Open source DIY TEC and Fan controller

[hello im sean]

Hey Ultrasonic2 just chiming in to say I've learned a ton from your posts and website. You may not realize how many lurkers take in your valuable info. The ones who comment a lot of times are either trolls, or too full of themselves to acknowledge your tested facts.

[Liam_G]

I think the thread view count speaks for itself, but be careful of the "tested facts" part of that comment. Ultra is very intelligent but this is a work in progress, don't just rush out and build this and expect it to work perfectly. We are just lucky that ultra is doing the leg work for us and testing along the way to make sure the end result will do the job and not cause any hazards. I think regardless of how you get there, if it works, it works, so i'm eager to see the fully functioning unit to see where this design is at. I truly hope it works out the way ultra wants because it would be an amazing piece of work.

In the meantime i'm loving learning about TEC's and gleaning some knowledge of electronics along the way, although electrics is still a bit of a dark art to me, need to give myself a project that requires me to learn all this stuff. I understand what ultra is trying to do but i have no idea how to come up with this stuff in the first place. I'm at "read only" status with electrics, lol, need to get "write" permissions :P

PS, ultra, I haven't looked at first page of this post for a while but if you haven't already could you EDIT the first post to contain your current parts list and modus operandi to make sure new people to the thread know where you are at, I only say this because you have been evolving the design quite a bit and are honing in a bit more on final components. good luck!

[Ultrasonic2]

I think the thread view count speaks for itself, but be careful of the "tested facts" part of that comment. Ultra is very intelligent but this is a work in progress, don't just rush out and build this and expect it to work perfectly. We are just lucky that ultra is doing the leg work for us and testing along the way to make sure the end result will do the job and not cause any hazards. I think regardless of how you get there, if it works, it works, so i'm eager to see the fully functioning unit to see where this design is at. I truly hope it works out the way ultra wants because it would be an amazing piece of work.

In the meantime i'm loving learning about TEC's and gleaning some knowledge of electronics along the way, although electrics is still a bit of a dark art to me, need to give myself a project that requires me to learn all this stuff. I understand what ultra is trying to do but i have no idea how to come up with this stuff in the first place. I'm at "read only" status with electrics, lol, need to get "write" permissions :P

PS, ultra, I haven't looked at first page of this post for a while but if you haven't already could you EDIT the first post to contain your current parts list and modus operandi to make sure new people to the thread know where you are at, I only say this because you have been evolving the design quite a bit and are honing in a bit more on final components. good luck!

hay guys sorry for the late reply oddly i haven't been getting any Updates from this site .

Thanks for your kind words guys.

Yeah it's a good idea to update the first page as it has changed alot and will still change as im making it up as i go along.

im still ......... Waiting on parts

[mindchill]

Here is an interesting arqtik chiller/controller if you don't want to do something diy

[Elloquin]

Even for you this was a pretty di$k move lol.

[pby5cat]

Im sorry i disappered, my login info got messed up.

Ultrasonic i have a way to vary the voltage to a peltier, but i need to know that is the current supplied to the peltier?

[Ultrasonic2]

about 25 amp for me or 50+amps at 12v

what idea did you have ?

[pby5cat]

Most buck controllers have a feedback pin that is used to keep the output voltage stable. The feed back pin gets a divided voltage of the output. so insted of using a fixed voltage divider like normal, you use a digital poteniometer. As the digital pot changes the feedback voltage the contoller adjusts the output to keep the feedback voltage the same as an internal refrence voltage. Have the digital pot talk to the arduino through SPI and you made a variable power supply. I cant seem to find one that is ok with 50A at 12 something volts.

Here is an example controller that has the feedback pin (FB) http://www.ti.com/product/lmz14203h

Recommended digital pot http://www.analog.com/en/digital-to-...s/product.html

Arduino SPI http://arduino.cc/en/Reference/SPI

[Ultrasonic2]

To much effort. Im probably going to use one of these PSU's which i can use the Arduino to control 20% up

http://www.meanwell.com/search/SPV-300/SPV-300-spec.pdf

Doing so means i dont need a low pass filter or a pwm circuit just one simple capacitor


this was Liam_G idea so he should have the credit

[Liam_G]

To much effort. Im probably going to use one of these PSU's which i can use the Arduino to control 20% up

http://www.meanwell.com/search/SPV-300/SPV-300-spec.pdf

Doing so means i dont need a low pass filter or a pwm circuit just one simple capacitor


this was Liam_G idea so he should have the credit

Hey, thanks to XS and you Ultra and also foxrena at OCN for sparking my interest in TEC's and confirming my crazy ideas and answering my noobish comments. I'm glad I've been able to contribute to this idea, now I just hope these PSU's do what we want :P

[pby5cat]

That is a way better solution, as it also isolates the TEC power from the computer power.

[mindchill]

That is a way better solution, as it also isolates the TEC power from the computer power.

Interesting solution if you want an external power supply.

[MadHacker]

Interesting solution if you want an external power supply.

I think it will fit in two 5.25 bays

[Ultrasonic2]

Well i had a bit of a bad weekend and ended up blowing up my arduino so im back to waiting on parts

however this is the circuit for the fan outs this will enable the 5v arduino to control 12v and results in a analogy (change in voltage) out instead of digital (change in current) and this could be used to power your TEC's if you want to like i had originally planed

actual fan low pass.PNG

If you just want a PWM out (PWM BOOSTER) you can make this

PWM Booster.PNG

[Liam_G]

bummer about the Arduino. keen to see some results with the new PSU idea though. good luck!

[Ultrasonic2]

bummer about the Arduino. keen to see some results with the new PSU idea though. good luck!

the whole thing has been put on hold as the company i work for went into liquidation

[Liam_G]

the whole thing has been put on hold as the company i work for went into liquidation

Sorry to hear that ultra. You're in IT right? Plenty of work about here in oz in the IT industry. NZ is doing it pretty tough at the moment.

[Blindsay]

Here is an interesting arqtik chiller/controller if you don't want to do something diy

That is pretty nifty, if i wanted to add a chiller to my loop to bring down the water temp a little bit (id be happy keeping it at ambient air temp) and as small and compact as possible. Do you think that would be my best bet? Its for a crazy mATX build im working on

[MadHacker]

That is pretty nifty, if i wanted to add a chiller to my loop to bring down the water temp a little bit (id be happy keeping it at ambient air temp) and as small and compact as possible. Do you think that would be my best bet? Its for a crazy mATX build im working on

adding a chiller won't make it more compact, if anything it will add to the size.

[Blindsay]

adding a chiller won't make it more compact, if anything it will add to the size.

I know, im saying i want the chiller itself to be as compact as possible, ive seen some huge chillers that are housed in their own case

[Ultrasonic2]

i wouldn't conceder mine compact but everything fits in

[chaotic]

Any update here?

[Ultrasonic2]

No im still getting parts together

I've had a failed TEC for ages so it was time to sort that out and now im replacing my rads and may add another TEC water block

So i've got a 200mm rad up front and im saving for a 200 x 400 uptop. I getting some custom 200mm fans in the post right now. Then i need to sort out the failed TEC and then i still need to buy a new PSU to power everything so there's plenty of money still to be spent

[Ultrasonic2]

right this project is making a come back i've ordered new part and am waiting on them they'll probably take a month to get here. In the mean time i've been re thinking what temp to keep my load at

1 keep temp at dew point
the problem i have with this is that when the humididty is low this temp could be quite low and therefore require significant power to achieve this and when i load up the cpu there will be a significant increase in temps.

2 keep temp x as long as above dew point
same as above but not as bad. it will still mean lots of power in summer and in winter my target temp may be above ambient.

3 keep cold temp X degrees below hotside temp.
This seems like the best idea as tec power usage will be the same year round and so will fan noise. Also the TEC's will always be having a positive affect on temps.
at this point im going to try for a 10c delta but im worried that will still be a fear amount of power at idle

[bds71]

the Boreas system i use is based on static (water) temp (like 2 above) - ie: it keeps the cool-side water at a set (but adjustable) temp whether at load or idle. i like it, but here in germany the temperature simply varies too wildly for it to be effective (no A/C so hot in summer - sometimes in excess of 30-35C, but cool/warm in winter with heating available). personally (for germany - due to the hugely variable ambient temps) i like the 1st idea. if the room is cool enough and the load is light (idle) you wouldn't use much electricity at all. but yes, in the summer (assuming variable temps like here in germany) yea - could be expensive running it that way, but still cheaper than 2 (if temps get above set point you run at 100%, and even if temps are reasoanable you still run a high dT (water to ambient) so not worth the effort to use TEC really). 1 & 3 are really your best options for year round use (my personal opinion of course) with 3 being the most cost effective/efficient, but least effective performance - wise