I played today around with PSU-Designer 2 from www.duncanamps.com in order to see wether it is not possible to get similar good results like with a choke-input -supply without chokes but resistors. Result was that it should be no problem, only difference is that the resistor-input might add a bit more resistance to the power supply, but it should be possible to reduce Z again through the capacitor banks following that. Other difference is that we have only a 6Db-Filter compared tho the 12 Db-Filter with a choke. Besides that all the goodies are there: Reduction of the current spikes and HF-nois by the factor of 30 in my example, possible reduction of oversized transformers to handle those nasty spikes etc.
A few resistors are so much easier to get and much cheaper than the chokes nowadays, I am wondering now why not many more people tried this configuration and report back (hopefully the positive) listening results ?
When I follow the discussion around capacitance multipliers I don't get it as well: I understand that people use them not because they improve sound, but to reduce ripple significantly while not having to buy big caps. On the other handside a passive RCRCRC-Filtering would give you similar ripplerejection without the need to use hundredthousands of uF. In tube amps this is quiet ususal stuff, why not here ? Why building a more complicate setup, knowing that it will downgrade sonics when there is as well a simple passive solution ?
Maybe somone will eplain a bit more on the effects of transient response, but than I would like to understand as well: IN which frequency-range ? Do I need perfect response at 3 Hz ? Or is it more important to have a ripplefree, HF-free supply which has enough beef for the range of 30Hz upwards ? Than the last RC does not have to be very big but should be of high quality, right ?
As I am planning to build different projects I am highly interested in your listening experiences.
Best Regards
Frank
A few resistors are so much easier to get and much cheaper than the chokes nowadays, I am wondering now why not many more people tried this configuration and report back (hopefully the positive) listening results ?
When I follow the discussion around capacitance multipliers I don't get it as well: I understand that people use them not because they improve sound, but to reduce ripple significantly while not having to buy big caps. On the other handside a passive RCRCRC-Filtering would give you similar ripplerejection without the need to use hundredthousands of uF. In tube amps this is quiet ususal stuff, why not here ? Why building a more complicate setup, knowing that it will downgrade sonics when there is as well a simple passive solution ?
Maybe somone will eplain a bit more on the effects of transient response, but than I would like to understand as well: IN which frequency-range ? Do I need perfect response at 3 Hz ? Or is it more important to have a ripplefree, HF-free supply which has enough beef for the range of 30Hz upwards ? Than the last RC does not have to be very big but should be of high quality, right ?
As I am planning to build different projects I am highly interested in your listening experiences.
Best Regards
Frank
Choke input
The reason you do not see choke input filters is the size of choke needed. Using you Ben Duncan program it is easy to see.
Most of the choke with current handling (amps) are mH range chokes. The Duncan program will not let you model them with 10,000 - 40,000 ufd filter caps in the choke input mode.
I tried building a choke input power supply for my buffered linestage. It never acted like a choke input. It worked like a RC filter instead. The dc output for a choke input should be close to the AC. Mine was running 150% of the AC.
Playing around with the Duncan program it took a 500 mH choke to get it to model with the caps I was using. My choke was only 20 mH.
For a power amp where 6 - 10 amp current rating is needed, a 35 mH choke is a serious chuck of iron. Plus the field coming from it will create other problems.
Hope this helps.
George
The reason you do not see choke input filters is the size of choke needed. Using you Ben Duncan program it is easy to see.
Most of the choke with current handling (amps) are mH range chokes. The Duncan program will not let you model them with 10,000 - 40,000 ufd filter caps in the choke input mode.
I tried building a choke input power supply for my buffered linestage. It never acted like a choke input. It worked like a RC filter instead. The dc output for a choke input should be close to the AC. Mine was running 150% of the AC.
Playing around with the Duncan program it took a 500 mH choke to get it to model with the caps I was using. My choke was only 20 mH.
For a power amp where 6 - 10 amp current rating is needed, a 35 mH choke is a serious chuck of iron. Plus the field coming from it will create other problems.
Hope this helps.
George
Energy loss: yes ( I assume you mean the voltage losses across the Rs).
Extra costs ? I am not sure. Most people double or tripple the size of their Xformers to deal with the high current spikes of a pi-filter (CRC; for a 400mA Phono-Amp they can be 15A) . As you don't have these spikes in a RC or LC-Filter, you can right-size the transfomer. As well a pi-filter produces 1,4 times the voltage, but only 0,6 times the current needed, while a choke-input produces 0,9 times the voltage, but proportionally more current, so the filter as it delivers less voltage does not destroy energy, but transforms it differently. Or in other words, when I meet my voltage-target with a pi-filter I needed in any case to calculate VA with 40% higher as in the choke-input-scenario.
Only exception is the power dissipated in the Rs though...but is there not always a price to pay ? And if I consume extra 30W when heating up an Aleph-Design, well that's only 10% more...but with good HF-rejection like in a choke-Input Design (maybe even better, as resistance is higher and diodes love to work into resistance).
Extra costs ? I am not sure. Most people double or tripple the size of their Xformers to deal with the high current spikes of a pi-filter (CRC; for a 400mA Phono-Amp they can be 15A) . As you don't have these spikes in a RC or LC-Filter, you can right-size the transfomer. As well a pi-filter produces 1,4 times the voltage, but only 0,6 times the current needed, while a choke-input produces 0,9 times the voltage, but proportionally more current, so the filter as it delivers less voltage does not destroy energy, but transforms it differently. Or in other words, when I meet my voltage-target with a pi-filter I needed in any case to calculate VA with 40% higher as in the choke-input-scenario.
Only exception is the power dissipated in the Rs though...but is there not always a price to pay ? And if I consume extra 30W when heating up an Aleph-Design, well that's only 10% more...but with good HF-rejection like in a choke-Input Design (maybe even better, as resistance is higher and diodes love to work into resistance).
resistors use too much power...
inductors are better, and it is not really necessary to make the circuit behave like a "choke input" all the time, only at high DC current.
For a power amp supply i recently designed i used a small inductor (0.5mH) to limit diode current. this is followed by a NFB-regulator (IOW zener referenced darlington emitter follower) because the regulation of the LC supply with small L is not good and the C is not too large either because it defeats the purpse of low diode current.
the reg also has low output impedance (0.12 ohms).
heavy RC filtering works great for preamps etc - low current apps. for pure class A, really constant load power amps it might be ok.. never tried it. also, as NP wrote, very lossy.
inductors are better, and it is not really necessary to make the circuit behave like a "choke input" all the time, only at high DC current.
For a power amp supply i recently designed i used a small inductor (0.5mH) to limit diode current. this is followed by a NFB-regulator (IOW zener referenced darlington emitter follower) because the regulation of the LC supply with small L is not good and the C is not too large either because it defeats the purpse of low diode current.
the reg also has low output impedance (0.12 ohms).
heavy RC filtering works great for preamps etc - low current apps. for pure class A, really constant load power amps it might be ok.. never tried it. also, as NP wrote, very lossy.
A friend of mine has a Carver Signature preamp in the closet as a spare in case his CAT throws a hissy fit. The phono section of the Carver has a loooong string of CRCRCRCRCR...going from the main supply down to the first stage of the phono. Each successive stage in the phono (four or five stages in that thing--way complicated) taps into the next higher rung in the ladder, running higher rail voltages.
Yes, it works. Yes, it's cheap. Yes, it's wasteful. But there's another aspect that hasn't been touched on yet, and that is the fact that there are fashions in the electronics world just as there are in other things. Right now inductors are out of fashion, mainly due to cost and weight. When performance becomes an issue again, the pendulum will swing and they'll be easier to find once more.
Grey
Yes, it works. Yes, it's cheap. Yes, it's wasteful. But there's another aspect that hasn't been touched on yet, and that is the fact that there are fashions in the electronics world just as there are in other things. Right now inductors are out of fashion, mainly due to cost and weight. When performance becomes an issue again, the pendulum will swing and they'll be easier to find once more.
Grey
inductor availability
hello all
Grey, i agree with you.
But Inductor availability isn't bad..
for power amp filters speaker crossover inductors can be used (air-core) and these are available in a number of values and sizes.
For low current applications (and low voltage) also a number of PCB-mount inductors are available, my supplier's catalog lists a 36mH choke for 36cents, good for 50mA. With a string of these and some big caps you can make a mighty filter...
My old preamp used RC filtering too, i think 6 stages.. needless to say the supply was very quiet.
really big inductors are also available for tube use (in the tube world inductors don't go out of fashion so easily) so you could of course build a preamp filter with 10H chokes (be ready to pay for it), i think it would be appropriate to connect a smaller inductor in series with these because of the rather large capacitance of the big chokes.
regards
k
hello all
Grey, i agree with you.
But Inductor availability isn't bad..
for power amp filters speaker crossover inductors can be used (air-core) and these are available in a number of values and sizes.
For low current applications (and low voltage) also a number of PCB-mount inductors are available, my supplier's catalog lists a 36mH choke for 36cents, good for 50mA. With a string of these and some big caps you can make a mighty filter...
My old preamp used RC filtering too, i think 6 stages.. needless to say the supply was very quiet.
really big inductors are also available for tube use (in the tube world inductors don't go out of fashion so easily) so you could of course build a preamp filter with 10H chokes (be ready to pay for it), i think it would be appropriate to connect a smaller inductor in series with these because of the rather large capacitance of the big chokes.
regards
k
Choke inputs do a lot of good things.
If the choke is big enough than you can expect.( No particular order here just what pops into my head)
Perfect for class A constant current draw. This reduces the size of the choke needed when compared to AB types.
Inrush current reducton. Yank out those skinny wired themisters.
Big reduction in ripple. You actually get a sinwave (100 or 120HzUS) and not the usual saw-tooth tiger. Which is actual noise when you think about it.
You will not know your amp is turned. That's how quite it can get.
Lower capacitor input ripple. Which may increase capacitor life due to lower ripple current across it. Can you hear your caps say thank you?
Reduction of peak diode and transformer current. They will thank you too. Lowered current also results in lowered transformer losses.
You loose voltage, but current is near the full theoretical limit of the AC/VA rating of the power transformer.
Cooler diode and transformer temperatures. Thanks again!
Acts like a regulator. Helping to keep the voltage constant.
Acts as an energy storage device.
Finally it just sounds better. Smooth and musical would be the effect you get. Your basic capacitor input sounds nervous and hyper.
The reason it is not used is increased cost and size. But that is no deterrent for nutty DIY types.
If I missed any please chime in.
If the choke is big enough than you can expect.( No particular order here just what pops into my head)
Perfect for class A constant current draw. This reduces the size of the choke needed when compared to AB types.
Inrush current reducton. Yank out those skinny wired themisters.
Big reduction in ripple. You actually get a sinwave (100 or 120HzUS) and not the usual saw-tooth tiger. Which is actual noise when you think about it.
You will not know your amp is turned. That's how quite it can get.
Lower capacitor input ripple. Which may increase capacitor life due to lower ripple current across it. Can you hear your caps say thank you?
Reduction of peak diode and transformer current. They will thank you too. Lowered current also results in lowered transformer losses.
You loose voltage, but current is near the full theoretical limit of the AC/VA rating of the power transformer.
Cooler diode and transformer temperatures. Thanks again!
Acts like a regulator. Helping to keep the voltage constant.
Acts as an energy storage device.
Finally it just sounds better. Smooth and musical would be the effect you get. Your basic capacitor input sounds nervous and hyper.
The reason it is not used is increased cost and size. But that is no deterrent for nutty DIY types.
If I missed any please chime in.
Also, you may be able to use a smaller transformer and less caps and still get the performance your looking for. This is because of the superior, voltage regulation, ripple filtering and energy storage of the choke. The expense of a choke input power supply can come down. But, we are still stuck with the large size to get all the benefits.
Paul, I totally agree. I started this threat though to discuss the experiences whcih may or may not be out there to play with resistor-input instead of choke-input and what this gives us. I gues one clear difference is that resistor don't store energy, but destroy engergy, but what does this mean soundwise ?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.