I need help building a power supply

[i am]

Hello everyone. I'm new to the forum so I would like to take me Lpervrom else in the case that this forum is not appropriate.

I would like to help build your money with a force following values​​:

The current limit control 0-124A
Limit control voltage V0-30

My power supply is welder values:

166A 47V Maximum Power

Under load of 124 amps voltage is 25V

i ask u to guidance. with project process. Thanks to all contributors

[AndrewT]

What is your language?
Maybe get someone to translate your questions to english.

What are you building?

[i am]

Originally Posted by AndrewT

What is your language?
Maybe get someone to translate your questions to english.

What are you building?

First My language is Hebrew. The second thing I'm building a power supply with current limiting and voltage limiting.

The main problem is the schema. I've done a schema but I do not know whether it will work. If you want to see it I will share it with you

the power supply is not a issu beacuse i bouth a 166A 51V SMPS . but the SMPS have not a voltage and current limitations . so i wanna biult the boost circuit that can push 0-124A and 0-30V reguleted. caz my SMPS isn't reguleted.and i need for the voltage section 2 pot . 1 pot is the max voltage that i allow the SMPS uotput and the 2 pot for 0-100% out of pot 1 . ( i need that with both sections voltage and cuurent ) so i need total pot of 4 . tha op-amp will be aLT1012 . and i let u tell me what the schem should be caz i dont know precisely what the Values should be . oh and the mosfets will be thous http://search.digikey.com/scripts/Dk...075T2&x=9&y=20

and thank u for ur comment

[hesener]

are you sure about 124A? At say 5V output there will be (51-5)V*124A = 5.7kW power dissipation in your regulator (if it will be a linear regulator, as implied by your component choice). that's a lot of heat....

please post your schematic!

[i am]

Originally Posted by hesener

are you sure about 124A? At say 5V output there will be (51-5)V*124A = 5.7kW power dissipation in your regulator (if it will be a linear regulator, as implied by your component choice). that's a lot of heat....

please post your schematic!

yup i am sure . look i edit my post . i add the mosfets . they can handel alot of power ( Easily 600W at 130C or 400W at 85C ) [IMG] Uploaded with ImageShack.us[/IMG]

that is the schematic. but i do not shure about the values . please fix me if i wrong

and thank u alott !!

[indianajo]

Welders don't use 5V much, but they do use 24 V a lot. You'll need some huge heat sinks and fans to lose 50% of the voltage at 124 A with linear techology. Actually for welding currents switch mode using fets to decrease the current might be much more useful. This schematic is all analog linear, which is 30 year old technology. Something like a lamp dimmer to decrease on time percentage from 100% to 33% maybe, but operating at 10 Khz instead of 50 hz. You could also weld AC and use triacs or back to back SCR's at 50 hz, delaying the turn on time in the phase with an actual lamp dimmer. Would make an awful RF noise in radios & TV's, however.

[i am]

Originally Posted by indianajo

Welders don't use 5V much, but they do use 24 V a lot. You'll need some huge heat sinks and fans to lose 50% of the voltage at 124 A with linear techology. Actually for welding currents switch mode using fets to decrease the current might be much more useful. This schematic is all analog linear, which is 30 year old technology. Something like a lamp dimmer to decrease on time percentage from 100% to 33% maybe, but operating at 10 Khz instead of 50 hz. You could also weld AC and use triacs or back to back SCR's at 50 hz, delaying the turn on time in the phase with an actual lamp dimmer. Would make an awful RF noise in radios & TV's, however.

Mm I think that you misunderstood me (or that I can not explain: P)

Look. I want a power supply that his OP is pure DC. Without switching.

Note that if I conect parallel to the power supply a load of 10-15A voltage dropped to 45V.
So the calculation is 45 * 124 = 5580W which means that the MOSFETS in 110C hold 523W each. So I'll have about 10 units of MOSFETS: P. While the load (10-15A fall just about 800W or 765W exactly)

agine . biggg thank u guys !!

[macboy]

Let's do a little thermal engineering.

First, 5580W split by 10 is 558W each MOSFET, not 523W.

So you have to dissipate 558W with a junction temperature of 110 degC. I'll assume the ambient temperature is 25 degC. That's a difference of 85. 85/558 = 0.152 deg/W, which means that your junction to air thermal resistance needs to be 0.152 degrees/W or less, in order to avoid overheating of the MOSFET junctions.

You need to add together all of the thermal resistances in the system: the MOSFET's internal junction-to-case, the case-to-heatsink (don't forget to include to add thermal resistance of any insulating washers), and the heatsink-to-air. That will give you the total thermal resistance (junction-to-air), which needs to be less than 0.152 deg/W. You will find this to be impossible. Just the junction-to-case is 0.16 in the datasheet (R_thJC). Instead, you are likely to find that the practical juction-to-air thermal resistance is about 1 deg/W. Assuming that value, you will need to dissipate no more than 85 W per MOSFET, which means you need to use 66 or more MOSFETs, not 10.

What you want to do is not impossible but you will find it to be quite a bit more difficult than you think.

[i am]

well. thank u for the comment .

and umm i am not the pro here . i came to learn from u guys . so i need ur shcematic for this specific circuit . ill thank u if u can post a shcematic that can handel that total 5580W . ( i dont know if u calculat that . the junction thermal resistance of the mosfets is 0.21 u can chek out on the detasheet )

agine thank u !! :P

[indianajo]

Originally Posted by macboy

. You will find this to be impossible. Just the junction-to-case is 0.16 in the datasheet (R_thJC). Instead, you are likely to find that the practical juction-to-air thermal resistance is about 1 deg/W. Assuming that value, you will need to dissipate no more than 85 W per MOSFET, which means you need to use 66 or more MOSFETs, not 10.
What you want to do is not impossible but you will find it to be quite a bit more difficult than you think.

I think you see why this is usually done with transformers instead of mosfets. Cost. If you have a 4:1 transformer (220 VAC:51 VAC) and you actually want 24 V, you could buy the proper transformer, or SMPS. (switch mode power supply). There is sometimes a cost benefit to using 2:1 "isolation" transformers, maybe you could put one of these in front of your 4:1 transformer for voltages below 24 VDC. Deciding what you want to do saves cost a lot. Charge truck batteries?- you want 26-27 VDC. Golf carts?- 38-40 VDC. Mig or tig weld? 24 VDC or AC. (tig requires a current limiter). refine bauxite into aluminum? 5 VDC at this amperage. A cheap supplier of transformers in the US is antekinc.com. You would obviously do better to buy in europe or asia- see various threads on this forum for the best european and far eastern transformer suppliers.