-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
Build a Peltier PSU
- Thread starter silvershad0w
- Start date
- Joined
- Mar 19, 2001
- Location
- Bakersfield,CA
keep up the good work!! I cant wait to build myself a powersupply!!
can you give us a idea of how much this whole thing is going to cost? maybe $15-20 or $20-30?
hell I dont even care if its a $100 I just want to build something with my own 2 hands!
can you give us a idea of how much this whole thing is going to cost? maybe $15-20 or $20-30?
hell I dont even care if its a $100 I just want to build something with my own 2 hands!
- Joined
- Mar 24, 2002
- Location
- Plymouth, MN
I think this thread needs a bump. I am interested in the progress of the psu and I hope the project hasn't been scraped.
- Joined
- May 13, 2002
- Location
- Disney Land, CA
homer simpson type of voice:
oscope .. me want oscope
BTW: 31S1C (Satellite Network Controller)
ETSing 22 Aug 02 ...
oscope .. me want oscope
BTW: 31S1C (Satellite Network Controller)
ETSing 22 Aug 02 ...
http://www.mpja.com/ is a good source for electronic parts.
- Joined
- Mar 24, 2002
- Location
- Plymouth, MN
I can't wait. I get stuff from www.allelectronics.com.
- Joined
- Aug 25, 2002
- Location
- bleek abyss
Bump
silvershad0w said:
A switching PSU is not easy to design and build. Testing and
safety are BIG concerns for any PSU. Some of you
guys think that this a big free lunch project. Don't!!!
There is a reason that power supplies for Pelts are so expensive
and it is not greed. You guys who are not EEs have a good chance
of coming home to a dead system. Ebay is way cheaper.
Don't get the idea that I am against this project. I am not. I have
built several power supplies. Getting the design right is especially
tough. Linear power supplies are just power hogs.
http://forums.procooling.com/vbb/showthread.php?s=&threadid=4269
An older post I made that mgiht be of interest. Perhaps you could use some of the info contained within.
An older post I made that mgiht be of interest. Perhaps you could use some of the info contained within.
Wow, I just checked out this link. It's the best DIY article I have
ever seen, too.
Do heed the warning about EMI/RFI when picking the switching
frequency. Annoying sub-harmonics are also possible in the
audio band. A beast like this could easily make your great sound
system sound like...well, you know what.
The design output voltage may still not be optimal for your TEC.
Check the expected operating point of your TEC to see if it's
within acceptable limits.
ever seen, too.
Do heed the warning about EMI/RFI when picking the switching
frequency. Annoying sub-harmonics are also possible in the
audio band. A beast like this could easily make your great sound
system sound like...well, you know what.
The design output voltage may still not be optimal for your TEC.
Check the expected operating point of your TEC to see if it's
within acceptable limits.
I read, (actually, more like skimmed through) the front page article.
Overall a nice article.
A few comments though.
This is not likely true, and without use of a proper fuse, might be dangerous. Any given transformer core will only support a certain voltage per primary turn before the core will saturate. Once the core saturates, the impedance of the core will effectively drop to the resistance of the primary winding itself. If this occurs, the result will be huge current draw by the primary during a portion of the AC sine wave.
There are transformers that are designed for 120V or 240V and 50/60 Hz operation. They have two separate primary windings. When using them with 120V you connect the two primary windings in parallel. When using them with 240V you connect the two primary windings in series. (They also have a large enough core to meet their specs at 50Hz.)
I've seen recommendations from TEC manufacturers to keep the the ripple below 10% to avoid shortening the TEC's lifetime. Considering the result of TEC failure in this case, heeding this recommendation seems wise to me.
Have you measured the efficiency of your supply? If you run a 226 Watt TEC 24/7, the money paid for wasted power would probably pay for one of these in well under a year.
Overall a nice article.
A few comments though.
This is not likely true, and without use of a proper fuse, might be dangerous. Any given transformer core will only support a certain voltage per primary turn before the core will saturate. Once the core saturates, the impedance of the core will effectively drop to the resistance of the primary winding itself. If this occurs, the result will be huge current draw by the primary during a portion of the AC sine wave.
There are transformers that are designed for 120V or 240V and 50/60 Hz operation. They have two separate primary windings. When using them with 120V you connect the two primary windings in parallel. When using them with 240V you connect the two primary windings in series. (They also have a large enough core to meet their specs at 50Hz.)
I've seen recommendations from TEC manufacturers to keep the the ripple below 10% to avoid shortening the TEC's lifetime. Considering the result of TEC failure in this case, heeding this recommendation seems wise to me.
Have you measured the efficiency of your supply? If you run a 226 Watt TEC 24/7, the money paid for wasted power would probably pay for one of these in well under a year.
- Joined
- Mar 21, 2002
- Location
- pittsburgh pa
hey the best supplier for electronic parts i know is allied electronics.. i want a new pelt power supply to power dual 226's....
Since87 said:
I second your comments.
silvershad0w, the only other thing I might add is that after you
have further refined the design you might want to have one
of the online printed circuit places make a few boards for those
who want to build your design. Just be careful the traces are
wide and thick enough.
Good job.
- Joined
- Jul 6, 2002
- Location
- Good old UK
Not to leap in and start slating silvershadows article - a good read by the way - but if you know the reason why something doesn't work, and not just that it doesn't, it often helps. So, would a transformer meant for 120V work on 240V, simply doubling the output? No. Here's why:
US houeshold mains is, as stated, 110 - 120 volts, at 60Hz. The frequency - 60Hz, is the important factor. It means that a transformer is built to operate at 60Hz, and likewise, a British transformer is built to operate from 240V at 50Hz. So what happens in a transformer? The alternating voltage across the input terminals allows current to flow through the primary coil, causing a change in the magnetic field strength. The coil(s) on the secondary are in this changing magnetic field, and so when a load is connected across the secondary, current flows through the secondary coils and powers the load, basically. You need an iron core to make it work too. This is where the "catch" is. The more power you need, the more iron you need, for a bigger magnetic field, to transfer more power to the secondary. You also need more iron if the frequency is lower or your transformer core *** saturates. ***
Imagine you are pumping water into buckets, that are 1 gallon, and fill in one minute. It takes you a minute to take a full bucket to somewhere else and come back with an empty bucket to swap for the one you're filling. Now imagine you try this with 1/2 gallon buckets. Same pump, same time to take the water where you want it. By the time you get back though, the bucket has overflowed, and your floor is wet. Dang.
The size of the bucket is analogous to the size of your transformer core.
So, with the American supply being 60Hz, you can have slightly smaller transformers than us Brits - because at 50Hz, taking 16.67% longer to reach the peak of the mains cycle, a transformer with not enough iron in the core would saturate. What happens when it saturates? Well, when you reach saturation point, the core is holding as much magnetic energy as it can, and so the transformer coils begin to act more like a resistor, and dissipate the extra energy as heat. Bad news obviously. Your transformer draws more current, overheats, and if it doesn't blow a fuse, burns.
There ya go. By the way, you can plug a brit transformer into american supplies if it has a 120V tap - the dual primary winding thing posted earlier about the windings in series or paralell, for 240 or 120 volt operation respectively is fine as long as the core is rated to work at the frequencies you're using it on.
PC power supplies though - that's a whole different ball game. The transformers in them are switched at anything up to 300 kilohertz or so, that's why they can be so small. and they're ferrite, not iron, which eliminates eddy currents. Another pitfall of iron cored transformers.
Oh yeah, and if you're thinking of going ahead with a PWM design silvershadow, look at the TL494 and SG3524 IC's - or failing that, you can make a 555 timer IC pulse width variable. 555's man, they're a godsend, you can do TONS of stuff with those. PM me if you want any tips or help, or I'll post a design up. I'm going switched mode myself, gonna switch 600 watts with mosfets and some homebrew controller. Project's in the pipeline. After my radiator. I'm building that at the mo.
US houeshold mains is, as stated, 110 - 120 volts, at 60Hz. The frequency - 60Hz, is the important factor. It means that a transformer is built to operate at 60Hz, and likewise, a British transformer is built to operate from 240V at 50Hz. So what happens in a transformer? The alternating voltage across the input terminals allows current to flow through the primary coil, causing a change in the magnetic field strength. The coil(s) on the secondary are in this changing magnetic field, and so when a load is connected across the secondary, current flows through the secondary coils and powers the load, basically. You need an iron core to make it work too. This is where the "catch" is. The more power you need, the more iron you need, for a bigger magnetic field, to transfer more power to the secondary. You also need more iron if the frequency is lower or your transformer core *** saturates. ***
Imagine you are pumping water into buckets, that are 1 gallon, and fill in one minute. It takes you a minute to take a full bucket to somewhere else and come back with an empty bucket to swap for the one you're filling. Now imagine you try this with 1/2 gallon buckets. Same pump, same time to take the water where you want it. By the time you get back though, the bucket has overflowed, and your floor is wet. Dang.
The size of the bucket is analogous to the size of your transformer core.
So, with the American supply being 60Hz, you can have slightly smaller transformers than us Brits - because at 50Hz, taking 16.67% longer to reach the peak of the mains cycle, a transformer with not enough iron in the core would saturate. What happens when it saturates? Well, when you reach saturation point, the core is holding as much magnetic energy as it can, and so the transformer coils begin to act more like a resistor, and dissipate the extra energy as heat. Bad news obviously. Your transformer draws more current, overheats, and if it doesn't blow a fuse, burns.
There ya go. By the way, you can plug a brit transformer into american supplies if it has a 120V tap - the dual primary winding thing posted earlier about the windings in series or paralell, for 240 or 120 volt operation respectively is fine as long as the core is rated to work at the frequencies you're using it on.
PC power supplies though - that's a whole different ball game. The transformers in them are switched at anything up to 300 kilohertz or so, that's why they can be so small. and they're ferrite, not iron, which eliminates eddy currents. Another pitfall of iron cored transformers.
Oh yeah, and if you're thinking of going ahead with a PWM design silvershadow, look at the TL494 and SG3524 IC's - or failing that, you can make a 555 timer IC pulse width variable. 555's man, they're a godsend, you can do TONS of stuff with those. PM me if you want any tips or help, or I'll post a design up. I'm going switched mode myself, gonna switch 600 watts with mosfets and some homebrew controller. Project's in the pipeline. After my radiator. I'm building that at the mo.
- Joined
- Jul 6, 2002
- Location
- Good old UK
Just read Since87's post up there... Sorry to repeat it in more detail, Just didn't notice the complete post until after I posted.
Oh well
Oh well
Similar threads
