Hello everyone,
I was wondering if anyone ever thought of using peltier cells to have a power supply for sensible part of the audio circuits. I'm not sure of how much power a DAC would drain, apparently a DSD1794 would just need 10ma.
I saw once someone using two fotovoltaic panels and a really powerful light bulb to supply the galvanic isolated DAC side of the circuit board, apparently was too big as a system and not very efficent.
My idea was to have a "sandwich" with a peltier cell in the middle, then over it another peltier cell from the cold side (cooled with a fan) and a 10w resistor as heat radiator under it.
I guess in this way you would get enough cold and hot so that the cell will produce enough energy to power up the converters ICs.
Make it X2 and you can have dual power supply.
My only question is if temperature noise will play a big part here.
Probably someone more technical than me would know. 🙄
I was wondering if anyone ever thought of using peltier cells to have a power supply for sensible part of the audio circuits. I'm not sure of how much power a DAC would drain, apparently a DSD1794 would just need 10ma.
I saw once someone using two fotovoltaic panels and a really powerful light bulb to supply the galvanic isolated DAC side of the circuit board, apparently was too big as a system and not very efficent.
My idea was to have a "sandwich" with a peltier cell in the middle, then over it another peltier cell from the cold side (cooled with a fan) and a 10w resistor as heat radiator under it.
I guess in this way you would get enough cold and hot so that the cell will produce enough energy to power up the converters ICs.
Make it X2 and you can have dual power supply.
My only question is if temperature noise will play a big part here.
Probably someone more technical than me would know. 🙄
Why not just use a well designed standard power supply linear or SMPS instead of some whacky idea with peltier cells
Of course you do have to build a power supply for the fan and resistor!
Of course you do have to build a power supply for the fan and resistor!
http://www.diyaudio.com/forums/digital-source/142562-microsd-memory-card-transport-project-62.html
You should read from post n.616,
Two users of this forum tried powering their DACs with a solar panel stating that it influenced the sound less than using batteries.
I really believe that having galvanic isolated clean power supply for the parts of the circuit very sensible to audio like DACs or ADCs does make a difference, so I thought using peltiers instead of bulbs and solar panels was a even tinier, more efficient way to make this kind of power supply.
As the DAC circuit used has already galvanic isolated digital audio inputs, seems right to keep it isolated from the building's AC line that can be carrying noise.
A pro PSU that would be able to filter out any noise in the line maybe it would end up costing a bit more, and the peltier solution maybe can be the poor DIYer way to get around the problem! 🙂
You should read from post n.616,
Two users of this forum tried powering their DACs with a solar panel stating that it influenced the sound less than using batteries.
I really believe that having galvanic isolated clean power supply for the parts of the circuit very sensible to audio like DACs or ADCs does make a difference, so I thought using peltiers instead of bulbs and solar panels was a even tinier, more efficient way to make this kind of power supply.
As the DAC circuit used has already galvanic isolated digital audio inputs, seems right to keep it isolated from the building's AC line that can be carrying noise.
A pro PSU that would be able to filter out any noise in the line maybe it would end up costing a bit more, and the peltier solution maybe can be the poor DIYer way to get around the problem! 🙂
Hello everyone,
I was wondering if anyone ever thought of using peltier cells to have a power supply for sensible part of the audio circuits. I'm not sure of how much power a DAC would drain, apparently a DSD1794 would just need 10ma.
I saw once someone using two fotovoltaic panels and a really powerful light bulb to supply the galvanic isolated DAC side of the circuit board, apparently was too big as a system and not very efficent.
My idea was to have a "sandwich" with a peltier cell in the middle, then over it another peltier cell from the cold side (cooled with a fan) and a 10w resistor as heat radiator under it.
I guess in this way you would get enough cold and hot so that the cell will produce enough energy to power up the converters ICs.
Make it X2 and you can have dual power supply.
My only question is if temperature noise will play a big part here.
Probably someone more technical than me would know. 🙄
Good idea.
I have bought some peltier elements for evaluation of this use.
Will be interesting to check this out and see if it works at all 🙂
The analogue JFET I/V - outputstages burns off enough heat as it runs class A 0.5 Watt into 8 ohms load, and the JFETS needs cooling..
As the Peltier elements have a ceramic surface I can mount the JFETs directly.
If this works I get two Peltier PSUs from the left and right channel, and possibly another two Peltier PSUs from the rectifier diodes heat..
Did a quick test with the first Peltier element I found..
I do not know the spec or part number,
but heating it from the top of the heatsink of a passively cooled PC with below 40 degree celsius created 0.5 volt and 30mA. Flipping it and heating the other side gives -0.5 volt...
3.3 volt and 200 - 300mA looks to be possible with the power source I have in mind with a Peltier element optimal for this use.
I do not know the spec or part number,
but heating it from the top of the heatsink of a passively cooled PC with below 40 degree celsius created 0.5 volt and 30mA. Flipping it and heating the other side gives -0.5 volt...
3.3 volt and 200 - 300mA looks to be possible with the power source I have in mind with a Peltier element optimal for this use.
Last edited:
Ordered 20 pcs peltier elements so I have some to play with.
The goal are to use the heat from the analogue JFET output stage (and increase that from 0.5 to 1 Watt RMS output into 8 ohm. The increase are just a doubling of the bias current) and power each mono DAC section from each of the peltier elements. And make a test if the rectifiers can make enough heat to have a third peltier element powering the USB, SPDIF and re-clocker stages..
As I have some identical peltier elements already at hand I will build one class A JFET channel ASAP and test the voltage and current the peltier element will give, and make a peltier PSU and power the USB section as a test (if I get enough voltage in the first test).
The goal are to use the heat from the analogue JFET output stage (and increase that from 0.5 to 1 Watt RMS output into 8 ohm. The increase are just a doubling of the bias current) and power each mono DAC section from each of the peltier elements. And make a test if the rectifiers can make enough heat to have a third peltier element powering the USB, SPDIF and re-clocker stages..
As I have some identical peltier elements already at hand I will build one class A JFET channel ASAP and test the voltage and current the peltier element will give, and make a peltier PSU and power the USB section as a test (if I get enough voltage in the first test).
wow RayCtech, it's cool what you're doing!
I still wonder if to get enough energy we would need to push the delta T of the cell to its limits (around 60/70C). For this reason I imagined to use an even bigger sandwich of 5 these units where the peltier elements are sequentially 1 active (providing hot/cold via power supply) and 1 passive (generating electricity from delta T) in this way there would be the best heat transfer, best efficiency and.. enough power?
Does it makes sense to you?
I still wonder if to get enough energy we would need to push the delta T of the cell to its limits (around 60/70C). For this reason I imagined to use an even bigger sandwich of 5 these units where the peltier elements are sequentially 1 active (providing hot/cold via power supply) and 1 passive (generating electricity from delta T) in this way there would be the best heat transfer, best efficiency and.. enough power?
Does it makes sense to you?
The peltier elements I ordered today was just some $2.3 pr. pcs peltier elements from ebay 🙂
I will later calculate (with help from a manufacturer) the optimal peltier element for my use. "Later" due to I want to make some prototypes first and find the effect I actually will transfer to the peltier elements and also what cooling system I will use.. Then I will know the delta temperature the peltier will get etc.. I want to use one peltier element giving the voltage and current needed, and if that requires a custom made peltier element I will order that 🙂
http://cds.linear.com/docs/en/lt-journal/LTJournal-V20N3-01-df-LTC3108_09-David_Salerno.pdf
There are better ways of creating a supply! But the above gives some info where energy harvesting is done using peltier devices.
Not the best way to engineer a solution...the voltage out of the peltiers needs regulating, the temperature gradient needs regulating, with the low conversion rate you are going to have to supply power to create power, more supplies... true audiofoolery, have fun.
Marc🙂
There are better ways of creating a supply! But the above gives some info where energy harvesting is done using peltier devices.
Not the best way to engineer a solution...the voltage out of the peltiers needs regulating, the temperature gradient needs regulating, with the low conversion rate you are going to have to supply power to create power, more supplies... true audiofoolery, have fun.
Marc🙂
Last edited:
- Status
- Not open for further replies.