I recently got two AMP3 kits and now have a bunch of PSU questions. I have done a search. Surprisingly few AMP3 and 6 threads. Anyway, I will use a linear PSU. I will not use 14.6 V, but 14.2 at most.
1. 41Hz recommends 10,000-22,000 uF for C1 in the schematics. I have asked a similar question before, but probably failed making the question clear. I might fail this time as well. Anyway, when I try this, or similar designs, in PSU Designer (sans bypass caps), I get so lousy results I would never think of using something like it. I get best result when I use something like five 2,200 uF caps and resistors or, better, huge, say, 10H chokes. I'm obviously missing something here. I mean, the SS guys use huge amount of capacitance. So what is going on here? Am I putting to much emphasis on the ripple? Or am I simply missing something, like the effect of those bypass caps?
2. I will use a 50-60VA, 15VAC transformer. I was thinking to fuse the secondary as well as the primary (230VAC). Which values?
3. What Amp rating for the rectifier diodes? Any recommendations? Schottky? Bridge?
4. Any benefits from using Zener diode for D1?
1. 41Hz recommends 10,000-22,000 uF for C1 in the schematics. I have asked a similar question before, but probably failed making the question clear. I might fail this time as well. Anyway, when I try this, or similar designs, in PSU Designer (sans bypass caps), I get so lousy results I would never think of using something like it. I get best result when I use something like five 2,200 uF caps and resistors or, better, huge, say, 10H chokes. I'm obviously missing something here. I mean, the SS guys use huge amount of capacitance. So what is going on here? Am I putting to much emphasis on the ripple? Or am I simply missing something, like the effect of those bypass caps?
2. I will use a 50-60VA, 15VAC transformer. I was thinking to fuse the secondary as well as the primary (230VAC). Which values?
3. What Amp rating for the rectifier diodes? Any recommendations? Schottky? Bridge?
4. Any benefits from using Zener diode for D1?
Attachments
The schematic you posted is very similar to Jan's onboard regulator for the AMP6. No surprise there.
I don't think you need to fuse the secondary, I never do.
10,000uF for C1 works well. There will be some ripple, but the regulator kills almost all of it.
A 3 amp bridge will work well. A bridge has the advantage of matched diodes. I can't see the need for Schottky here, the amp switching noise would drown out any noise advantage they "might" have.
I don't use D2, don't see the need. Zeners can be noisy, but see above.
2200uF works well for C2. You can use more if you want. A big inductor in that spot isn't likely to help much and could make the circuit unstable, I don't know. An inductor after C1 and before the regulator might be worth a try. You don't need a big one.
I don't think you need to fuse the secondary, I never do.
10,000uF for C1 works well. There will be some ripple, but the regulator kills almost all of it.
A 3 amp bridge will work well. A bridge has the advantage of matched diodes. I can't see the need for Schottky here, the amp switching noise would drown out any noise advantage they "might" have.
I don't use D2, don't see the need. Zeners can be noisy, but see above.
2200uF works well for C2. You can use more if you want. A big inductor in that spot isn't likely to help much and could make the circuit unstable, I don't know. An inductor after C1 and before the regulator might be worth a try. You don't need a big one.
some suggestions
1) " Anyway, when I try this, or similar designs, in PSU Designer (sans bypass caps), I get so lousy results I would never think of using something like it. ..." The "bypass capcitors" are all pretty important to a well regulated supply. (I assume here you mean C3 & C4 in the diagram.) Without these in place, the linear regulator LMS1585 will not have much of an impedence load to "work against". Heaven forbid if you actually mean C1 & C2 as these are the caps that do most of the work of filtering off the 60/120 Htz chop from the D1 diode bridge. The 10,000 uF cap is most important, if this is for audio. Leave it in circuit or make it bigger. The "snubbing" cap (C2, 0.1 uF.) should be "plastic", MKT, Tantalum or "poly". This is where the lively "sweetening" action begins ... C4 cleans up the leading edge of the stuff coming off of the diode bridge and scraps off some of the HF noise that bleeds through the regulator.
2) A 15 VAC secondary of the transformer means that the potential DC voltage ahead of the regulator could spike as high as 25 Volts ... so use 35 or 50 volt rated capacitors at C1 & C2. In the event of a regulator failure, to be safe use 35 to 50 volt rated downstream as well for C3 & C4.
No fuse in the secondary is required, as long as the fuse exists in the primary circuit.
3 & 4) Bridge diodes should be >> than 100 volts (D1). Cheap insurance. The use of Schottky diodes in a linear supply is not necessary, but you can if you want = in parallel with each of the diodes in the bridge ... But using a large thermal area diode instead is just as good (some would say better) without increasing the parts count.
Use 1N4004 or equal in the diode bridge (D1a, D1b, D1c, D1d) or get a "diode bridge in a block" (all 4 in one solid state unit). The 1N4004 are quite common and not expensive and rated above several hundred volts ... and will switch faster ('cause of the larger amount of diode thermal area) than the cheaper 1N4002 or '03 or equal. (I once used 200 volt zeniers in a diode bridge, 'cause I had 'em to spare, but a Zenier is capable of leakage and noise up in the HF range, so for audio purposes they are a no, no in the full or half wave bridge.)
Zenier diode at D1?? Don't do this.
panomaniac:
" ... I don't use D2, don't see the need. Zeners can be noisy, but see above. ..."
I agree, the Zenier is unnecessary and could be a source of noise, audio and otherwise...."
" ... A big inductor in that spot (around C2) isn't likely to help much and could make the circuit unstable, I don't know. ..."
Absolutely correct ... a better place for any inductor is downstream from the rectifier, but chosing the wrong one could set up an oscillation in the load further downstream ...
Important: using a 230 VAC primary instead of a 120 VAC primary will effectively double the secondary winding output ... so you would have to pick a secondary and primary that matches. Keep in mind that the LM series voltage regulator has an Absolute upper limit (usually 28 VDC) at the input pin ... so:
use: 230 X 15 or 125 X 15 but not 230 X 30 ... (30 VAC times 1.414 is about double what the LM regulator can handle.)
1) " Anyway, when I try this, or similar designs, in PSU Designer (sans bypass caps), I get so lousy results I would never think of using something like it. ..." The "bypass capcitors" are all pretty important to a well regulated supply. (I assume here you mean C3 & C4 in the diagram.) Without these in place, the linear regulator LMS1585 will not have much of an impedence load to "work against". Heaven forbid if you actually mean C1 & C2 as these are the caps that do most of the work of filtering off the 60/120 Htz chop from the D1 diode bridge. The 10,000 uF cap is most important, if this is for audio. Leave it in circuit or make it bigger. The "snubbing" cap (C2, 0.1 uF.) should be "plastic", MKT, Tantalum or "poly". This is where the lively "sweetening" action begins ... C4 cleans up the leading edge of the stuff coming off of the diode bridge and scraps off some of the HF noise that bleeds through the regulator.
2) A 15 VAC secondary of the transformer means that the potential DC voltage ahead of the regulator could spike as high as 25 Volts ... so use 35 or 50 volt rated capacitors at C1 & C2. In the event of a regulator failure, to be safe use 35 to 50 volt rated downstream as well for C3 & C4.
No fuse in the secondary is required, as long as the fuse exists in the primary circuit.
3 & 4) Bridge diodes should be >> than 100 volts (D1). Cheap insurance. The use of Schottky diodes in a linear supply is not necessary, but you can if you want = in parallel with each of the diodes in the bridge ... But using a large thermal area diode instead is just as good (some would say better) without increasing the parts count.
Use 1N4004 or equal in the diode bridge (D1a, D1b, D1c, D1d) or get a "diode bridge in a block" (all 4 in one solid state unit). The 1N4004 are quite common and not expensive and rated above several hundred volts ... and will switch faster ('cause of the larger amount of diode thermal area) than the cheaper 1N4002 or '03 or equal. (I once used 200 volt zeniers in a diode bridge, 'cause I had 'em to spare, but a Zenier is capable of leakage and noise up in the HF range, so for audio purposes they are a no, no in the full or half wave bridge.)
Zenier diode at D1?? Don't do this.
panomaniac:
" ... I don't use D2, don't see the need. Zeners can be noisy, but see above. ..."
I agree, the Zenier is unnecessary and could be a source of noise, audio and otherwise...."
" ... A big inductor in that spot (around C2) isn't likely to help much and could make the circuit unstable, I don't know. ..."
Absolutely correct ... a better place for any inductor is downstream from the rectifier, but chosing the wrong one could set up an oscillation in the load further downstream ...
Important: using a 230 VAC primary instead of a 120 VAC primary will effectively double the secondary winding output ... so you would have to pick a secondary and primary that matches. Keep in mind that the LM series voltage regulator has an Absolute upper limit (usually 28 VDC) at the input pin ... so:
use: 230 X 15 or 125 X 15 but not 230 X 30 ... (30 VAC times 1.414 is about double what the LM regulator can handle.)
fuses? we don' need no stinking ...
" ... As for fuse on secondary, it's recommended by one transformer manufacturer. I don't think I have seen many fuses on the secondary outside of commercial ss amps. ..."
The transformer makers would certainly recommend a fuse or circuit breaker somewhere = law suits without this recommendation.
One fuse or breaker is enough and the primary side is better from an audio perspective (keeps those noisy fuses further away from the LM regulator = isolated from the secondary) and engineering point of view = easier, cheaper, no PCB board real estate needed ...
" ... As for fuse on secondary, it's recommended by one transformer manufacturer. I don't think I have seen many fuses on the secondary outside of commercial ss amps. ..."
The transformer makers would certainly recommend a fuse or circuit breaker somewhere = law suits without this recommendation.
One fuse or breaker is enough and the primary side is better from an audio perspective (keeps those noisy fuses further away from the LM regulator = isolated from the secondary) and engineering point of view = easier, cheaper, no PCB board real estate needed ...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.