Hi.
I've been thinking about building a small dual PSU.
I'm looking for the following qualities:
- Tracking: neg rail will follow pos rail
- +/- 0-20V, 0-1A
- Current-limiting with a single pot to set both pos and neg rail.
- voltages shall track even when current-limit is active
- No (overly) temperature-sensitive components in voltage or current-reg.
I don't need:
- super-precise current-limit, but I would like it to be stable over time and temperature
- Super low noise and very precise voltage reg.
Inspiration was this:
Paul's DIY electronics blog: Tuning a 0..30V DC 0..3A PSU DIY kit
Based on those ideas, I have played with the idea below:
V1 and V6 are two X-former windings with smoothing (not ideal sim, but ok).
V11 sets current-limit. should be a pot.
V2 sets voltage.
V4, V5, V7 and V8 are voltages needed for the opamps.
U2 and U3 is the current-sensors (voltage-drop over R9 and R8 relative to gnd
U5 inverts the current-ref for the pos rail.
U4 tracks the pos rail.
V3 and V9 is used to simulate step-response and such.
What do you think? Is the idea valid? can the current-regulators work?
I've been thinking about building a small dual PSU.
I'm looking for the following qualities:
- Tracking: neg rail will follow pos rail
- +/- 0-20V, 0-1A
- Current-limiting with a single pot to set both pos and neg rail.
- voltages shall track even when current-limit is active
- No (overly) temperature-sensitive components in voltage or current-reg.
I don't need:
- super-precise current-limit, but I would like it to be stable over time and temperature
- Super low noise and very precise voltage reg.
Inspiration was this:
Paul's DIY electronics blog: Tuning a 0..30V DC 0..3A PSU DIY kit
Based on those ideas, I have played with the idea below:
An externally hosted image should be here but it was not working when we last tested it.
V1 and V6 are two X-former windings with smoothing (not ideal sim, but ok).
V11 sets current-limit. should be a pot.
V2 sets voltage.
V4, V5, V7 and V8 are voltages needed for the opamps.
U2 and U3 is the current-sensors (voltage-drop over R9 and R8 relative to gnd
U5 inverts the current-ref for the pos rail.
U4 tracks the pos rail.
V3 and V9 is used to simulate step-response and such.
What do you think? Is the idea valid? can the current-regulators work?
These type PS are called Master-Slave or Tracking regulators. They are excellent for audio apps. You should be able to find many such circuits under one of these names.
THx-RNMarsh
THx-RNMarsh
Hi. Yes the voltage-tracking is found in many schematics on the web. I don't claim to be first on that part.
I cannot find any simple tracking regulators with variable current limiting that limits both pos- and neg-supply with a single pot.
TroelsM
I cannot find any simple tracking regulators with variable current limiting that limits both pos- and neg-supply with a single pot.
TroelsM
Not sure about simple, but the HP 6114A can be wired in master-slave configuration. Using two 6114A you can get a very very very precise tracking setup.
Sent from my LGLS992 using Tapatalk
It looks promessing but I you need two separate windings on the trafo and two rectifiers.
If you like experimenting I have another circuit, just an idea, never tryed it.
The noname C's are perhaps nessesairy to came down the comparators ??
Have fun,
Mona
Attachments
Hi Ketje. That circuit looks a bit like an old Elektor-schematic from my collection (attached, hope it's OK with copyrights).
I don't see how your version works. Do you think that the comparators will work linearly? - or will they just cut voltage completely when activated?
Also not sure how the voltage regulation will work.
Looks very interesting though.
I don't see how your version works. Do you think that the comparators will work linearly? - or will they just cut voltage completely when activated?
Also not sure how the voltage regulation will work.
Looks very interesting though.
Attachments
Last edited:
Yes indeed, it looks a bit like my proposal.The LM317 is used directely to regulate the voltage from its ohne 1,25volt reference and the LM337 via an op-amp.
I regulate both sides with an op-amp (comparator= op-amp with open collector out), the reference is 10V on the Vpotm derived from the 7815.
The output from the op-amp goes from approx.-10...+10V for 0...20V out thanks to the 1k5 drop (has a constant current 1.25V/220Ω)
Elektor has two current sensor circuits coupled mechanically (tandem potm).
By inverting one side one can use a single (adj.)treshold voltage.The 100Ω potm gives 0..1V for 0...1A on 1Ω.(A bit better would be the output of another op-amp giving 0-1V)
Too much current pulls the voltage on the Vpotm down until current becomes acceptable.
Mona
I regulate both sides with an op-amp (comparator= op-amp with open collector out), the reference is 10V on the Vpotm derived from the 7815.
The output from the op-amp goes from approx.-10...+10V for 0...20V out thanks to the 1k5 drop (has a constant current 1.25V/220Ω)
Elektor has two current sensor circuits coupled mechanically (tandem potm).
By inverting one side one can use a single (adj.)treshold voltage.The 100Ω potm gives 0..1V for 0...1A on 1Ω.(A bit better would be the output of another op-amp giving 0-1V)
Too much current pulls the voltage on the Vpotm down until current becomes acceptable.
Mona
Hi Ketje. I've been looking a bit more and I think your schematic makes more sense now 🙂
I'm still not sure that a comparator is the best choice for a linear circuit, but maybe it will work.
Have you simulated it?
Troels
I'm still not sure that a comparator is the best choice for a linear circuit, but maybe it will work.
Have you simulated it?
Troels
Simulated ? no I don't like that stuff (too old ?)
There is something wrong with the schematic
It looks nice symetric but the output of a comparator isn't.The 1k5 from the LM317 is pulled down by the open collecor but the 1k5 on the LM337 side must be pulled UP ! An op-amp pulls both sides but here we need a emitter follower between.
Using real op-amps not comp's is another sollution but then the current limit need diodes like in Elektors schematic, you can't connect two ouputs together like with O.C. outputs.
Mona
There is something wrong with the schematic
It looks nice symetric but the output of a comparator isn't.The 1k5 from the LM317 is pulled down by the open collecor but the 1k5 on the LM337 side must be pulled UP ! An op-amp pulls both sides but here we need a emitter follower between.
Using real op-amps not comp's is another sollution but then the current limit need diodes like in Elektors schematic, you can't connect two ouputs together like with O.C. outputs.
Mona
Attachments
I know this thread has been dormant for a long time, but I've been trying to figure out which elektor article this was from, and couldn't identify it.
Can you shed some light on this please?
Which year and month of publication?