust the emissivity setting on your IR temperature gun.
-
For temperature control--rather than using a relay and switching the Peltier on/off--I'm thinking that varying PWM in response to temperature changes might avoid the hysteresis issues with on/off and also eliminate whatever the relay opening might generate. The Peltier I'm using is listed on Amazon as a TEG, however reviews suggest that it is a TEC. In any case, it draws 1.5A with an applied 12VDC from a bench PSU, dropping to 5VDC under load.
In actual operation, the power for the Peltier would come from a source that doesn't make the reference useless from its contributed noise. Perhaps a battery backed by a charger that is turned off when the reference is being used.
Suggestions?
-
Metal box? Metal is rather poor thermal insulator choice. ^-^
Otherwise good start, just don't forget that Peltier device is NOT a cooler but a heatpump with poor efficiency.
P.s. I still see no point in unneccessary informational noise "Making Do with What You Got" in every thread title. ;)
There will be foam insulation inside the box--later possibly on the outside as well. The box has been hanging out on my work bench for over a year--it was to have been part of an educational experiment for a kids radio telescope to investigate HII spectra in our galaxy. COVID put that on hold.
Peltier junctions and efficiency don't even live in the same country. I am hoping that in practice Peltier will just be managing stability of temperature around variations in my lab, however I do not harbor high hopes. If anyone wants a really efficient heat transfer, check out 'ammonia refrigeration cycle'.
-
Peltier and high efficiency don't work that well, but for an occasional test efficiency is not that critical. The temperature range is limited, but should be OK for the usual near room temperature tests.
Controling a peltier cooler with PWM is however a poor idea: it is quite some stress to the TEC and the efficiency is quite bad - just a DAC and linear control is higher efficiency for cooling: This way the extra power is burnt in the power transistor and not in the TEC, where it interferes with cooling. For higher efficiency a kind of switched mode controller can be used.
For the outside one usually needs a massive heat sink, likely with a fan.
-
ammonia refrigeration cycle
I used to build refrigerators for computer CPUs and GPUs back in my student years, with temperature points around -40°C with heatload ~300-500W. High-power PWM drivers will also spew lot of crap around, so you would need well shielded wiring and lot of other mitigations to make it useful. A9 board will go nuts if you have PWM power source running nearby :D
But cycling system is hard to get stable uniform temperatures. As of suggestions, use metal box for something else and build larger foam insulated box instead. You'd want to have some wiring, stirring fan inside and other toys to be measured eventually, not just little PCB or A9 module, which is only thing that will fit into box you show. I should actually order up some sheets and build new big box myself. :popcorn:
-
For a quick result, I would consider a ready made TEC based cooler box for the car. So the mechanical part is essentially already done, just add a sensor, maybe some metal inside for a more uniform temperature and the regulator.
-
ammonia refrigeration cycle
I used to build refrigerators for computer CPUs and GPUs back in my student years, with temperature points around -40°C with heatload ~300-500W. High-power PWM drivers will also spew lot of crap around, so you would need well shielded wiring and lot of other mitigations to make it useful. A9 board will go nuts if you have PWM power source running nearby :D
But cycling system is hard to get stable uniform temperatures. As of suggestions, use metal box for something else and build larger foam insulated box instead. You'd want to have some wiring, stirring fan inside and other toys to be measured eventually, not just little PCB or A9 module, which is only thing that will fit into box you show. I should actually order up some sheets and build new big box myself. :popcorn:
Back in my "student years", computers had raised floors and needed more cooling than the school library. LoL
Right, PWM is out. My testing has been done with a bench power supply, so perhaps a constant current power source with temperature control? My intent is to have the little box for small stuff like the A9 board. The cycle of ambient temperature is my present challenge. I built the little box to see if it can stabilize temperature. Depending on time of day and what work is in progress, heat load from meters will vary greatly.
Home Depot stocks those pink polystyrene boards so knocking out a proper thermal chamber as you describe will be a good near term project. Right now, space is at a premium.
-
An aluminum case is good for the inside, as it can reduce thermal gradients inside. One may still want some extra surface / "Heat sink" were the TEC is attached. The outside should be isolated reasonably well, so some 2-4 cm PS board sounds reasonable. No need to overdo, as the TEC itself is a significant thermal path to the outside.
If one is really after mK stability, it helps to also look at the temperature of the outside half of the TEC and correct as feed forward. The TEC can be a major path for disturbance through a varying outside temperature. So it is also important to have quite a large / good heat sink at the outside. For protection it may be good to measure the temperature anyway.
Driving the TEC with constant current or constant voltage is OK, it does not make a big difference as the TEC behaves essentially like a resistor with only a little (AFAIR some 10-20 mV/K) added thermal EMF. The nominal current of the TEC is the maximum useful one under optimal conditions. Real world (especially limited heat sink) the current for best cooling would already be lower. So for the design I would plan with a maximum of about 50-80 % of the nominal current, no need for more. Only the heating case could use a little more, but likely not needed, as the upper temperature is limited.
-
Photos of my little box before testing.
It was connected to a bench PSU overnight supplying 1A on CC mode at about 3V. Three watts of power seems reasonable regardless of efficiency. Temperature decrease from 27C to 26c approximately. There was no control of temperature, so no surprise, the measured temperature decreased more or less in line with ambient. The good news is that the decrease was linear so adding a means of adjusting CC based on ambient should be workable.
The ambient in the lab is 21.6C at present and the box PSU is set on 960mA at about 2.8V for 2.7W into the Peltier. The temp is measured by a 10K thermistor in an aluminum enclosure inside the thermal box. The 34465 is recording the thermistor output. The task today is to see what the relationship is between ambient and box with no adjustment to the settings
-
Now for something completely different . . . keeping your pet Iguana warm
A major part of "Making Do" is finding alternate means of reaching one's objective. {For example, I sailed a very long way (Tonga to SF CA) using PVC water pipe as battens for the mainsail--there being nothing else available.} In the photos attached are a thermo controller and infrared heating pad designed for starting plants, brewing beer, hatching eggs and keeping lizards comfortable.
Sold on Amazon for a pittance in comparison to a genuine electronics thermal chamber. Considering that TIN has 'egged me on' to building a bigger, better styrofoam box, what is not to like about the gadgets in the picture? (from an electronics POV)
I'm guessing that the controller might be hackable and that any relay inside is solid-state. What about the infrared pad? Some kind of LED's from the photos on Amazon. Surely they are not running on mains power? Perhaps the whole rig could be made to run on DC?
UPDATE:
Not what I would call hackable after all. Only regulates +/-2C even though shows ##.##C. Full mains voltage with standard (NotSS) relay. Only regulates heat. Back to Amazon . . .
-
Here's a link to the video of the Thermal Stability Fixture:
http://mthly.co/p/PQHprX (http://mthly.co/p/PQHprX)
It was built as a part of Creative Engineering with Mark Rober on the Monthly Classroom (Monthly.com)
It is finished! (almost)
It works! (temperature is controllable by setting Amps)
I tested over a 25-hour period and measured the temperature inside the Fixture and the ambient temperature in the room. The Fixture was able to maintain 2 degrees Celsius above ambient over the test. Ambient temperature in the room ranged from 25.2C to 27C, being cooler at night and warmer at noon.
Using a Keysight 36313 bench power supply for power for the Peltier junction, with current limited to 960mA to establish the desired temperature, the Peltier functioned on less than 3 watts.
Further testing will determine the maximum reliable temperature for operational use in the lab.
Testing revealed that the small size of the enclosure and the inherent quantum nature of the Peltier junction made the Fixture very sensitive to small changes in supplied current. Further study into alternative methods of controlling temperature will be required.
-
The curve show now compensation of room temperature variations and a rather fast response.
The maximum reliable temperature is better not determined by experiment, but from looking at the TEC datasheet and than some math. There is ususlly an upper temperature limit allowed for the TEC and this setts the upper temperature limit. The lower temperature limit can be found by experiment. So measure the final temperature or temperaure difference to room temperature (may be more accurate) for a few currents (e.g. 40%, 60% and 80% of I_nom) and than fit a parabola to the temperature difference versus current to find the maximum useful current. This will nusually be somewhat less than the nominal current givebn for the TEC, as the heat sink is usually not perfect. It is still a slow process.
Alternatively start at some 70% of the nominal current and wait to reach near final temperaure and than do small steps in the current to see if the temperature sinks or goes up over a few minutes. Initial the temperaure may still go down with a higher than optimal current as the heat sink needs time to heat up. Relevant is the later change with the heat sinks in equilibrium.
-
Testing TEC heating capability:
The goal is not to find how hot it can get, rather it will establish by stages what to expect. The limit on the TEC (claimed to be a TEG) is 1.8A, so that is as far as I would go. It was initially tested at 1.5A just to see the result and it certainly made the DUT warm to the touch and had an effect on the output from the DUT.
To have some fun with the Keysight PSU programability and the logging on the various DMM’s, I’ll probably set up an automated test protocol. Maybe step the amps up from 800mA to 1500mA by 100mA increments.
Results posted here, of course.
-
Calculation is good way of course, but better way is test thermal response of system with some controller and tune it bare metall.
But i don't see any thermal response tests in this thread.
For small boxes you can use with fully hardware controllers like as MAX1978 with fast thermal response on TEC->Metal->NTC way for high thermal resonant frequency 20...300 mHz.
Big boxes with slow and weak thermal response as example TEC->Gun->Air flow->Massive metal box->NTC provide very low thermal resonant frequency 0.1...20mHz, full hardware PID can't regulate it directly. It can be use with only software P.I.D controller.
I don't see direct contact between DUT and TEC on yours pictures :o is slow and weak thermal response, be careful!!!
TEC->Still Air->Massive metal box->NTC may be extremely low thermal response in sub mHz range.
PS. Still air - is best ever thermal insulator!
This is work in progress. TEC is mounted to thick (6mm) aluminum 'plate', the other side of which is open to the interior of the box. Insulation is inside the box and some also on the cold side of the AL plate. Box is for durability and TEC and plate are mounted to it with HDPL washers to minimize contact between box and plate. Inside the box is "still air" with DUT inside another metal box. (Is this a bad idea? Should DUT be in contact with heat plate? I don't think so, but???)
Yes, there must be a controller and your graph shows excellent stability. Also, your cat. What a view from the top[!
-
Some thoughts.
For temperature stability thermal mass is good. For uniformity of temperature good thermal conductivity is beneficial. So the aluminum box has its place. Particularly if it has a large wall thickness. It is analogous to a low pass filter in electronics for both spatial and temporal variations. You make a secondary chamber around the aluminum box and control that temperature. The drawback is a long time to reach stable temperature, but for an application like this it isn't a big problem.
Another way to get passive stability is to use the high heat of melting of appropriate solids. It is analogous to using a zener diode in electronics. And it is why an ice bath is one of the standards for temperature references. It takes a lot of heat flow to move an ice bath away from the freezing point. Glauber's salts and paraffin can perform similar duty at temperatures more appropriate for a lab voltage reference.
-
No, metal box here hurts the purpose on all counts. One need decouple inner volume from external thermal variations (e.g. increase thermal resistance between inner volume and outer world).
Otherwise poor peltier is just wasting little useful work it provides to pump unwanted heat sneaked into inner volume. Same reason why we have multiple walls and thermal insulation in house walls/windows :)
-
A small Aluminum box is good, if it is at the inside. The design can be a little different for a small box and a large box: The small box could be from the inside out:
Aluminum case with minimal electrical isolation on the inside (e.g paper on the bottom)
TEC at one point and Idolation around
Good heatsink , likely with fan
For mechanical protextion some case around the isolation may help, but not really needed.
With a large box, e.g. to test a whole DMM, one may want a extra Fan at the inside and maybe a heat sink to couple the TEC stronger to the air flow.
With a fan the inside case no longer needs to be highly thermal conductive.
For the control is help if the temperature sensor reacts fast to the TEC. A fast reaction makes the control easier. The relevant part is the dead time, not so much the rate of rise.
The problem is however than than the measured temeprature is no longer the actual DUT temperature. With digital control one could use a 2nd sensor closer to the DUT. The sensor at the DUT would provide some of the P and the I part. The sensor closer to the heater would provide some of the P and the D part of a PID control. Alternatively have the sensor at the DUT just for a check and measurement.
For digital control I would also measure the heat sink temperature, as this can be the main path for disturbance coming in. The power from the TEC depends on the current and the temperaure difference. Including the outer temperature can improve the regulation, by handling the effect from the outside in a feed forward mode instead of only by the regulation.
-
First off, thanks for helpful advice and just to be clear, I'm not suggesting that building what I did is a good idea for all purposes. Clearly, what I have is suitable for only the smallest of DUT's and maybe not even good enough for that.
Now for the "still air" aspect of thermal transfer. "Heat transfer is classified into various mechanisms, such as thermal conduction, thermal convection, thermal radiation" (Wikipedia) Yeah, we all know that, right? However, unless there is a difference in temperature between regions, there is no transfer to worry over is there? In my situation, there is maybe 5.0 C max difference between ambient and what is needed to maintain constant temperature. As configured, I am heating the interior, not cooling.
"Still air" is a factor in convection, true, but a quick touch on the tip of your soldering iron will confirm that conduction is the king of heat transfer. In my little box, the Peltier transfers heat by conduction to a relatively massive AL plate. The plate radiates heat (or if the inner DUT box is in contact, conducts heat) into the box. That gets the air molecules moving faster and Brownian motion and entropy spread the heat around. I doubt that there is a significant convection component in the little box. Convection would be a major factor in a larger box and a stirring fan would be essential as TIN pointed out.
Consider the outer aluminum box. It is exposed to ambient, however it is also in direct contact (by four 4mm steel cap screws) with the metal plate that is the heat source for the box. In a way it could act like a "guard" being warmer than ambient and thereby lowering the differential temperature with the interior.
Now I have a controller that is sensitive to < .1C and can set hysteresis to as low as .1C, so now it is time for some experiments. I should have something to report by tomorrow.
-
Hello,
This is my controlled thermal volume solution.
https://www.eevblog.com/forum/metrology/manufacture-an-oven-for-measurements-in-a-temperature-controlled-environment/msg2532159/#msg2532159 (https://www.eevblog.com/forum/metrology/manufacture-an-oven-for-measurements-in-a-temperature-controlled-environment/msg2532159/#msg2532159)
Positive temperature only, project made from an insulated beverage can. The heating element is a 75 W flat ceramic resistor on an old graphics card heatsink.
-
That is an impressive rig IRFP. What level of temperature and what sort of tests are you doing with it? What is the range of temperature variation when operating? You have a 75W heater; what’s your total operating power budget?
-
Hello,
It's a low-budget platform.
With PID management and a stabilization of about 2 hours, the variations recorded are about 0.1°C. Stable in time over several days with variations in ambient T°.
The system can reach 65°C max, the power is proportional to the setpoint and then becomes low during the regulation phase.
We are not in the budget of a 5305 TECsource control at 0.004°C...!
I measure voltage references, resistive voltage dividers and other OCXO.
The platform is simplistic but it does the job.
-
No need to buy new :)
-
As you get down to small fractional degrees questions about uniformity across the chamber volume become very real. Minor sources of heat and variations in conductivity are issues.
It does pay to ask what is good enough. Most equipment has had considerable effort to reduce temperature sensitivity. How many orders of magnitude below normal room temp variation is required to achieve the results you want.
-
As promised, attached is a plot of temperature in my little box overnight. Ambient decreased from 25C to 21C during the test period. The PSU for the TEC was set at 960mA CC and 3.2V CV power was < 3W. The temperature controller was set at 30.0C with 0.1C hysteresis.
Over the 16 hour test, temperature in the box interior averaged 29.5C with a variation of +/- 0.5C. Both average temperature and the variation decreased slightly with the trend of ambient.
Whether this test is significant depends on two factors: the intended purpose of the device and the amount invested in the equipment. My intention was to build a small temperature controlled environment to hold things like voltage references and resistors when verifying and calibrating meters. So far my "out of pocket" incremental costs have been under $20 for the temp controller used in this test and a (maybe) better one for a test currently in progress, and $15 for the TEC. The other components were leftover from dead computers and other projects.
The enemy I face is hysteresis, who I believe is the Greek God of procrastination, or being late or something. Whatever.
My question to the Delphian Oracles in this forum is "Will a +/- 0.5C temperature stability on a consistent base (like 30.0C) be good enough to improve the quality of whatever verification and calibration I do now?"
Glauber's Salt (old-school metrology temperature standard) and paraffin have been proposed as ways of using latency to manage temperature. Thermal mass as well.
It looks like there are two paths: use a latent energy eutectic to hold its inherent temperature or get a super-duper temp controller and a fast response heat source to counteract temperature changes. Hmmm . . .
-
Here's what might work for more precise RT control:
"The LDI-800 Laser Diode Driver may be used to power laser diodes requiring up to 1000 mA of drive. Current control or optical power monitor modes may be selected. An ad- justable current limit helps protect the diode from accidental overdrive. The LDI-800 contains a thermoelectric cooler driver which is compatible with many laser products."
Sitting on a shelf gathering dust.
-
eBay find:
ThorLabs TEC2000 TEC controller.
TEC current range -2 A…+2 A
Compliance voltage >6 V
Maximum output power 12 W
Resolution TEC current 1mA
Noise and ripple <1 mA
This should smooth out temperature variations.
-
There is also Aliexpress "TCB-NE-AH" called "TEC thermostat" with claimed +-0.01°C up to 18A@24V for ~120€. Serial control fom PC. They have multiple models. Anybody try that?
-
This is from the AliExpress vendor listing:
"Temperature measurement accuracy is another concept. It represents the error between the currently measured temperature and the absolute real temperature in the physical sense. For general systems, this value is not important because, for example, the currently measured temperature is 25 degrees, and the absolute real temperature is 24.5 degrees, so if the object is stable at the measured temperature of 25 degrees, the object is also stable at the real temperature of 24.5 degrees. This is enough for general systems. As for the current actual temperature, it does not matter (for laser applications, it is necessary to adjust to find an optimal temperature to obtain the best output). " [emphasis added]
As other contributors have remarked, the better control would come from a device for laser management. This device isn't for a laser, but is it good enough for stabilizing temperature for an electronic DUT? Based on TIN, Shodan posts and my recent eBay purchase, a laser-precision controller can be found for $150-300 vs $125-150 for a bare board from China in April. Hmmmm . . .
-
To me it sounds like they want to say: it is not a precision (as in high accuracy) thermometer.
I think the chinese option handles the most power? Also as for entry barrier, this should be more deterministic to get hold of than something from ebay? In USA, supply of old Laser TEC Controller seems good, but not so sure for the rest of the world.
Shodan solution certianly takes the cake for stability, and can be repeated. It is not automated, but is that needed in home lab?
But power is a lot lower and entry barrier is higher again (order pcb and parts, smd soldering).
If china solution works as advertised, and someone integrated it into automation - it could be repeated into other lab without much fuss, has high enough power for bigger chamber.
-
I'm really surprised - why voltnuts don't use such modern TEC ICs, it can solve a lot of problems :o
I think reasoning is: if you can have known working solution for 150$ (ebay used), why invent something for 100$ in parts?
-
Those modern switched mode TEC controller ICs are a 2 sided thing: they allow for efficient control of a TEC, but they can also introduce EMI problems. The analog control loop is also only good for small / fast setups.
Slower systems are better controlled digitally. This can also take into account the nonlinear action of the TEC and can include the temperature of the other side or feed forward correction.
A TEC also has the problem of introducing a relatively low thermal resistance to the outside.
For some systems the alternative is using better insulation and than only heating to a temperature like 50 C, like used in the LTZ1000. A slightly elevated temperature also avoids possible problems with condensation if the temperature would go too low. With the LTZ1000 I still don't like the resistive heater, as it is nonlinear and not very effective at low power.
-
Seems like the key to controlling one of these chambers with an analog process is to calculate where the DUT temperature will be using the current temperature in the chamber as one of several inputs to the algorithm. Other inputs would include the daily fluctuation in lab ambient, the latency of the chamber and the position in the power cycle of the TEC driver. First, the independent variables: Lab Ambient can be forecast from a model derived from historical data; Chamber Latency is another historical model-derived forecast; same with the Power Cycle Position. Those three models need to be independent statistically to reduce the cross-correlation effect.
The dependent variable is the temperature of the DUT (or perhaps the output of the DUT ??). The analog process "throttle" is the instantaneous setting of the power to the TEC and whatever "hand" is on the throttle must anticipate the time-dependent effect of the throttle setting.
Looks like an industrial control/autopilot/Mars-mission kind of challenge . . .
-
The analog control is usually limited to a simple PID or often even PI controller. So one sensor in and one heater out.
The difficulty with thermal systems is often that they tend to get slow when it gets large and long time constants above some 100 s get tricky analog. So one may have to compromise with the sensor position - closer to the heater gives fast response, though not exactly the DUT temperature.
Even digital control usually works essentially like PID. The point that is usually included better than in the analog part is anti wind-up. If the sensor or actuator is non-linear one can usually at least approximately correct for this. Even this relatively easy part is often ignored - though no real need for this.
If the system is well behaved the regulator does not need to be more complicated.
A first next step with a TEC may be measuring the heat sink temperature and than compensate for the effect as a proportional feed forward effect or include the temperature as part of the TEC model to control the heat flow. The result should be the same, just different way to look at it and maybe different ways to get the parameter.
Even just getting a PI regulator right is a nice exercise - there is quite a bit of theory behind it.
Pure try and error can be slow with a thermal system.
-
Three time intervals to keep in mind when designing/evaluating controllers..
1. Diurnal variations- This 24 hour cycle is likely to be dominant in many situations. Controller time constant around a tenth of this is likely to be a good choice.
2. HVAC cycle time - Dominant factor in some situations. Will vary from 30 minutes or so on up. In my house there are two zones, one controlled by an automatic heater and the other heated by a wood stove. The automatic heater provides heating cycles of around 1-1.5 hours. The wood stove (which is where my shop/lab is) is controlled by the person stoking the fire and by the burn characteristics of the wood. Tends to be 2-3 hours.
3. Thermal time constant of the DUT. For a small exposed circuit card this might be in the neighborhood of ten minutes. For more complex and packaged assemblies thermal time constants well over an hour are not uncommon. Making chamber controller time constants significantly faster than this will provide small benefit.
These three cycles inform sample rate and test duration when evaluating your system, and are the fundamental reason that Kleinstein says that empirical design of the control loop is so tedious. Each test cycle has to take a day or more.
-
As you get down to small fractional degrees questions about uniformity across the chamber volume become very real. Minor sources of heat and variations in conductivity are issues.
It does pay to ask what is good enough. Most equipment has had considerable effort to reduce temperature sensitivity. How many orders of magnitude below normal room temp variation is required to achieve the results you want.
Don't forget heatpiping effects - thermal conduction along the cables from the DUT to outside the chamber.
If, say, you want to test tens of devices simultaneously, each with independent Kelvin connections, heatpiping along the test leads can be significant. And that thermal path will have its own time constant.
-
My Russian article about chamber link (https://ampnuts.com/micron-tec/) now translated to English link (https://hackaday.io/project/177631-micron-tec).
WOW !!
Thanks for the translation--the photos and explanation are well worth a visit to the link.
-
As you get down to small fractional degrees questions about uniformity across the chamber volume become very real. Minor sources of heat and variations in conductivity are issues.
It does pay to ask what is good enough. Most equipment has had considerable effort to reduce temperature sensitivity. How many orders of magnitude below normal room temp variation is required to achieve the results you want.
You don't even need to go to fractional degrees, just try to make a regular 100 liter chamber (freezer) to be stable* without bucket of circulation fans (home brewing is science these days, I need a hipster beard next)... ...and with turbocharged air circulation you don't anymore have same heat flow out of DUT compared (probably) to actual use case (self heating).
* For described "fermentation science nuttery" required accuracy it is easy, but the process shows nicely how the air circulation and contact effects easily at 0.1 .. 0.3 °C level. (while controller might have 0.0001°C resolution and accuracy, but what about the ie. copper wires on sensors (both as heat resistance and seebeck-effect etc. etc.). What sensor is used, how it is placed etc. etc.) Also NTC and PTC thermistors are really poor for accuracy and stability. Use something else, Platinum sensors PT100/PT1000 are the best for resistive ones as far as I know (long term stability and linearity, latter doesn't take much effect when ∆T=±1°), but they are slow (physically BIG relatively to the system). K-type thermocouples (like all of them) are unstable as is most others. Maybe silicon junction is good stablity vs. heat capacitance vs. stability (but assuming not any smaller than PT100/PT1000 as of naked component without casing ... protective that is, not chip package ie. TO-92).
Heat nuttery is like INV(voltnuttery).
I would probably go with two phase measuring, if I would build something like this. One reference sensor with best possible drift vs. £€$ and then something quick for actual ∆T-control. If this is only a automated "pinch grip heater" then .. well arduino and lm35.
-
Thermistors have an undeservedly bad reputation: very high stability thermistors are available. The subject is covered in D. R. White, K. Hill, D. del Campo, and C. Garcia Izquierdo, Guide on Secondary Thermometry - Thermistor Thermometry, BIPM, Sèvres, Temperature Measurement Guide, Aug. 2014 (https://www.bipm.org/utils/common/pdf/ITS-90/Guide-SecTh-Thermistor-Thermometry.pdf)., which states:
Traditionally, thermistors have had a reputation for instability, but thermistors are now readily available for temperature ranges within -20 °C to 60 °C with stabilities of a few tenths of a millikelvin per year. For applications in this range, their stability, high sensitivity and simple instrumentation enable a short-term measurement accuracy approaching that of standard platinum resistance thermometers (SPRT), but at a much lower cost.
The most stable thermistors are glass encapsulated, but there are precision epoxy coated types available. For example, https://www.mouser.co.uk/ProductDetail/Semitec/103AT-11/?qs=wgO0AD0o1vvXePzqWokBCw%3D%3D (https://www.mouser.co.uk/ProductDetail/Semitec/103AT-11/?qs=wgO0AD0o1vvXePzqWokBCw%3D%3D). We have tested a batch of these and found their resistances all vary with temperature in the same way: a common set of Steinhart & Hart coefficients gave in-calibration errors of less than ±0.2˚C between any pair of devices over the range 0˚C to 85˚C.
The great advantage of thermistors is their high sensitivity (change of resistance per K) and the relative simplicity of interfacing them. They make a good choice of sensor for high-stability temperature control.
-
Pt100 are relatively tricky to interface, as the voltages are really small. So unless you really need it (e.g. for higher temperature like 200 C) or very high accuracy as with SPRT, I would avoid it. The small cheap thin film PT1000 are also not as stable as the classical glass encapsulated ones - so you can't have small form factor and highest stability at the same time.
A NTC or a simple diode type sensors is much more practical. They can use a 2 wire interface and have low self heating and not need to look at the sub µV range.
Comparing the resistance change is a bit biased - a more fair comparison would the voltage, as thermal EMF is a common problem.
Type K thermocouple is some 40 µV/K
PT100 at 1 mA gives some 350 µV/K
PT1000 at 0.3 mA some 1 mV/K
a diode some 2 mV/K
and a NTC maybe 10 mV/K , depending on the power level
With the NTC the relatively high resistance can make leakage current a slight problem already are more moderate conditions. For higher temperature they are not an option anyway, but leakage can become a problem with lower temperature and high humidity too.
-
This is the first test of my little box with the ThorLabs TEC2000 temperature controller (used $1,500; for me on eBay: $170) Thanks to shodan, TIN advice on controllers.
I have made no attempt to adjust the PID controls and the set temperature is in ohms as the sensor is a thermistor. I have an 890 sensor on order and the TEC2000 will report that in degrees. The whole rig was set up in an unheated space, photo attached. The thermistors are playing hide and seek, the Keithley 6510 center stage is capturing data from four thermistors, the RTD in the box and the voltage applied to the TEC. Voltage is also visible on the meter at the top of the photo.
The attached plot shows the box interior temperature (4-wire RTD) in red, the box exterior top in orange and the ambient temperature in blue. Superimposed upon the ambient is the temperature from the thermistor taped to the heatsink. If the heatsink is below ambient, the interior is being heated and vice versa. The orange plot illustrates the efficiency of the insulation, not so good, but better than nothing. Yes, I intend to fix that.
UPDATE: Added bubble wrap around exterior (trash-can is full and wife said "not with MY towel!")
-
Still very horrible and useless as thermostat. Putting DUT in a towel and inside metal trashcan will get better stability than this... :bullshit:
-
PS. My sensor chart:
1 - NTC - extremely high sensitive to small temperature changes, cheap, easy to drive, different dimensions are available. 8)
Also the linearity errors are high (compared to PT100/1000), because of the high non-linear sensitivity. However it seems I must refresh my knowledge about current NTCs. The annoying thing is that I do not have personal equipment to actually measure the stabilities nor do I have access to cal-lab. |O
2 - TMP117 - Guaranteed high precision, reasonable price, easy to use with flex-PCB(link (https://oshpark.com/shared_projects/rqIdFGTS)) and Raspberry Pi Zero W. :-DMM
This is new sensor to me. I just wonder if there might be switching noise introduced to DUT because of digital bus and contact measurement, for most it probably doesn't matter.
3 - PT100/1000 - Guaranteed sex with low resistance, high price for metrology grade samples, hard to use. :box:
Kelvin or bridge. One thing that someone somewhere here at forums stated, hard to find drift values. What I have understood the RTD drift is strongest (platinum contamination) at maximum operation temperature and somewhat limiting on that, ie. some thin film (eRTD Variohm) proposes R0 drift as 0.04% after 1000h at 500deg (which is +200 °C for A-type 300°C max operation). It would be actually nice to know how these would drift when used strictly at 0..100C range similar as NTC. It still lack the natural sensitivity (for me sensitivity of these elements have always been the response time, but anyway) to ∆Ω/∆T.
Good article about sensors, specially for "TEC beginners" dudes: Wavelength electronics AN-TC11, Thermistor Basics - link (https://www.teamwavelength.com/download/applicationtechnotes/an-tc11.pdf)
The article states stability only as relative process-loop stability, not actual absolute stability as of thermal reference.
-
Still very horrible and useless as thermostat. Putting DUT in a towel and inside metal trashcan will get better stability than this... :bullshit:
Hello Illya,
Are you sure about the can and towel? Or is that a suggestion for a comparison test? We have recycling here and metal trash cans are nowhere to be found, but I could maybe use a cook-pot . . .
Have a look at the attached plot that covers a two hour period last night from roughly 22:40 to 00:40. Ambient temperature was 19C and inside the box was a relatively toasty 28.4C +/- 0.4C. Did I give the impression that this project was completed? That I was satisfied with the results? I don't think so.
There are two very good reasons to participate in a forum like this: to learn something new from folks with experience and to share experience with others so that they might learn. Sometimes sharing one's efforts, mistakes and failures can be more instructive than showing off a finished product as if it were easily done. My mathematical vocabulary developed in a narrow segment of finance, metrology is like learning a new language so if I have given offense, tut mir leid.
-
Wait a second. So your chamber makes worse temperature stability than ambient room temperature without anything?
Maybe I don't understand what is the purpose of this project then? Wasn't it to make a box with TEC module to allow fixed stable temperature for A9/etc reference boards, while changing ambient in 19-29C range in your lab? :-//
Are you sure about the can and towel?
So based on your plot above, yes, simple can with a towel will be better stability than what is on your last graph.
Don't take it personal, it is just unclear to me what you trying to do here, as posters above already suggested "traditional" and already proven working solution for thermal stabilization. Get rid of metal box (use it for something else), mount TEC on finned heatsink inside the enclosed inner volume, add little fan for inner volume to mix air, implement active temperature-power control for TEC drive (with PID or even just simple relay on/off method) and you will get much better than +/-0.1C stability with ease and settable temperature in the box. There is nothing much to add here, concept with controlled heater to maintain oven constant temp is straight-forward and common task.
-
Question, Is it a mistake to try to set the desired "stable" temperature above the highest normal ambient? I assumed that way would be more in line with what the heater in the LTZ1000 was trying to do. If that is a bad idea, I can change settings, no problem.
The instability that TIN noted was due to untuned PID and also poor insulation, both are work-in-progress and should be working by the weekend. The sole purpose of the last test was to see whether the TEC was capable of holding temperature significantly above ambient under the worst of conditions. The cyclical instability was to be expected due to untuned PID; holding the box temperature 9C above ambient on average, was a positive result I thought.
Now, what about metal boxes? Starting from the DUT (probably an A9 board), it should be in metal box to reduce interference, right? Temperature sensor for the TEC controller inside the box with the DUT, or stuck on the metal plate, which? Metal box containing DUT in close thermal contact with the thick plate that mounts the TEC, right? Next, enclose DUT box and thick metal plate with insulation and leave the TEC and heatsink in open air with fan on heatsink, right? That all make sense.
This is where I have a problem. The boxes TIN and shodan describe are bigger than what I would like, and have open space inside and a fan to circulate heat. I would prefer a small unit that can be run 24/7 as a stable voltage reference whenever needed.
If the insulated DUT metal box is in thermal conductivity with the TEC, that box should be at a uniform temperature and the DUT inside heated by radiant energy from the box. The only job I see for a fan to do is move air around and that would take heat away from the DUT box in the configuration described above. If I missed something here, please set me straight.
-
The periodic temperatur indicate oscillation due to poor PID parameters. The waveform is however a little odd - normally one gets more like a near sine.
A temperature slightly above room temperature is OK. With a DUT that produces only little heat one can get this easier with just a heater instead of the TEC. Just a heater allows for better isolation. So a TEC may not be the best solution, especially for a low power DUT (e.g. ref. capacitor). The temperature above room temperature can be OK for many DUTs (especially those meant to be used inside an instrument, or those that need low humidity).
For a really small DUT (and preferrably low power), one has the option to work without a fan, just a relatively thick walled metal case. An extra alumunum plate may be used to thicken the wall where the TEC is mounted. The whole metal case would than be controlled and transfer the heat to the DUT. The A9 boad should be still acceptable small.
If not just single purpose, I would still plan for space for a small fan (e.g small CPU / GPU fan from 486 times). There should still be an extra temperature sensor at the DUT, not just the sensor at the case for regulation.
Those small units could run 24/7 also with a fan. Inside a closed box the fan would not make much noise. 24/7 operation would still need some safety features to avoid overheating (e.g. thermal fuse) and possible fire hazzard.
-
TiN will be stabilize after success tune >:D
::) ;D
I still do not understand how you and other voltnuts (I have been trying to avoid this sickness for many years) will keep these chambers at correct absolute temperature ("self-calibration" sort of) say 25C if there is no (at all) long term temperature reference.
-
The periodic temperatur indicate oscillation due to poor PID parameters. The waveform is however a little odd - normally one gets more like a near sine.
[...]
I suspect the reason for asymmetrical waveform is the different gain (as in energy transferred) of the TEC when heating/cooling.
-
The periodic temperatur indicate oscillation due to poor PID parameters. The waveform is however a little odd - normally one gets more like a near sine.
[...]
I suspect the reason for asymmetrical waveform is the different gain (as in energy transferred) of the TEC when heating/cooling.
Might be just bang-bang control mode and what we see is thermal mass integrating the function. Too early and not enough information to say anything.
-
NIST claims ice melting point can be easily realized with 0.01C uncertainty. One point is not ideal for tracking long term drift but still better than nothing.
-
Might be just bang-bang control mode and what we see is thermal mass integrating the function. Too early and not enough information to say anything.
Integrating? Probably, however the TEC controller is not supposed to be "bang-bang" as it is bipolar and designed to maintain the frequency of lasers by stabilizing temperature. One addition to my tracking protocol will be the amperage through the TEC. Voltage is currently tracked.
Any suggestions as to what should be expected from a ThorLabs TEC2000? ThorLabs is a high-end supplier of optical gear to universities and other research facilities. I would like to think they know their job.
-
The "ice melting point" used for temperature calibration is in the latency zone where water temperature is stable due to the latent heat. Ice by it self is colder and water with no ice is warmer. There should be a mixture at the bulb of the thermometer. Not just touching ice or in the melt water.
Is there a place in Metrology for temp-nuts? Probably.
-
Might be just bang-bang control mode and what we see is thermal mass integrating the function. Too early and not enough information to say anything.
Integrating? Probably, however the TEC controller is not supposed to be "bang-bang" as it is bipolar and designed to maintain the frequency of lasers by stabilizing temperature. One addition to my tracking protocol will be the amperage through the TEC. Voltage is currently tracked.
Any suggestions as to what should be expected from a ThorLabs TEC2000? ThorLabs is a high-end supplier of optical gear to universities and other research facilities. I would like to think they know their job.
...Reality is that it is just fast response thermostat (it seems) >:D and Bang-bang controller can be bipolar. There could have been a bang-bang mode for some odd use case, implementing that costs nothing if done with software which I do suppose is the case in here. The ripple looks like a non-symmetric saw-wave (on-off-controller if you wish) with thermal mass rounding it (acting as integrator). Well research can be done with really odd pieces of equipment (means poor or modest), I'm not stating that it would be the case here with ThorLabs, what I thought that the device is some "as is delivered / lastly used" mode and doesn't do it job as out of tune / wrong mode for the job in hand.
With quick skim TEC2000 is PID-only (as pointed). Only PID and there is a rather good (at least longish) tuning instruction there.
https://www.thorlabs.com/drawings/241fa595cd73d614-F2A048EE-AD74-E27F-6E71BEA26112560B/TEC2000-Manual.pdf (https://www.thorlabs.com/drawings/241fa595cd73d614-F2A048EE-AD74-E27F-6E71BEA26112560B/TEC2000-Manual.pdf)
But first thing you might try out is to make some step-response test (or pulse response by doing up and then down steps) to your chamber with increasing and decreasing set-point (desired set process temperature) as one discrete Heaviside step function (or as fast it goes.), magnitude of 0,5..1,5 degrees maybe in this case. With making a known disturbance to the system you actually can see something about quality of control and behavior of the system.
I have politically incorrect place in mind where all temp-nuts might belong. >:D
...But I'm not here to argue, these are interesting projects and subjects and I will follow them (..and learn probably some new things or viewpoints along the way).
PS. shodan@micron, I totally didn't notice two things A) it is for medical B) the NIST traceability, I need to order one or two of these... :D
Also it is full traceability sensors: "The TMP117 units are 100% tested on a production setup that is NIST traceable and verified with equipment that is calibrated to ISO/IEC 17025 accredited standards"
-
Still very horrible and useless as thermostat. Putting DUT in a towel and inside metal trashcan will get better stability than this... :bullshit:
Hello Illya,
Are you sure about the can and towel? Or is that a suggestion for a comparison test? We have recycling here and metal trash cans are nowhere to be found, but I could maybe use a cook-pot . . .
Have a look at the attached plot that covers a two hour period last night from roughly 22:40 to 00:40. Ambient temperature was 19C and inside the box was a relatively toasty 28.4C +/- 0.4C. Did I give the impression that this project was completed? That I was satisfied with the results? I don't think so.
There are two very good reasons to participate in a forum like this: to learn something new from folks with experience and to share experience with others so that they might learn. Sometimes sharing one's efforts, mistakes and failures can be more instructive than showing off a finished product as if it were easily done. My mathematical vocabulary developed in a narrow segment of finance, metrology is like learning a new language so if I have given offense, tut mir leid.
How familiar you are with control tuning and control theory (in this case Proportional-Integral-Derivative aka PID)?
Please do set-point step with that setup something like 2 degrees, it is oscillating if you are using the TEC2000 for control. Read the manual from above post, you can try to tune the system from the instructions given there. If I would need to give an educated guess there is too much integral time set to controller, maybe also derivative which tries to stabilize the thing ending up steady oscillation. One way to solve this is if you educate yourself with some articles of Ziegler–Nichols method PID tuning method (which is some times duped as oscillation method, but I can not remember anymore what books did say) or step-response PID tuning method, for fast and harmless system oscillation method might be usable approach.
If I would need to tune this box with specialist methods aka intuition (or expert method aka autotuning button ;D) I would go with step method with both integral and derivative terms turned off so the controller is only proportional. Setting the set-point nominal value to ambient temperature and doing steps at both directions. First by just adjusting the P-term so that I would find fast response to set-point step. As there I would enable so slightly the integral term of the TEC2000 so that the offset-error would go away. D-term is a bit two fold and I would leave it out from the first test round. After I would have noted that the system is following the set-point step with damping oscillation, without offset error and rather rapidly I would run first process test (aka leave it logging for a few hours and see how it react to ambient).
-
Thank, I'm certain there will be many mistakes as "learning opportunities" for all. I'll try to report what I observe without bias as to the (assumed) provenance of the hardware.
1) I don't know anything about PID setup. Other than an excellent trading video from ABB on YouTube last week, my history with industrial control was "send a guy out with a wrench" trial and error, followed by "if it's working, don't f--k with it."
2) I have the manual for the TEC 2000 from ThorLabs and have read it. The link in the post above gets a 404 error, but it is probably the same manual. No worries.
3) "Bang-bang" is what I see from the data I'm collecting. A sawtooth with 240sec period with about 10sec turnaround at top and bottom. Not a sine curve at all.
4) I have decreased the set temp from 33c to 30C and controlled the ambient to 25C +/- 0.5C
5) The manual setting instructions start with PID all fully CCW--it looks like 'no effect'--so i set 'P' to middle (better), so then full CW and now nice little sine wave, 80sec period.
6) Now the measured temperature at the DUT is 30.4C +/- 0.05C
Now what? The manual suggest tweaking 'D' next . . . .
7) Setting 'D' to max CW made small improvement.
8) Moving 'I' off full CCW made things worse.
Amps are set to max for the TEC installed. Any reason to adjust to lower value?
-
Small correction, which probably is because my average english. Bang-bang controllers are nickname for on-off-controllers, iirc one description is hysteresis controller. You are (at least by manual) using PID-controller. Why I did refer to bang-bang controller at beginning is that I falsely (which quick reading of manual titles etc.) thought that it does have such included, now it seems that I was mistaken.
I think the best thing for you is play with the controller a few days and try to google some hands-on PID tuning instructions. No need to go full academic with mathematics, since you do not have a model of your system anyway (and I think the TEC2000 is also black-box, so actual algorithms on controller is not known) and probably it will be changed many times more.
The D-parameter try to resist the change what I&P terms do, it doesn't hurt you to tweak it, but I would imagine that it is still not fully optimal at all since you are saying there is steady (but small :-+ oscillation going on).
The a graphed setpoint step is good way to see how system behaves.
You have a good setup to familiarize yourself for tuning of PID.
Like I said try to find some good (authorative) hands-on articles about PIDs. While this is TEC-controller it is still just a PID controller in the end. After also reading a bit about the control subject return to tec2000 manual, it most probably open differently what everything do mean on it.
You have a good learning setup there, also play a bit the box and sensor itself to see how it might effect to controller outcome. That is with different measurement point and added thermal mass (ie. Cold soda can to the box etc.).
Not so quickly from phone.
Ps. Graphs should have (if possible) setpoint, controller output and desired target (box or dut temperature at this case) with those three you can see how the controller behaves. For long term measurements add athmosperic temp as that is source of disturbamce for you (by not knowing exactly all the whys and wheres).
-
And these advices are given for you for reason you get understanding (of PID tuning) only then you can start to get where you actually wan't to go. Because if you do not have basics sorted out you will not get where you wan't to go as metrology thermal box.
-
A TEC is a nonlinear device, especially near its rated current. Usually the rated current in the DS is for a perfect heat sink. Real world the maximum useful current is lower (e.g. some 10-30%) and with more current it will not cool more but less. It absolutely makes sense to limit the current to a lower level, more like 50% of the rated current, so the TEC is in the relative linear range and no danoger of overheating. I would use more only if absolutely needed to reach the final temperature.
There are many different methods to tune a PID system.
It absolutely makes sense to record the step resonse for the system, that is apply a defined step in the TEC current (e.g. off to some 20% of rated current - maybe more if the response is weak, or less if the temperature gets too high/low). This could show the characteristic of the system an tell how easy it is to regulate or if one may need a different postion of the sensor. A system with a long delay gets hard to control, especially with an anolog control loops with limited paramters.
For the step response on can calculate approximate suitable PID setttings via several approximate rules made for slightly different systems.
2 such systems are Ziegler, Nichols and Chien, Hrones, Reswick.
https://infosys.beckhoff.com/english.php?content=../content/1033/tf4100_tc3_controller_toolbox/9007199500260107.html&id=
This should be at least a suitebale starting point and OK to check if the parameters are feasable with the given controler.
This needs a little more math, but it can be worth it. The try and erros method is really time consuming with a slow system. A 3D parameter space is quite large when every try takes hours.
-
Kleinstein did write it up clearly. I want to give a warning those tables can be extremely confusing, since the controller (I don't know if it is the case with this one OP have) might not have the same input parameter levels what those gives out as values and the markings differentiates highly (as seen on those two tables). Also there is differences on those tables at which way the parameters do go, I have forgotten the 'why', but relates to the mathematical model.
I'm actually surprised that this is TEC-unit is still analog :o , but it seems it is rather old (does its job). Makes sense in that regard. Uh, I feel relieved I do not need to use analog day to day, there is enough headache from million different digital implementations to remember, but auto-tune algorithms and process tuning programs are from heaven.
4) I have decreased the set temp from 33c to 30C and controlled the ambient to 25C +/- 0.5C
6) Now the measured temperature at the DUT is 30.4C +/- 0.05C
What I was after was to adjust the set temp "zero level" to be ambient temperature (room temperature) and then make these steps below and above from that zero level. By doing so you can see both cooling (down) and heating (up) and how the response of controller do look like for each direction (when graphed, if you have real time graphing possibilities that makes this so much easier).
5) The manual setting instructions start with PID all fully CCW--it looks like 'no effect'--so i set 'P' to middle (better), so then full CW and now nice little sine wave, 80sec period.
The tuning instructions usually (you can read always) are written from state that controller is turned to as close as possible to P-only controller, removing I and D -terms effect. (this leaves usually permanent offset (error value) between set-point to response, mathematically always, but at practice especially with analog devices not always). Also notice how the manual says to open the device and removing jumper if you need to disable I-term (removing capacitance from operator amplifier circuit most probably).
Kleinsteins method is way to go if you can wrap your head around those tables, I were hoping you would end up to some articles explaining the tuning and these tables.
There is caveats you will find out, ie. when you modify your sensor position the system will change and the tuning might not be as good anymore (or improve).
-
Those tables for PID tuning are partially heuristic and partially based on a model for the system. Slighlty different models give slightly different values for the tables. The other point is that there are different targets on what to consider a good tuning. This can be just without overshoot or with an "acceptable" level to get fast relatively close.
There is an awful lot of tables and tuning methods - though the principle is the same: do a test (e.g. step response, pulse response, limit of oscillation or a bang bang type crude regulator) to learn about the sytem and than calculate suitable PID parameters based on approximate formulas. It depends on the details of the system which approximate model fits best - so there is no univesal best method. The resulting parameters are still rather similar most of the time.
Within the range given by different tables is about where one can see with try and error and look at the specific target which one fits best. One may still have to look at the parameters as sets - mixing the gain from one table with the times from another one may not work. Often there is also a good enough - so the first result from the tables may work good enough - especially if the system is easy to control. Thermal systems are often slow, but still usually relatively well behaved. If really bad, it may be more about sensor placement than fine tunig the PID parameters.
Many of the so called auto-tune regulators one can buy do a similar test and calculation method inside the box. This allready is a big plus for the users. I find especially the method starting with a crude band bang regulator interesting though this is more a method to integrate in a regulator, not so much to do by hand.
There are true auto-tune algorithms, that do the system test while doing the regulation, but this is not without problems and mathematical complicated. Controling the speed of learning may be as tricky as setting the PID paramerters :scared:.
The PID parameters can be confusing as there are different formulas used to expess the regulator. So one can have the gain or the invers and similar one can expess the I and D parts with times for the cross over or as factors including the gain. The units give a hint, at least for the I and D part.
With analog regulators one may not have a meaningful scale :(.
-
Well, thanks to all the advice and references, things are looking better. The interior of the DUT box has been holding 30.4C +/-0.01C and I've managed to cpntro; ambient to 25C +/- 0.5C using a "small-room" air conditioner. Adjusting 'P' was made more difficult due to the 'P' control being 'high' at the CCW setting, not 'low' as the instructions seemed to indicate. Right now all PID are max CW and temperature is stable in the box. It looks like I'm have reached the "don't f--k with it " stage for now.
Screenshots attached.
-
For testing the control loop, one should intentionally not have a really stable room temperature - more like the opposite of intentionally increasing the external temperature varaition to see how much lower the variations are at the inside. A stable temperature is more like something for later to help the meters and in case the regulation does not work well.
The temperature still seems to show some relatively fast oscialltion. Because of the low level I am not sure if this real, or just something like a beat frequency to mains.
So the parameters may be still far off.
Another way to test is to look at steps, like the turn on transients, or a short disturbance, like opensing the case for a short time while running.
-
The fixture being measured is designed to hold a small calibration reference, most likely a 10VDC source for monitoring autocal on DMM's. The plan is to leave it powered up and connected, probably to the 3458 for some long time testing.
The same effect as changing ambient can be had by making a change in the set temperature. That was just done and the attached photo shows the measured temperature on the plate side of the TEC rising to the new setting without overshoot and settling in with the same relative swings in temperature as before the change. The periodicity is about 65 seconds, which seems more consistent with latency in the fixture than a beat frequency. There is some new 'space-age' insulation to be installed and that should reduce the fluctuations.
-
Unless something very strange is going on I would not expect your new insulation to reduce the magnitude of these oscillations, though it could possibly lengthen the period. Your graph seems to show either an oscillation or a dead band. It would be very telling to show TEC current on the same trace as the temperature.
The real question is does this behavior meet your needs? For what I would be doing with voltage references it would be more than good enough.
-
There are two temperature sensors in the system: one is connected to the TRC controller and reads the temperature of the 'cold plate' in the chamber; the other is close to the DUT (a TI REF102-based 10VDC reference) and reads the temperature affecting its voltage output. Both are thermistors of the 10kOhm type. The TEC controller reports 'actual' temperature as resistance, leaving the calculation of temperature TBD. There is another source of actual temperature--this time reported on a 0 to 10VDC scale--presumably for a more precise laser diode controller.
The short-term fluctuations shown in the plots above are of the 0 to 10VDC signal representing the cold plate temperature. To be honest, given that the thermistor temperature (resistance) is inverted, I am not sure whether the voltage signal is as well. In any case, my focus is on its magnitude and periodicity compared to the DUT temperature. In rough numbers, the cold plate fluctuates about one percent of the full scale on a cycle of sixty seconds and the DUT temperature is stable at 25.5C +/-0.05C.
The TEC controller may not yet have enough data to integrate the fluctuations. As reported above, the current is 800mA and there is a fairly rapid shift of polarity at the inflection points.
The real test of "good enough" will be the stability of the DUT over a lengthy testing period. We'll see . . .
-
If the TEC power is really changing polarity and reaching some 800 mA this sounds quite a bit like oscillation of the regulator loop. So it would be a good idea to look at the loop tuning again. Even if not used for the tunig, the setp response of the sensor to actuator step (not a step in the set point) could tell if the system is easy to control or more difficult and may thus need a different sensor position.
The suppression of external temperature variation (some 15 min cycle) is also not that good. Some 0.1 K ambient and 0.015 K inside. Not bad, but also not especially good. With so little change the suppression factor is a little hard to judge.
-
As reported above, the current is 800mA and there is a fairly rapid shift of polarity at the inflection points.
A rapid change of polarity only generates heat.
I would avoid that.
And always think of: the TEC acts also as generator when there is a temperature difference on both sides.
So if you change polarity quickly the voltage generated on the TEC adds to the supply voltage.
This increases the intended current. (maybe above the system limits).
With best regards
Andreas
-
Here's the latest:
First, a link to a YouTube video on PID control https://www.youtube.com/watch?v=6OH-wOsVVjg (https://www.youtube.com/watch?v=6OH-wOsVVjg)
It is by a drone pilot--with a lot more at risk than a voltage reference .5 PPM variation--and by his own admission, an over-simplification. A graphic, math-free explanation of P, I and D contribution to process control.
Second, my power to the TEC at 800mA was the cause of oscillations. I reduced it by halves down to 50mA, checking the effect at each stage, and no more oscillations. Considering that this particular DUT is drawing just over 1mA from the PSU, what is left to improve?
Finally, I got over "making do" with a 4.0C swing in ambient temperature in my lab and found a small "room air conditioner" on Amazon that now keeps the ambient at +/-0.4C. It sends the hot air into the attic via a 5" flex hose.
-
Is that an Arroyo 5305 (12V 5A)?
Ambient temperature is 29°C, is this the water temperature or room air temperature?
The datasheet of the quad module has limits of 14.8 V and -50 °C. Did you use both Krytherm modules on top of each other or or only the quad module? Will the "tower" crack when going lower?
Regards Dieter
-
The 4 stage version is for an even smaller system, more like a IR photodiode.
If a slightly lower temperature is needed, there may still be the option to relace the external radiator with a bucket of ice water. It is low tech, but OK for short time use.
-
The temperature control looks really good. Getting a nice stable temperature is easy, a nice fixed slope is a little more tricky, but the sharp corners are the art.
The orange foam looks a bit messy. For the low temperature range is can be OK, but in the higher temperature range the ready made isolation foam is usually better.
-
Very nice work.
A controlled slope is important for some types of tests. But square corners are kind of like ornate artwork on baroque buildings. Pleasing to the designer, and satisfies a currently popular style but really has little or no functional importance. A minor savings in test time is about all you can claim for it.
-
Hi Shodan. I follow with interest since I'm doing a similar project.
You may see, temperature of Arroyo feedback have large shift between DUT temperature, when Tset away from ambient.
Reason is simple: heat leakage trough DUT measurement wires ;)
Could you make the DUT temperature wires loop around inside the box before exiting? In that way, the wires can cool down better, with less ambient temperature conducted to the sensor tip.
-
Great that the loop made such difference!
I don't have time to continue my own project right now, but I will post details when I do.
-
try loctite tite foam for making this, its slightly less insulating but the materials parameters in terms of toughness is vastly superior to normal crappy expansion foam.
it is around 7$ a can vs the normal 4$ but it is worth it
-
Shodan, not quite sure what the plan is, but if the cost of milling holds you back I can share what I did. I bought blocks and cylinders of aluminium and managed to saw and drill them at the local makerspace. Copper should work too, except I heard that drilling deep holes in copper is hard.
-
I do machining on a small lathe for fun in free time. When I do design most often I tend to over engineer designs.
Manual lathes are often available in each city and hourly rate is much lower.
Maybe it is worth redesigning coper part in order to be doable on lathe?
If I need to do something like this I would redesign coper part to only have circular features and few bores that doesn't need to be precisely aligned.
Will increase size of coper part and will replace aluminum part with plastic part that can be produced on 3d printer.
If aluminum thermal bridge is needed, then will make it as cylinder and plastic part around it.
It will not look so fancy but it will have comparable performances.
-
Article for Russian EEVBlog readers: https://ampnuts.com/deep-cold/
(sorry, my English too poor to create big text...)
No worries, google translate is doing a good job. Always interesting to read your post Shodan :-+
-
Some people even use web translators to read the russian description linked above.
When doing that, i thought about your little theory about reducing air convection losses by cooling from above. Think about it again: The coolest spot is your cooler, so it should be at the bottom. I know the cooler has a hot side but that's another problem to solve by air tight and thick foam isolation. The heat "buffer" you insert between the two TEC stages will help with that.
Regards, Dieter
-
I like those nylon screws. A pitty we don't have nylon wires.. But seriously, what is the heat conduction of a TEC in comparison to some wires or screws?
Regards, Dieter
-
Ok, report: The China control module arrived & I had a play. It's the "TCB-NE PT100" which is specified at 12V/12A up to 24V/10A. The unit operates as a buck converter with switchable output polarity with +-1000 steps resolution. It can reduce maximum power going to the peltier, at the expense of control resolution by setting a lower maximum allowed duty. The duty is calculated by PID from temperature difference "Set Value"-"Read Value" at an interval between 100ms to 3s. To have stable regulation, the interval has to fit the speed of your thermal feedback. Each time a new duty value is calculated, it reports over serial. Also you can have manual interrogation instead. So far I did run it at 12V/4A max and it stayed totally cool so I have no reason to doubt the numbers. PCB seems well made, "Sam Young" capacitors. The unit has two error & ready logic outputs and a disable logic input. The Aliexpress seller send only the chinese manual, but shafan-oe.com seem to be the primary source and I got a prompt response in good english and the english manual from them. The official seller is shafan-store on taobao. PID values can be modified on the fly and will be used on next cycle, so with a bit of scripting on the pc side more complicated stuff (different P for heating and cooling as a first step?) can be wrapped around. All in all it is a simple unit without much bells and whistles, seems to be originally made for industrial temperature stabilisation of optical elements. Verdict: It does deliver for its price, at ~1/5 compared to more professional units I am happy with it.
Edit: Also, if you have hints on how to optimally set interval with regard to thermal feedback delay, I'm all ears ;)
-
The update interval sets the upper frequency limit, as the 4th (often hidden) paramter to a real word PID regulator. As a rule of thumb one wants upper limit frequency about 5-10 times the P to D cross over. So the PID sampling / recalculation should be at something like 1/5 the thermal delay.
Faster calculation would increase noise of the D term - some noise may be actually wanted to get some kind of dithering to avoid the 1000 steps quantization.
Normally the same PID paramters should work for heating and cooling. This is at least true if the PID loop include linearization of the TEC. The TEC is nonlinear with a power level about a parabola with the maximum cooling near the nominal current. So if no linearization is used, one may consider to adjust the overall gain depending on the power level, e.g. less gain for heating and maybe more close to the lowest temperature.
-
I would not torture the low temperature setup with so high temperatures. Some +100 C should be plenty. Of cause for some commercial test a -40 to +125 C range in a single setup is nice. Some of the stress comes from the temperature span.
The action of the PID is nonlinear, giving more gain when heating. For a very large range one may have to take that into account and reduce the PID gain for higher heating power, or directly add linearization for the ouput part if one has it's own PID controller with full control.
Instead of 1 higher power water pump one could just use 2 (and a 2nd heat sink).
-
Did you find a RS-232 programming manual for the AE-800?
Regards, Dieter
Edit: See here: https://www.kepcopower.com/support/cotek_communication_protocol_spec_a5_0928.pdf (https://www.kepcopower.com/support/cotek_communication_protocol_spec_a5_0928.pdf)
-
Those mosfets are specced with Rdson = 1.4 milliOhm typ and 1.8 milliOhm max, a pretty solid build. When looking at your image and thinking about heat, i found the solution for low output operation of my Cotek AE-800. As of the AE-800 spec sheet operation below 10 % of nominal 30 V output isn't recommended, but i can use the polarity switch to implement PWM between -3 V and + 3 V. With a little bit more firmware one can keep "linear" operation of the AE-800 outside of +/- 3 V interval. Have to try yet.
Thanks again for sharing the info!
Regards, Dieter
-
A TEC is quite nonlinear, and even the parabola approximation is not perfect (this ignores the resistance going up with temperature). So I would not worry so much about the linearity of the output stage.