Pelt: Sub-Ambient but not Sub-Zero
- Thread starter JN
- Start date
Mr. Smith
New member
I'll get back to you on this one as I did something to mine and now performance is awful - idles about 1 deg and full load mid to high 20's...
I think I knocked it and it isn't seated properly, not really what you were looking for...
If you give the pelt less wattage then it doesn't perform as well, so this would give you your desired effect.
Perhaps a 'pot' on the X3 to turn down the power.
I think I knocked it and it isn't seated properly, not really what you were looking for...
If you give the pelt less wattage then it doesn't perform as well, so this would give you your desired effect.
Perhaps a 'pot' on the X3 to turn down the power.
Mr. Smith
New member
Hmmm, internally would be some huge resistor, or maybe ghetto mod a large potentiometer - isn't a dimmer switch essentially capable?
I know a sparky but I'd have to get him over and open up my booster to get any better ideas... I only did gcse electronics and that was 8 years ago lol
I know a sparky but I'd have to get him over and open up my booster to get any better ideas... I only did gcse electronics and that was 8 years ago lol
This is something i've been wondering as well. If looks did not matter is it possible that contecting the pelt to a variable power supply, like this http://www.maplin.co.uk/Module.aspx?ModuleNo=30391&criteria=power supply&doy=29m7
would do the job. I appreciate that the pelt you are talking about may have different specs from the swiftech 6500 that I am looking at but is it possible?
Dave
would do the job. I appreciate that the pelt you are talking about may have different specs from the swiftech 6500 that I am looking at but is it possible?
Dave
quietfreek
New member
XMS what power TEC are you looking to control ?
Are you looking to to apply the TEC directly to the IHS or are you looking to cool the liquid in a W/C loop to ambient as dell do in there H2C which is much more efficient as you arnt looking to pump the full TDP of the PROC through the TEC.
Are you looking to to apply the TEC directly to the IHS or are you looking to cool the liquid in a W/C loop to ambient as dell do in there H2C which is much more efficient as you arnt looking to pump the full TDP of the PROC through the TEC.
quietfreek
New member
Holy 
thats a lot of juice
i would want to be building a PWM controller, do you have any experience in power electronics?
are you driveing this from your PC PSU ?
if you budget will allow id use something like the m-cubed tbalencer
http://www.t-balancer.com/english/bng.htm
use this to create a PWM signal with a custom profile based on ambient temp and proc. temp sensors. Id then input this into the base of what is esentially a power amplifer, a suitably spec'ed darlington driver / mosfet (s) that can cope with the current you need, As your not driving an inductive load the circuit would quite simple, with a little care and posably a couple of capacators to smoth the current draw from then power supply and delivery voltage you can have very low power loss control of you pelt / cpu temp tuneable in windows and microprocessor controled.
if your really interested ill pull out some of my old 3rd year electronic eng. notes and spec up some components / sugested circuit total cost shouldnt be too much.
thats a lot of juicei would want to be building a PWM controller, do you have any experience in power electronics?
are you driveing this from your PC PSU ?
if you budget will allow id use something like the m-cubed tbalencer
http://www.t-balancer.com/english/bng.htm
use this to create a PWM signal with a custom profile based on ambient temp and proc. temp sensors. Id then input this into the base of what is esentially a power amplifer, a suitably spec'ed darlington driver / mosfet (s) that can cope with the current you need, As your not driving an inductive load the circuit would quite simple, with a little care and posably a couple of capacators to smoth the current draw from then power supply and delivery voltage you can have very low power loss control of you pelt / cpu temp tuneable in windows and microprocessor controled.
if your really interested ill pull out some of my old 3rd year electronic eng. notes and spec up some components / sugested circuit total cost shouldnt be too much.
quietfreek
New member
Thats good its less critical that the current draw is ripple free if its off a 2nd PSU.
Does anyone have any contacts with M-Cubed i'd need to know the switching frequency of there PWM signal to spec the mosfets
Does anyone have any contacts with M-Cubed i'd need to know the switching frequency of there PWM signal to spec the mosfets
quietfreek
New member
cool approx 500 hz is not going to be a problem - you can get away with big dirty relatively slow power electronics transistors.
although the phrase "It is only a reference for global orientation and does not reflect real pulses per second" does worry slightly.
Ill have a bit of poke around on there forum see if i can find anything
although the phrase "It is only a reference for global orientation and does not reflect real pulses per second" does worry slightly.
Ill have a bit of poke around on there forum see if i can find anything
quietfreek
New member
excelent just been onto the m-cubed forums found this
http://www.mcubed-tech.com/forum/showthread.php?tid=606
is scemes that someone is trying to just what you want to do but it gives the signal as being max 900Hz - although the person in the post is worried about the signal not being high enough if the signal is smothed to provide effectively a class d amplifier with analogue output the pelt will be fine
Max power drop accross the transistors should only be 25A * 0.7v = 17.5w easyly pasively cooled and continiously variable
compared to 56w onto 0.16 ohm to take the pelt to 170w output with a potentiomiter
http://www.mcubed-tech.com/forum/showthread.php?tid=606
is scemes that someone is trying to just what you want to do but it gives the signal as being max 900Hz - although the person in the post is worried about the signal not being high enough if the signal is smothed to provide effectively a class d amplifier with analogue output the pelt will be fine
Max power drop accross the transistors should only be 25A * 0.7v = 17.5w easyly pasively cooled and continiously variable
compared to 56w onto 0.16 ohm to take the pelt to 170w output with a potentiomiter
