Could this be the control circuit IC that you are looking for: LT4320/LT4320-1 - Ideal Diode Bridge Controller - Linear Technology ?
Nice supply Simon!
I have question, there is Zener in your schematic of the regulator. Do you actually a zener there. I've found zeners to be a wonderful source of noise, but maybe you used LEDs?
Other thing is why use 8 diodes for rectification when you can use two? I do that a lot and use soft switching rectifiers with snubbers (.47 - 1uF in series with a 10 ohm resistor).
I'd post a schematic but I can't from phone.
The secondaries are wired antiphase with separate rectifiers because this causes DM noise from the line to cancel. It won't do much for resonances.
I have not used less than 15R to critically damp a resonance here. Now if the capacitor's self-inductance is resonating with a capacitor elsewhere, the 1R might help. This would be at a much higher frequency though.
I have not used less than 15R to critically damp a resonance here. Now if the capacitor's self-inductance is resonating with a capacitor elsewhere, the 1R might help. This would be at a much higher frequency though.
That's a great find. I will need to try it. Its expensive as are the associated FET's but these mistresses of ours are demanding.
OK, let me 'nit-pick' a little bit about this supply. I am not against multiple caps, IF the caps are of the highest quality, but if they are standard parts, why not combine them into a quality (not necessarily ultra) cap like a 5,000 uF Panasonic, and reduce the connections and wiring inductances?
The LT controllers are designed for PoE PD applications. They claim they have a wide frequency response that exceeds the typical mains frequencies. The mains line, however, can carry noise and interferences in a much wider bandwidth. It is not only because of breakdown voltage that a similar solution designed for direct mains connection is still unavailable for the SMPS market. This will be the next frontier for ultimate efficiency of AC/DC supplies. For low noise application it may be worthwhile trying the latest SiC diodes which have no reverse recovery losses. These are typically designed for PFC boost converters and are high voltage and high cost but if you want to go to the expenses of a fully synchronous bridge they may ultimately result simpler and probably even better.
Just my 0.02$
Giorgio
Just my 0.02$
Giorgio
Most computer power supplies, advanced motor drives, etc. use power factor control input circuits, and achieve quite high efficiency.
I would think these the extreme of fast switching and exactly the opposite of soft diodes.
Last edited:
Thanks for the suggestion. They are really interesting. However they are better suited to high voltage supplies. The forward drop is high and the current ratings are not high enough for a solid state power amp. They would be ideal for a tube amp.
rayma,
I'm quite familiar with the subject. Available ICs enable power supplies in the KW range to be 95% - 96% efficient. This is commonly achieved with an interleaved PFC rectifier followed by a resonant LLC stage for isolation and regulation. To go even further the bridgeless PFC topology is available but still difficult to implement because of the lack of dedicated ICs. In a modern SMPS the high voltage rectifier is really the hottest part.
Demian,
The LT controllers are designed for a power range of 13W to 90W and possibly a little more. MOSFETs are driven with charge pumps that cannot afford the large capacitance of high power switches for fast turn-on, etc. My suggestion was for the SiC diodes to be used for low level applications. This is the Blowtorch thread after all
I'm not sure I'd really want to use MOSFETs and charge pumps in low noise applications. The charge pump switching frequency can easily feed through the circuit, thus vanishing the purpose of using MOSFETs instead of diodes.
I agree SiC are ideal for tube amps, for solid state power amps they belongs to the PFC stage for which they were originally designed. High current parts will soon be available for the traction market if you want to go the passive way with that bridge (not that I recommend this!)
Giorgio
Giorgio-
You are probably right about the noise from the charge pumps etc. Its still interesting enough to learn more about, maybe build a breadboard to try. Same for the SiC rectifiers. The SiC rectifiers may make sense in a classic full wave rectifier with just two diodes. Makes it easier to accept the price.
The underlying question is how much noise comes from the rectifiers switching and how much from the current changing in the transformer and caps being interrupted. Maybe Ed can get some samples and see how much difference it makes in his instrumented setup.
You are probably right about the noise from the charge pumps etc. Its still interesting enough to learn more about, maybe build a breadboard to try. Same for the SiC rectifiers. The SiC rectifiers may make sense in a classic full wave rectifier with just two diodes. Makes it easier to accept the price.
The underlying question is how much noise comes from the rectifiers switching and how much from the current changing in the transformer and caps being interrupted. Maybe Ed can get some samples and see how much difference it makes in his instrumented setup.
In low power, low noise applications an inductance input rectifier is the way to go. Its PF can be made reasonably high and the slope of the current waveform sufficiently slow at turn on to be as good as it gets without resorting to an active, high frequency switched, solution (boost PFC) with accompanying HF noise issues.
If the output load is approximately constant the output voltage regulation can be quite good. Turn on and turn off can bring some unexpected overvoltage issues that may need to be addressed with an active shunt regulator or a preload, for instance.
Giorgio
If the output load is approximately constant the output voltage regulation can be quite good. Turn on and turn off can bring some unexpected overvoltage issues that may need to be addressed with an active shunt regulator or a preload, for instance.
Giorgio
I modified several commercial components coming from the angle of the non-ideal transformer having secondary current switched off in each cycle; ignoring what the 'quality' of the diodes was - no measurements, but subjectively very worthwhile improvement.
I found the SIC to change the bass character in a negative way especially on low level stuff
Measuring soft fast versus conventional on power amps was quite dramatic. Strange that in some circuits it is audible and others it isn't.- Status
- Not open for further replies.