i have been working on a linear bench power supply, and have come to the conclusion that the cost of parts will be more than what i cam buy an assembled one for, however i will share what i have come up with so far so that any one interested will be able use what i have started.
the idea is to have 0-20v adjustable voltage and a 0-5A current limit/shutdown
the nor gates are arranged to in a way that there is an over-current tripped indicator
a reset button, and an load off switch.
at current the voltage regulator will not go below the zener voltage, a better reference system needs to be used to allow voltage down to null.
my hope was to to use only 2 ICs and have an accuracy of 1mv and 1mA using 4.5 digit volt meters.
the lm324 will output from nearly 0 to vcc-1.5v
the lm7805 is for the 74ls02 and relay.
this is being released CC BY-SA v3.0 http://http://creativecommons.org/licenses/by-sa/3.0/
I'd love to read a bit about this design, maybe an explanation and some of the background on your design decisions. Would you be willing to do a write up?
i can cover some sections and the theory around it
for the voltage regulator IC1A is feeding a voltage to the base of Q3,
R19 and the voltage pot is an adjustable resistor divider, being fed into the negative input of IC1A.
R15, D1 and C5 form a reference voltage into the positive input of IC1A.
The configuration of the op amp is a non inverting, feedback is taken on the output side R3 so the burden voltage will not affect the output voltage.
R3 is an 0.01 ohm 5watt 0.1% power resistor. at 5 amps the voltage across R3 is 50mV and dissipates 0.25 watts.
R1, R2, R19, and R20 with IC1b form a differential amp. measuring the voltage across R3 and preforming ohms law by multiplying the voltage by 100. allowing 1 volt output to equal 1 Amp.
R6, IC1C and the current pot form a voltage follower for the current limit.
IC1D is used as a comparator. comparing the current and shutdown current, when the voltage on the positive input equals or exceeds the voltage on the negative input, IC1D briefly outputs high.
ic5a and ic5b form a flip-flop when an over current occurs ic5d is used to light an indicator of over current condition.
ic5c allows the output to switch off through a relay driven by Q1, and allows manual shutdown of output.
the current set, output voltage, and current can all be read by 1:1 volt meters allowing any meter to be used.
i have thought of using a micro controller and replacing a couple of op amps and the nor gates, but is there is a bug in software it could be catastrophic if an over current does not trip correctly.
i have breadboard the circuit, and the response time of the over current is about 10-12µS much faster than a low power micro codded in c could accomplish, and 99.9% more reliable.
i hope this will answer your questions, and if you build one post some pictures id love to se it fully working
Thanks, that really helps. I have a lot of learning to do about analog design, esp. op amps. Can you explain why you're passing the center tap of the transformer all the way from input to output?
the board is designed not ho hold the transformer, and to be isolated, the middle is an earth ground.
using two of these in series the center can be tied to earth and you could get a + ground - dual supply. or you can have 2 common grounded supplies.
For what it is worth, the complete designs for very nice HP analog DC supplies are actually in the manuals for the old 80s models. Worth a look if you want to design a really nice DC power supply.
Personally I think a manual adjust it is more sensible to buy one, now if you want a computer controllable one that is another matter... Because the agilents of the world will charge you a million dollars for such things it seems, I have not published any of my designs of such things (and won't) because of lawyers.
[quote author="sqkybeaver"]the board is designed not ho hold the transformer, and to be isolated, the middle is an earth ground.
using two of these in series the center can be tied to earth and you could get a + ground - dual supply. or you can have 2 common grounded supplies.[/quote]
Can you post an example of such a connection?
would be necessary a center tapped transformer?
would a 200 VA dual 20V((20*1,4)-1,4 =+-26,6V after rectification) toroidal transformer suffice ?
What would be the changes to the circuit to have, lets say +-35V@5A output?
as it is setup it will handle 5A but the lm324 is only able to handle about 28v max supply.
the 5v regulator is used as a stable reference for the voltage regulator and the current limit. the transistor has to meet a certain voltage before current will flow. so the max voltage will be a volt or 2 lower than the input voltage. same as the lm324(Vcc - ~1.5)
the input is setup to take isolated dc. and should not be used on a multi-tap transformer.
to have a dual voltage supply, you would need an transformer with 2 isolated outputs, and 2 of these circuits.
i am thinking about having a few boards created.
[quote author="sqkybeaver"]the 5v regulator is used as a stable reference for the voltage regulator and the current limit. the transistor has to meet a certain voltage before current will flow. so the max voltage will be a volt or 2 lower than the input voltage. same as the lm324(Vcc - ~1.5)[/quote]
So if I replace the lm324 with the lf147 which has a max voltage of 44V i should be able to use higher voltage transformer, right?
There are modification to be performed on the scheme in order to change the Opamp?
[quote author="sqkybeaver"]to have a dual voltage supply, you would need an transformer with 2 isolated outputs, and 2 of these circuits.[/quote]
sorry to bother you but I can't figure out how to use two of this circuit to obtain a dual power supply, I followed you down to using two isolated secondary of the transformer, but then you have to connect the two regulating circuits together somehow, in order for the currents to circulate correctly, right? how do you do that?
Thanks.
I thought I'd piggy-back on this thread although my PSU will be a little bit different.
I've got an SMPS from an old portable CDROM drive which gives out 12v @ 2A and I'm looking to make a small digitally controlled lab supply from it. I'll be using a 16x2 LCD and push buttons for the user interface. The PSU would have fixed 5v, fixed 3.3v and variable 0-12v options available. The fixed supplies would be able to source about 300mA. Ideally, I would like all supplies to have a current control feature, but that may make things too complicated. So to keep things easy I'll have current control on just the 12v. I've see the schematics for the PSU above an was wondering why a MOSFET can't be used instead of the TIP3055? Wouldn't that be better?
mosfets are typically used in switching circuits. you will end up with unpredictable results using a mosfet.
with a darlington or standard transistor you can adjust voltage much easier.