To bleed or not to bleed, that is the question.
The point of bleeders is 2 fold; first they help regulation by providing a minimum current set point. Second and more important is in the case of failure of the rail fuses, you will have them right? Where does the stored energy go if you don't have the bleeders? After the failure will be the time you are poking around in the amp with a probe and will inadvertently short out one of the caps. BANG! There goes your probe tip and now a trip down the hall to change your shorts!
The stored charge can last a very long time and if the dead amp has been sitting for some time we tend to discount thoughts of danger like this. Safety first!
Roger
The point of bleeders is 2 fold; first they help regulation by providing a minimum current set point. Second and more important is in the case of failure of the rail fuses, you will have them right? Where does the stored energy go if you don't have the bleeders? After the failure will be the time you are poking around in the amp with a probe and will inadvertently short out one of the caps. BANG! There goes your probe tip and now a trip down the hall to change your shorts!
The stored charge can last a very long time and if the dead amp has been sitting for some time we tend to discount thoughts of danger like this. Safety first!
Roger
Latest drawing
Back2diy,
The new drawing looks very good but please do include rail fuses. This setup still is not as good as separate transformers or even separate windings. The separated grounds are really an illusion as when the diodes are forward biased they are really connected together. Signal current will transfer between channels freely at this time. I did a design like this some time ago and had the transformer manufacture do separate windings so I could do full bridges for each needed voltage. At the same time you can get the exact voltage and power rating you need. Most of these transformers are done by hand one at a time so customs aren’t much of a problem or very much more costly. All you need to do is ask!
Roger
Back2diy,
The new drawing looks very good but please do include rail fuses. This setup still is not as good as separate transformers or even separate windings. The separated grounds are really an illusion as when the diodes are forward biased they are really connected together. Signal current will transfer between channels freely at this time. I did a design like this some time ago and had the transformer manufacture do separate windings so I could do full bridges for each needed voltage. At the same time you can get the exact voltage and power rating you need. Most of these transformers are done by hand one at a time so customs aren’t much of a problem or very much more costly. All you need to do is ask!
Roger
richie00boy, I will keep C2, C3 close to the rectifier bridges. Yes, I am keeping the diode caps only for curiosity 
Initial build will be without them, then I plan to add them to see if there is any sonic difference to my ears and I'll take them out if I can't find any. On the bleeders, I had planned on soldering one to each cap (across the terminals so to speak) - your option would mean I put them in the circuit potentially after the last caps. I guess that is ok (from a safety point - will it drain both caps if resistor rating is calculated correctly)? For the initial stages of my diy steps, I don't plan on using more than 300VA transformer - from what I have read, a soft start is unnecessary in this case? I do plan to use one when I move to the mono amp (which will most likely include a more powerful amp kit also - UcD400 for e.g.).
Bert and all, the reason I went for bleeders is mainly due to the safety factor. I plan to keep them that way.
Roger, on the rail fuses - from what I have learnt, they need to be fast blow type. What I couldn't figure out is the amp rating on those fuses - how do I calculate them (for calculating transformer primary fuse value I divide the VA rating of the transformer secondary by input voltage and add 10% for transformer loss, etc: e.g. - 300VA / 120V = 2.5A + 10% ~ 3A slow blow type fuse for F1 in previous diagram). Is it the same as the primary fuse except being a fast blow type?
[Edited to add comment below]:
Roger/Bert I understand your comments about separate windings/transformers - I do plan to get there in later stages of my diy steps. The intermediate step I had outlined is to educate myself on calculating the right values for the transformer, capacitors, bleeder resistors, etc., before I can take on the challenge of designing a psu for a mono amp with the various add-on circuits mentioned (previously).
Thanks again!
-bck2diy
Initial build will be without them, then I plan to add them to see if there is any sonic difference to my ears and I'll take them out if I can't find any. On the bleeders, I had planned on soldering one to each cap (across the terminals so to speak) - your option would mean I put them in the circuit potentially after the last caps. I guess that is ok (from a safety point - will it drain both caps if resistor rating is calculated correctly)? For the initial stages of my diy steps, I don't plan on using more than 300VA transformer - from what I have read, a soft start is unnecessary in this case? I do plan to use one when I move to the mono amp (which will most likely include a more powerful amp kit also - UcD400 for e.g.).Bert and all, the reason I went for bleeders is mainly due to the safety factor. I plan to keep them that way.
Roger, on the rail fuses - from what I have learnt, they need to be fast blow type. What I couldn't figure out is the amp rating on those fuses - how do I calculate them (for calculating transformer primary fuse value I divide the VA rating of the transformer secondary by input voltage and add 10% for transformer loss, etc: e.g. - 300VA / 120V = 2.5A + 10% ~ 3A slow blow type fuse for F1 in previous diagram). Is it the same as the primary fuse except being a fast blow type?
[Edited to add comment below]:
Roger/Bert I understand your comments about separate windings/transformers - I do plan to get there in later stages of my diy steps. The intermediate step I had outlined is to educate myself on calculating the right values for the transformer, capacitors, bleeder resistors, etc., before I can take on the challenge of designing a psu for a mono amp with the various add-on circuits mentioned (previously).
Thanks again!
-bck2diy
-bck2diy,
primary fuse a 3.15AT fuse(300W toroid) and dc line fuses for the UCD400 is 8AF per supply line. UCD180= 6.3AF. So 2 modules 4 fuses. This is how I use it with a fast type of output protection relay. You don't need a softstart for a 300W toroid with a 230V AC mains. At 115AC I don't know? Bleeders for safety are wise as stated by Roger.
primary fuse a 3.15AT fuse(300W toroid) and dc line fuses for the UCD400 is 8AF per supply line. UCD180= 6.3AF. So 2 modules 4 fuses. This is how I use it with a fast type of output protection relay. You don't need a softstart for a 300W toroid with a 230V AC mains. At 115AC I don't know? Bleeders for safety are wise as stated by Roger.
Lots of questions
The bleeders should not be mounted on the caps. If you use a low enough resistance value to drain the caps quickly they will run quite warm or even hot enough that you wouldn't want the cap to be the heatsink. Only one on each rail is necessary.
400VA seems to be the dividing line according to most and me as well. Above this number you would be better served with a soft start. Inrush current can get very high with the bigger transformers. I would guess in the 50-100 amp range. This is enough to stress parts and cause premature failures.
Rail fuses? 180 watts/4 = 45, SQ root= 6.7 amps total shared by 2 rails + some margin = 5 amps. This should be a workable value for the 180 modules. This is also the value Hypex recommends.
Roger
The bleeders should not be mounted on the caps. If you use a low enough resistance value to drain the caps quickly they will run quite warm or even hot enough that you wouldn't want the cap to be the heatsink. Only one on each rail is necessary.
400VA seems to be the dividing line according to most and me as well. Above this number you would be better served with a soft start. Inrush current can get very high with the bigger transformers. I would guess in the 50-100 amp range. This is enough to stress parts and cause premature failures.
Rail fuses? 180 watts/4 = 45, SQ root= 6.7 amps total shared by 2 rails + some margin = 5 amps. This should be a workable value for the 180 modules. This is also the value Hypex recommends.
Roger
richie00boy, Bert, Roger ... thanks once again!!
Roger, I'll take up your advice on not mounting the resistor across the capacitor terminals. I didn't think of the heat issues - I assumed a heatsinked wirewound resistor (as shown in the picture below) would take care of the heat dissipation issues! Also, thanks for showing an example of how to calculate the rails fuse.
Now, I ought to be able to complete step 2 without any further questions! I'll be back to ask more questions when I take on soft start, dc protection, etc!!
-bck2diy
Roger, I'll take up your advice on not mounting the resistor across the capacitor terminals. I didn't think of the heat issues - I assumed a heatsinked wirewound resistor (as shown in the picture below) would take care of the heat dissipation issues! Also, thanks for showing an example of how to calculate the rails fuse.
Now, I ought to be able to complete step 2 without any further questions! I'll be back to ask more questions when I take on soft start, dc protection, etc!!
-bck2diy
Attachments
Heat removal
Of course if you are remote mounting those resistors to a panel or something they could be connected to the cap as not much heat would go through the wires. I assumed you were talking about the typical sand cast or axial types mounted with short leads right on the cap pins.
I do hope that isn’t the actual resistor you are thinking of using. 100 ohm is way too low of value and will waste around 20 watts each. A value like 2000 2 watt would be far more appropriate using only around 1 watt. Anything from 2k to 10k will do fine with out generating excessive heat.
Roger
Of course if you are remote mounting those resistors to a panel or something they could be connected to the cap as not much heat would go through the wires. I assumed you were talking about the typical sand cast or axial types mounted with short leads right on the cap pins.
I do hope that isn’t the actual resistor you are thinking of using. 100 ohm is way too low of value and will waste around 20 watts each. A value like 2000 2 watt would be far more appropriate using only around 1 watt. Anything from 2k to 10k will do fine with out generating excessive heat.
Roger
Sorry for the confusion I calculated for a scenario of about 60V DC between one of the rails and GND:
1. Using a 10k resistor to bleed, say, 2x10,000uF capacitors would require the resistor to be rated atleast 1/2W (double that to 1W) and it would take about 35 minutes
2. Using a 1k resistor for the same would require the resistor to be atleast 4W (double - 8W) but the time taken would drop to about 4 minutes
I am going for a longer discharge time so my final choice would be closer to option 1 (use a 2W or higher rated resistor as you suggested).
-bck2diy
I calculated for a scenario of about 60V DC between one of the rails and GND:1. Using a 10k resistor to bleed, say, 2x10,000uF capacitors would require the resistor to be rated atleast 1/2W (double that to 1W) and it would take about 35 minutes
2. Using a 1k resistor for the same would require the resistor to be atleast 4W (double - 8W) but the time taken would drop to about 4 minutes
I am going for a longer discharge time so my final choice would be closer to option 1 (use a 2W or higher rated resistor as you suggested).
-bck2diy
Discharging caps
Discharging to safe levels and not to zero is what you are after. The times you stated are much longer than to a safe level. Even 6 time constants should be conservative and would give a time value 1/2 what you called out. This exercise does give one a good feel of just how much energy is stored in these caps. Certainly not something you want to dump across your test probe!
Roger
bck2diy said:Sorry for the confusion
1. Using a 10k resistor to bleed, say, 2x10,000uF capacitors would require the resistor to be rated atleast 1/2W (double that to 1W) and it would take about 35 minutes
2. Using a 1k resistor for the same would require the resistor to be atleast 4W (double - 8W) but the time taken would drop to about 4 minutes
I am going for a longer discharge time so my final choice would be closer to option 1 (use a 2W or higher rated resistor as you suggested).
-bck2diy
Discharging to safe levels and not to zero is what you are after. The times you stated are much longer than to a safe level. Even 6 time constants should be conservative and would give a time value 1/2 what you called out. This exercise does give one a good feel of just how much energy is stored in these caps. Certainly not something you want to dump across your test probe!
Roger
synchronicity
strangely, i am in the exact same boat.
i have a circuit quite similiar to your last one, although
i am using some 42000uF 90V caps i have lying around.
for what it is worth, i am slowly grinding my way through
making a small PC board for everything other than the big
caps (standard "use anything from the Mouser catalog" project).
although i am unsure when they would be ready
(i intend to use a commercial pc fab), i'd be happy
to share them with you.
andrew
strangely, i am in the exact same boat.
i have a circuit quite similiar to your last one, although
i am using some 42000uF 90V caps i have lying around.
for what it is worth, i am slowly grinding my way through
making a small PC board for everything other than the big
caps (standard "use anything from the Mouser catalog" project).
although i am unsure when they would be ready
(i intend to use a commercial pc fab), i'd be happy
to share them with you.
andrew
bck2diy said:
Just to show you the wiring layout of a 1 transformer stereo amp. with superb soundstage. I use this for the UCD amps. and the Coldamps.
http://www.grotel.nl/photo/wiring scheme class d stereo.jpg
This grounding scheme did not work for me, was to much humm and a lot of crosstalk. Touching the input cable changed my soundstage completely.(UCD400's had increased gain)
http://www.grotel.nl/photo/wiring scheme class d stereo2.jpg
Just waiting for Coldamps new SMPS(model30) when finished and to try them.
http://www.grotel.nl/photo/wiring scheme class d stereo2.jpg
Just waiting for Coldamps new SMPS(model30) when finished and to try them.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.