DIY Large thermal chamber for metrology T&M testing

[dietert1]

Left y axis with blue labels is for blue plots of R6581T, right y axis with red labels is for red plots with K2182A measurements. R6581T indicates about 12 °C more.
Maybe the R6581T is more unstable by its fan and the air intake is in a corner (lower left corner of image). Corners are more exposed to ambient temperature change. Also the R6581T runs from standard mains until now, while the K2182A has a special synthetic mains supply with low noise and +/- 0.05 % amplitude stability.

Regards, Dieter

[TiN]

MiniTEC air bath is now assembled together and ready for first tests.

Since it would be surrounded by measuring devices, I've decided to cut it for less depth. It is still big enough to fit two ESI SR104s or two typical half-rack DMM such as Keithley 2002. Also smaller volume is faster to cool/heat.

Here it is in expected habitat:



ManateeMafia wanted to see more details of actual TEC assembly, so his wish granted here.
Since we have just 40W nonamium TEC module, single 120mm CPU AIO watercooler used. It is cooled with 12V 0.6A fan and uses same voltage to run internal pump on heatexchanger. There is no filling required, such systems come prefilled.



Peltier is sandwitched between former Xeon 604 server copper heatsink and waterpump/waterblock. Four steel bolts provide tension between parts. Everything is mounted on 25mm thick foam panel.



Inner (facing chamber volume) side expose only copper heatsink. Thingy on blue coax with a wire is Pt100 Honeywell HEL-705 RTD. Volt-nuts must appreciate use of bent Fluke 5700A binding post. No scrap production



To mix air within chamber volume angry top quality Sanyo San Ace 80 is used, 12V rated 1.1A. I usually run this fan at 6-7V to reduce noise, and it still gives plenty airflow. Fan size is 80 x 80 x 38mm.



Other side with liquid tubing and pump head.



To seal any possible gaps and prevent air leakage to ambient all corners around copper heatsink (cold side) are sealed with strong adhesive aluminum tape.



Hot side gaps and cavities around waterblock are further insulated with small-expanding foam. It also provides little bit stronger support to assembly in place, so it stays happy and fixed.



All wiring escaped before foaming up.



On top I've added some armaflex HVAC insulation to further seal up the assembly.



Looks from inside :



Some MLI waiting to be applied on outside as well after I deal with wiring.



After a quick test run, I've ran TEC cooler with trusty Keithley 2510 a bit and it went down to +4.7ish C with 25W cooling. Then after 1 hour it went to +50.0 and settled there with +11W heating. This was without good seal on front door, so I hope can imporove numbers a bit after more tuning.

Now we can escape some permanently fixed wiring into chamber and seal up the back.



Cables that I captured into this chamber are : 1 mains cord, 1 GPIB cable, three PTFE 4-wire cables, twinax cable with triaxial connectors and BNC cable. This should be plenty for most of DUTs, and I still can run custom cables from the front when needed.



Some Fluke 335 heavy transformers and ductape put pressure for better bond



Test run snap-shots:





CH1 powered mixing fan at 8V, CH2 powered pump and radiator fan. I'll get small dedicated power supply for this power later, no need for lab supply here.

[ramon]

First of all, great work and thank you for sharing. But I have one question.

You are able to source multiple 3458 and 5720, a 6010B bridge, ESI SR104, Lakeshore cryogenic TEC controllers, vishay custom resistor networks ... but when it comes to environmental testing you take the DIY path.

Have you not found yet a nice and small temperature & humidity chamber (less than 150Kg, not too power hungry, that doesn't require 3-phase power, ...)?

[TiN]

Good question  . With volumes I built last two chambers it would be still huge and very expensive high powered heavy chamber, which I cannot move or transport when needed. Also all typical industrial/lab chambers are using compressor coolers and very lucky if have 0.1C stabililty. 0.1C fluctuation would make any 8-digit resistance measurement in most cases just useless noise. Sure, there are some metrology-targeted chambers but they are even more expensive and often have small volume. I've never seen MI or Guildline airbath on ebay for example.

I mean, there is a reason why I still have not used my Fluke 7015 bath for anything since I got it 2 years ago... Just cost of oil alone is over $5k, if we even forget for a minute of constant 1000W+ power draw.

So spending few days and $200 in parts to build DIY chamber for limited temperature range from say +5C to +50C makes much more sense for hobby uses. I'm not doing this for any business or monetary purposes and have no plans to do so, still science only 

[SeanB]

You could probably get away contacting one of the larger oil distributors and seeing what they have in transformer oil, which is generally a pure mineral oil with no additives in it, so it is basically inert, or getting medical grade mineral oil, which is the same thing, but with better filtering in production. Otherwise refrigerant compressor oil, plain mineral oil, for systems with R22 or ammonia, which is also a straight mineral oil fraction. By me the refrigerant oil is very cheap, not much more than regular engine oil, and sold in 5l containers. The transformer oil is generally in 20l only, and the medical grade in 1l bottles.

[branadic]

I came across this chamber, when I visited wekomm in Dec. 2019, that's what they use for their resistor characterization, beside an expensive oil bath.

https://www.cik-solutions.com/kalibratoren/temperatur-kalibrierkammer/

As you can see, the dimension are special and not beneficial for most of the equipment we all use. On the other hand they are based on the very same principles, TEC elements, TEC controllers, fan to move the air inside the chamber.
I fully agree with TiN, commercial temperature chambers like the ones made by e.g. CMS and the like are out of scope for almost every hobbyist, even used and suffer from large noise due to compressors and such.

-branadic-

[dietert1]

We built our own chamber to be flexible. In the beginning the requirements haven't been completely clear. It was clear that we can reuse electrical/electronics components and the controller with a different size. Chamber itself was very cheap, maybe € 200 for parts.
Look at the outer size of those Kambic chambers and their weight: Model TK-190 which may be big enough to put 19" equipment inside, weighs about 150 kg and has WxHxD = 875 x 1300 x 700 mm outer size. Of course the outer size has some reason. Anyway our chamber has about 40 kg and a wall thickness of 4 cm. If that won't be enough in summer we can still cover it with isolation on the outside.

Regards, Dieter

[ramon]

Thank you all for your feedback. Yes, the size and weight is not hobby friendly, but its not easy to find a DIY solution for humidity testing. And I have asked because he is Tin. You know, he is not the average hobby guy. Who in his sane mind is going to buy a 141 Kg Fluke 7015? ;-)

Nobody cares about humidity, or should I assume that we should give up in this subject for DIY?

I was tempted to test the rosahl type devices but after closely reading the datasheets found they are meant for 'static' closed air environments. There is also a recent humidistat solution (Veldscholte, 2021) made with gas bottles and electronic valves but meant just for really really small chambers.

[TiN]

Humidity is in general lesser problem for most high-end standards, unless you experience high extremes of it (very dry or very wet).
I usually have it around 40-60% in summer time and 15-25% in winter here. Also one thing is to control, but measurement is relatively easy, so you still can monitor how DUT behaves with change of humidity even without humidity control.

P.s. 7015 was less than $500 shipped, so insanity was justified .

[dietert1]

On the image with the R6581T and K2182A i posted above you can see some desiccant that has been in the chamber for several months now. Although the chamber is air proof while closed the desiccant turned green after about three months. Desiccant is a very simple means of humidity control. Not really control, but there will be a stimulus one can track and evaluate. Or leave green desiccant inside to provide for more stable humidity far from extremes. The Sensirion SHT35 sensor used in our chamber measures relative humidity.

Regards, Dieter

[miro123]

Nobody cares about humidity, or should I assume that we should give up in this subject for DIY?

 1. equipment used to control humidity is called climate chamber, the topic is about humidity
 2. Humidity is relative parameter - You dont need to take care if your sensitive equipment is inside of oven around 40..45C.  RH at this temperature are already under 15% - even in your region.
  https://www.lenntech.com/calculators/humidity/relative-humidity.htm

 3. fully agree with Tin - at work we have many pro grade climate and temperature  chambers - Lets ignore the  budget and size, still nor of them satisfy my metrology needs. I'm building my own TC with short term stability in mK range

[Anders Petersson]

I was tempted to test the rosahl type devices but after closely reading the datasheets found they are meant for 'static' closed air environments. There is also a recent humidistat solution (Veldscholte, 2021) made with gas bottles and electronic valves but meant just for really really small chambers.

I read the Veldscholte paper and I don't see them specifying a chamber size, only mentioning that the chamber should be air tight. What is the size limit, in your opinion?

Nobody cares about humidity, or should I assume that we should give up in this subject for DIY?

I care! I've started looking into water baths saturated with salts.

[ramon]

You are maybe right, nothing prevents to use that system on large volume chambers. I actually have not made any numbers, the paper says a flowrate of 1L/min for a 22.5mL chamber (3x3x2.5cm). Here I can rent small 7L dry Nitrogen cylinders at US $64, and smaller ones 3.5L at US $40.

[Anders Petersson]

I think you refer to the 25 mL pre-chamber that was used in the example... it's not the full chamber:

"If for a particular experimental setup it is undesirable or impossible to fit the humidity sensor in the measurement chamber, it also possible to construct a ‘pre-chamber’ containing just the humidity sensor, which is then connected between the humidistat and the measurement chamber."

[mzzj]

I care! I've started looking into water baths saturated with salts.

Saturated salts are reasonably easy starting point for relative humidity calibrations but as usual there is couple of gotchas.

I have built two DIY "primary" humidity generators, "constant temperature-two pressure" and "constant pressure-two temperature" systems.
Both are somewhat involving but the pressure based one might scale better for hobby DIY.

[miro123]

I'm struggling to find insulation material for mid=size oven. Requirements are

1. Not flammable - I don't want to burn my house
2. User friendly - glass wool is great material, but tricky to assemble
3. Reasonable price - aerogel is great material but far beyond my budget

[Anders Petersson]

I care! I've started looking into water baths saturated with salts.

Saturated salts are reasonably easy starting point for relative humidity calibrations but as usual there is couple of gotchas.

Please share the gotchas!

[TiN]

I'm struggling to find insulation material for mid=size oven. Requirements are

1. Not flammable - I don't want to burn my house
2. User friendly - glass wool is great material, but tricky to assemble
3. Reasonable price - aerogel is great material but far beyond my budget

Well solution is easy, weld double-wall stainless steel thin-wall chamber and evacuate it to vacuum. Essentially a dewar chamber.

[ramon]

is rockwool easier to assemble, or just the same as glass wool?

[Kleinstein]

I'm struggling to find insulation material for mid=size oven. Requirements are

1. Not flammable - I don't want to burn my house
2. User friendly - glass wool is great material, but tricky to assemble
3. Reasonable price - aerogel is great material but far beyond my budget

Well solution is easy, weld double-wall stainless steel thin-wall chamber and evacuate it to vacuum. Essentially a dewar chamber.

Thin walled may not work well with evacuation. However with a fully closed doule walled metal construction it does not really matter if the insulation inside is flammable or fluffy glass-wool.

Rock-wool is a little more sturdy and may be easier to handle than the usually softer stuff. How soft or stiff is gradual thing. The stiffer grades may be OK to glue and keep there shape and also hold the weight of the inner part or at least support the sides.

[CDN_Torsten]

"Rockwool" tends to be easier to work with and also doesn't itch as much as glass-based insulation.
It cuts like bread with a bread knife (or sharp utility knife) and is less prone to insulative-loss due to over compression.

[MiDi]

I'm struggling to find insulation material for mid=size oven. Requirements are

1. Not flammable - I don't want to burn my house
2. User friendly - glass wool is great material, but tricky to assemble
3. Reasonable price - aerogel is great material but far beyond my budget

Construction material in EU (e.g. certified insulation material EPS) needs to be fire retardant.
If you are unsure just test the suspicious material with a blowtorch outside.

[dietert1]

I like the idea of flooding the chamber with dry nitrogen as it will prevent fire, too. Maybe with a glas of water inside, the nitrogen flow rate can control humidity.

Regards, Dieter

[MegaVolt]

You can borrow some ideas from him.  Came across a video by accident.

s

[DavidKo]

I'm struggling to find insulation material for mid=size oven. Requirements are

1. Not flammable - I don't want to burn my house
2. User friendly - glass wool is great material, but tricky to assemble
3. Reasonable price - aerogel is great material but far beyond my budget

It depends if you need cooling. But you can use old oven (or the one that you have in kitchen, but the wife should not see it ).
1. Common types handles temperatures up to 250°C. Definitely build not to catch wire, probably also to handle some fire inside.
2. already assembled
3. Maybe you can get one for free, only you need to clean it up. If you are lucky, than you will get hot air one with fan.

I do not expect issues with power consumption for temperatures below 100°C.
Downsides will be:
- hard to cool (in winter you can have it outside or put it into cellar)
- you need to regulate cca. 3kW heating @230V (maybe less if you use for example only grill, bottom, top heating, put heaters to series ...)
- bigger power consumption in comparison with EPS solution
- big and heavy in comparison with EPS (by the way fire retardant only means that if will stop burning if you take away the heat source away, melted down EPS will burn like hell and sticks to everything - I have seen video in German TV, where firemen explained the big issues with EPS insulations of the houses)

You must choose between price/weight/flammability/size in your case.
As written above you can build double walled box, filled with EPS. You only need to put there some non flammable spacers to ensure, that during the fire, when EPS will melt down, only the inner box will be heated. Instead of EPS you can use the high density mineral wool which is sold in the same sizes as EPS (it is used for insulation of houses from outside), only the price is higher and it is not so rigid, but definitely easier to handle than the fluffy rock wool and good enough to support inner box. The materials that are used in construction have easy to access datasheets including their insulation properties and price is low.

In case that you will have the fire inside the chamber you need to solve the situation that the melted material can leak outside and start the fire (this is probably valid also for commercial chamber). Also you need to turn off the electricity going inside in such a case, use fire retardant cables or non flammable insulation on bare wires (in our case teflon insulation will be probably good enough) and probably some other measures need to be done. All depends on how much fireproof you want to be.