Hi, I want to design a power supply with outputs 0-12v @ 5amp (variable), 5v @1mp(fixed) and 3.3v @ 1amp(fixed). So, I am requesting to please give me schematic diagram and component values. Thank you.
It is actually very simple. If you can design a single power supply then this should not be a problem at all. However here is something that might be of help (attached to this message). There may be some component values that you have to determine - for example the resistor values for the 0-12V variable section (R1 & R2) - the ones I have used probably will allow variation from 0-6V (not sure). Refer to the LM317 datasheet for application examples - that will give you the formulas to determine the resistors. _[Edit: The schematic is not 100% accurate. The datasheet should have the correct high-current schematic for the variable 5A supply. The one here is made up to show the general idea.]_
The resistor for the zener (R4) may need to be calculated too - I have put that based on a hunch.
Also note that there is a caveat in the circuit in that if you draw 1A from both the 5V & 3.3V outputs, U2 (7805) will almost surely get destroyed - how to avoid that is something I leave to you as an exercise, and it will require redesigning of the circuit. As it stands the full 1A is available either from the 5V or the 3.3V output, but not both...and is limited to a max of 500mA/output in case of simultaneous loading of both. For practical purposes, do not expect the 7805 to give anything more than about 800mA total (max of 400mA each from the 5 & 3.3V outputs when loaded simultaneously) before it overheats and this is heavily dependent on the heatsinking.
If you plan to build this circuit than make sure that all the ICs and transistors are mounted on adequately sized heat-sinks.
_Note that you can simply repeat the 12V section with LM317 and use fixed resistors to get the required outputs - the above shows how you can use different types of circuit topologies to achieve your goal._
Well, there's important information missing such as the input voltage and AC or DC. At the currents you require I'd suggest a switching power supply. You can get schematics and parts lists based on your requirements with various switching power supply design programs on the net, such as webench - that's what I'd do. I'm not sure about a variable power supply down to 0v. When I last built a variable output power supply I used webench with a range of output voltages, in my case from 2 to 5v and looked at the results to pick resistors and a potentiometer to cover the variations.
1. The minimum load current on an LM317 is 5ma. 2. There is no proper feedback path for the LM317 to regulate this mess 3. The output will not go to 0 Volts (more like .55V - but that will be load dependant because of the lack of regulation)
The TI LM317 datasheet has circuit ideas for high current adjustable regulators and for how to get an LM317 to adjust to 0V (normally it will only go down to 1.25V):
Here's a few full diagrams for an adjustable high current supply (but the claim that it will do 0-30 is wrong. More like 1.25 - 30, but you can modify it using the technique illustrated in the datasheet. also, for 5A, remove all but two of the TIP3055s.:
Adjustment to 0V is obtainable (fairly easily) -- I have a bench supply that I built that does just that (using the technique offered in the LM317 datasheet).
The beauty of using an LM338 is you will get all the short circuit and thermal protection that are harder to achieve with a series pass transistor on an LM317 solution.
Yes, these circuits may well not work - and need some tweaking practically. Basically I wanted to convey some ideas instead of spoon-feeding. Going by his original post, I figured that he probably needs to experiment on his own rather than being given exact working schematics. That is why I added the note about referring to the datasheet(s).
Yes, but the circuit you provided for a 0-12V adj supply is hugely misleading in that it won't work (or if it does, it will be like a broken analog clock that is correct only twice a day).
Besides, is this EEWeb University (_I leave the rest as an exercise_)?
Agreed that it might be wrong but then a working ckt was not my intention anyways. I wanted to get him to think about it. If it doesn't work, fine - he will come back saying it did not work. But then he'd be better informed.
About leaving the rest as an exercise - my way of telling him to figure out what modifications to do. I am sorry if the language seemed improper.
All said, if what I wrote is unacceptable or seemed too overbearing/stupid then by all means have my posts deleted.
I have added the appropriate disclaimer in my first post. Hope that will alert anyone to its pitfalls.
The discussion addresses the design of a multi-output power supply providing 0-12V at 5A (variable), 5V at 1A (fixed), and 3.3V at 1A (fixed). Key challenges include achieving a true 0V adjustable output using linear regulators like the LM317 or LM338, which typically have a minimum output voltage around 1.25V due to their internal reference. Techniques from the LM317 datasheet can enable adjustment down to 0V, but require careful circuit modifications. The LM338, a 5A variant of the LM317, offers built-in short circuit and thermal protection, making it suitable for the high current variable output. The fixed 5V and 3.3V outputs are commonly implemented with 7805 and similar regulators, but simultaneous 1A loads on both may risk damage without proper design considerations. The discussion suggests that a switching power supply might be more appropriate for the required currents and efficiency. Several schematic references and design resources are provided, including links to datasheets and example circuits for high current adjustable supplies using LM317/LM338 and pass transistors like TIP3055. The importance of calculating resistor values for voltage adjustment and zener diode biasing is emphasized. Overall, the conversation encourages experimentation and consulting datasheets rather than relying on incomplete or approximate schematics. Summary generated by the language model.
