Electronics > Projects, Designs, and Technical Stuff
Evaluation of a 40V/5A DIY lab power supply
Kleinstein:
AT least center taped transformers are easy to get. So switching between 2 taps is also a possibility. However it is tricky in the way it is done in the normal cheap supplies, as the step is quite large.
One can use electronic switch over, a little like a class H audio amplifier - this is faster and thus needs less spare room. So 2 tap electronic switching may be as effective as 3 or 4 taps with relays.
For something like 4 A at 40 V one would definitely like some ways to reduce the heat. For a small 20 V 1 A version one can get away without and do it brute force with just enough heat sink and an over-sized transistor.
I would avoid high side current sensing - the extra amplifier adds delay and makes stability more tricky.
David Hess:
--- Quote from: Kleinstein on August 22, 2019, 08:45:49 pm ---AT least center taped transformers are easy to get. So switching between 2 taps is also a possibility. However it is tricky in the way it is done in the normal cheap supplies, as the step is quite large.
--- End quote ---
A lot of old HP supplies change the rectifier configuration so the high range doubles the output voltage and the low range doubles the output current.
--- Quote ---I would avoid high side current sensing - the extra amplifier adds delay and makes stability more tricky.
--- End quote ---
As I pointed out earlier, the high side sensing and error amplifier can be combined so only one operational amplifier is required. A lot of supplies do this for high side current sensing.
dom0:
--- Quote from: David Hess on August 21, 2019, 02:47:41 pm ---
--- Quote from: Kleinstein on August 21, 2019, 07:49:20 am ---The concept with a transistor as a follower is easy for low voltage, but no longer simple at more than some 30 V, as the error amplifier has to drive such a high voltage.
--- End quote ---
HP was fond of floating the entire control circuit to the power supply's output allowing operation on +/-15 volts or less even on their lower voltage supplies. This presents problems however for a digitally controlled supply unless the whole digital section is floated as well.
--- End quote ---
I think the obverse is true. Floating the control loop is annoying if you want analogue programmability, but is not much of an issue with digital control, since digital signals are easily isolated.
David Hess:
--- Quote from: dom0 on August 24, 2019, 12:19:04 pm ---I think the obverse is true. Floating the control loop is annoying if you want analogue programmability, but is not much of an issue with digital control, since digital signals are easily isolated.
--- End quote ---
It applies to both but few people consider analog programmability today. Digital isolation is definitely easier.
My Tektronix power supplies only allow analog voltage programming so no isolation is required for the current error amplifier which is floating.
TheJC:
Okay, I looked into quite a lot old schematics from Hp and others in the last couple days. And adopted some ideas into my new design.
I referenced everything on the output instead of at the input, which made CC circuitry simpler.
The only problem would be the generation of the +-15V rail. I thought about using a second smaller transformer for this.
Navigation
[0] Message Index
[*] Previous page
Go to full version