Another issue with the Rigol DP832 power supply: sense wires carrying 3 amps!

[rs20]

You should open up the 3633A at work and see how much capacitance it has it its output terminals, I'd be very interested to find out! I'm sure it's possible to make a supply with minimal output capacitance, but it's certainly not schoolkid-trivial to make a supply that can instantaneously transform between being constant voltage (i.e. infinite capacitance) to constant current (i.e., infinite inductance). Believe me, I spent a whole 30 minutes in LTSpice once, got nothing but oscillation in one mode or the other (or indeed, both). But like I say, there will be topologies out there that work, but they must be more expensive somehow.

In practice, I rarely find myself (deliberately) using CC mode. There's one exception, which is my massive 6 amp LED, but in that case I keep the CV setting close to the nominal forward voltage of the LED. If I was driving a simple red LED, I'd set the voltage to 5V or so. Setting it to 20V and relying entirely on the CC mode is a perverse thing to do. If you're working with something ridiculously sensitive, plug it in first and then enable the output on the PSU. Oddly I don't have any LEDs with me, it'd be interesting to see just how high I'd have to bring the CV voltage before it started killing LEDs, I'm not entirely convinced it'd do so even at 30V.

I don't think you need to test drive, what more would you learn by having a real DP832 in front of you? You need to think about what settings you'd sensibly use and when/why you'd be relying on CC mode. The energy that'll be delivered to faults in these conditions can be calculated.

[PioB]

Hi,

Ok, I was wrong, thank you very much for the explanation! At 20 Volts, the LEDs burn without fail, at 8 Volts it's better (no statistics there ) According to the service manual, the 3633 has 470 uF and 1 uF on the output.
So I got convinced and got probably the last DP832 from Conrad in Switzerland before the price increase today. (470 -> 700 Fr.  )

You should open up the 3633A at work and see how much capacitance it has it its output terminals, I'd be very interested to find out! I'm sure it's possible to make a supply with minimal output capacitance, but it's certainly not schoolkid-trivial to make a supply that can instantaneously transform between being constant voltage (i.e. infinite capacitance) to constant current (i.e., infinite inductance). Believe me, I spent a whole 30 minutes in LTSpice once, got nothing but oscillation in one mode or the other (or indeed, both). But like I say, there will be topologies out there that work, but they must be more expensive somehow.

[rs20]

Ok, I was wrong, thank you very much for the explanation! At 20 Volts, the LEDs burn without fail, at 8 Volts it's better (no statistics there )

Which power supply was this? What was the current limit you chose?