I have a china/ebay 5V linear power supply that I am re-capping. It looks like a shunt regulator circuit. The smaller 100uf/470uf appear to be helper capacitors for the pre-driver circuit. I can see four zener diodes in each power supply (it's a dual power supply board).
The DC voltage is generated by two IRFZ24N N-Channel MOSFETS wired in parallel. The output pins connect to 0.43 ohm resistors (white blocks in picture), which then both connect to +VOUT. Honestly, this almost looks like an amplifier board (which it is in a way).
However, there is a really big 6000uf capacitor as the output capacitor directly on the VOUT leg (as pointed out in the picture below - sorry for the bad quality). When I removed this, the board is marked for a 4700uf. This seems to me like an abnormally large value to use for a regulator output capacitor. I have a 2700uf Nichicon PM on order for this position (I love PM caps in digital power supplies), but now I'm thinking that I should really put a small cap here, such as a 330uf. I have been told that a really large output capacitor on a regulator could slow the regulator down in some situations.
Does anyone have any thoughts on this decision? This will be powering a digital S/PDIF output board.
Ack, it looks like diyaudio is no longer supporting hosting of inserted pictures, and I don't have a website where I can place this, so you'll have to just look at the original ebay pictures:
HIFI-Linear-power-supply-two-Way-LPS-50W-50W-DC-5V
The DC voltage is generated by two IRFZ24N N-Channel MOSFETS wired in parallel. The output pins connect to 0.43 ohm resistors (white blocks in picture), which then both connect to +VOUT. Honestly, this almost looks like an amplifier board (which it is in a way).
However, there is a really big 6000uf capacitor as the output capacitor directly on the VOUT leg (as pointed out in the picture below - sorry for the bad quality). When I removed this, the board is marked for a 4700uf. This seems to me like an abnormally large value to use for a regulator output capacitor. I have a 2700uf Nichicon PM on order for this position (I love PM caps in digital power supplies), but now I'm thinking that I should really put a small cap here, such as a 330uf. I have been told that a really large output capacitor on a regulator could slow the regulator down in some situations.
Does anyone have any thoughts on this decision? This will be powering a digital S/PDIF output board.
Ack, it looks like diyaudio is no longer supporting hosting of inserted pictures, and I don't have a website where I can place this, so you'll have to just look at the original ebay pictures:
HIFI-Linear-power-supply-two-Way-LPS-50W-50W-DC-5V
The large output capacitor could be necessary to keep the regulator circuit stable.
Test carefully if you decrease its value.
Test carefully if you decrease its value.
Something I was just thinking about. This power supply is designed to be a "generic target" to support anything from 5V all the way up to 24V. Could this 6000uf capacitor be that large to help provide enough current for 24V 1A? For 5V current, we probably don't need such a large capacitor. Am I thinking this right? Or am I totally off base?
The regulator output capacitor does not supply average current to the load. Its main purpose
is to provide a dominant pole in the feedback system to ensure stability. Since a high current
regulator will have a different circuit from a low current regulator, their minimum-value-for-stability
output capacitors will usually differ in value.
is to provide a dominant pole in the feedback system to ensure stability. Since a high current
regulator will have a different circuit from a low current regulator, their minimum-value-for-stability
output capacitors will usually differ in value.
Last edited:
Probably it's a standard series pass regulator, but the output capacitor concerns are similar in both.