Hi !
I've got another quick question for you 🙂
According to the power supply schematic posted here ,
do I need to connect the power supply ground to the chassis ?
(And the chassis to earth ?)
I don't have any real knowledge about safety and standards about isolation,
enclosure, grounding etc.. (I know however that it's not recommended to touch two wires
with both hands 🙂 Still I'm quite afraid to work on my Tube preamp with those big 220uF, 400V capacitors still full)
Quote from Boxbldr681
Thanks for your help !
I've got another quick question for you 🙂
According to the power supply schematic posted here ,
do I need to connect the power supply ground to the chassis ?
(And the chassis to earth ?)
I don't have any real knowledge about safety and standards about isolation,
enclosure, grounding etc.. (I know however that it's not recommended to touch two wires
with both hands 🙂 Still I'm quite afraid to work on my Tube preamp with those big 220uF, 400V capacitors still full)
Quote from Boxbldr681
Thanks for your help !
just keep the 'Hmmmm's to a minimum
Minimal "hmmmm" absolutey requires correct grounding. That was humor for those who fear that I am becoming an academic.
For the low noise and high speed soft recovery diode camp:
http://www.gensemi.com/appnotespdf/quik108.pdf
http://www.hagtech.com/pdf/snubber.pdf
Bungee jumping off the rectifier bridge,
Fred
Minimal "hmmmm" absolutey requires correct grounding. That was humor for those who fear that I am becoming an academic.
For the low noise and high speed soft recovery diode camp:
http://www.gensemi.com/appnotespdf/quik108.pdf
http://www.hagtech.com/pdf/snubber.pdf
Bungee jumping off the rectifier bridge,
Fred
Elkaid said:
I don't have chassis grounding in mine, and I don't have any hum. I do not use AC grounding either.
/Jan
Fred, thanks for that explanation, it's made things a little clearer for me.
I do however have another thought, please correct me if I'm wrong 😱
with a single bridge & twin secondaries the rails are supplied by both secondaries, swinging above and below the centre tap.
This means that the 0v centre-tap may not be quite in the mid-point of the two rails , depending on the accuracy of the windings and the resistance of the secondaries.
It also seems to imply that the 0v is only really held in place by the impedance/resistance of the trans secondaries, and a heavy load on either rail will drag the 0v point in that direction.
The upshot would be one of the rails exceeding its voltage somewhat:
e.g. 70-0-70 supply , the -ve rail takes a large load ( for whatever reason, perhaps an output device failure ) and drops to (say) -50v. The 0v reference swings -ve and the +ve rail now starts to show perhaps +85v or so.... ( pop go the psu caps which were only rated at 80v.....)
With dual bridges ( or the discrete equivalent ) , the rails are seperately supplied, there is no floating of the 0v point ( which itself would induce hum if it happened... ) , the rails are independent such that a large load on one rail has no effect on the other one.
now I realise that in a good amp the load on each rail should be well balanced, but in practice they are not since large transients will strain each rail and insufficient psu capacitance would cause an unevel load on each half cycle.
Am I wide of the mark here, or am I making sense?
🙂
ray
I do however have another thought, please correct me if I'm wrong 😱
with a single bridge & twin secondaries the rails are supplied by both secondaries, swinging above and below the centre tap.
This means that the 0v centre-tap may not be quite in the mid-point of the two rails , depending on the accuracy of the windings and the resistance of the secondaries.
It also seems to imply that the 0v is only really held in place by the impedance/resistance of the trans secondaries, and a heavy load on either rail will drag the 0v point in that direction.
The upshot would be one of the rails exceeding its voltage somewhat:
e.g. 70-0-70 supply , the -ve rail takes a large load ( for whatever reason, perhaps an output device failure ) and drops to (say) -50v. The 0v reference swings -ve and the +ve rail now starts to show perhaps +85v or so.... ( pop go the psu caps which were only rated at 80v.....)
With dual bridges ( or the discrete equivalent ) , the rails are seperately supplied, there is no floating of the 0v point ( which itself would induce hum if it happened... ) , the rails are independent such that a large load on one rail has no effect on the other one.
now I realise that in a good amp the load on each rail should be well balanced, but in practice they are not since large transients will strain each rail and insufficient psu capacitance would cause an unevel load on each half cycle.
Am I wide of the mark here, or am I making sense?
🙂
ray
TheFettler said:
Hi,
Ground point (0V tap) is not floating, and + rail cannot go up when -rail is loading too much. In single bridge loading secondaries wire warming is equal in this condition, in dual bridge is not.
This means for me also that the accuracy of output voltage is slightly better in single bridge connection (one + for single).
If voltage of secondaries is not equal the better performance have single bridge because both output voltage will be equal with some bigger ripple (one half period is greater then second). In double bridge in case of unequal secondaries, output voltage of + and - rail are different (second + for single).
Regards
Twin Rectifier Bridges
"If voltage of secondaries is not equal the better performance have single bridge because both output voltage will be equal"
There is really no reason that the positive and negative rails need
to be exactly the same voltage that I can think of.
" In single bridge loading secondaries wire warming is equal in this condition, in dual bridge is not."
I have no idea what this means........... in fact I can't follow the reasoning in moamps post. Am I missing something in moamps post that sounds reasonable to someone else?
I still maintain that a dual bridge is superior to a single bridge in every aspect except cost and the slghtly lower DC voltages from the extra diode drops.
Respectfully,
Fred
"If voltage of secondaries is not equal the better performance have single bridge because both output voltage will be equal"
There is really no reason that the positive and negative rails need
to be exactly the same voltage that I can think of.
" In single bridge loading secondaries wire warming is equal in this condition, in dual bridge is not."
I have no idea what this means........... in fact I can't follow the reasoning in moamps post. Am I missing something in moamps post that sounds reasonable to someone else?
I still maintain that a dual bridge is superior to a single bridge in every aspect except cost and the slghtly lower DC voltages from the extra diode drops.
Respectfully,
Fred
I tried dual bridge vs singe bridge, using the same diodes, in my Gainclone (as somebody suggested that he obtained better sound with one bridge). In my setup, dual bridge sounded significantly better. I was using MUR860 Motorola diodes.
Those Motorola diodes seem to be hard to get, so I ordered Fairchild MUR860 from Digi Key. I'm sending them back (100 of them), because again, they don't sound as good as Motorola parts. It seems like OnSemi is division of Motorola and I will try their parts (at least the diodes look exactly the same, the only difference being "ON" marking on a case, instead of "M").
Those Motorola diodes seem to be hard to get, so I ordered Fairchild MUR860 from Digi Key. I'm sending them back (100 of them), because again, they don't sound as good as Motorola parts. It seems like OnSemi is division of Motorola and I will try their parts (at least the diodes look exactly the same, the only difference being "ON" marking on a case, instead of "M").
Re: Twin Rectifier Bridges
Hi,
" In single bridge loading secondaries wire warming is equal in this condition, in dual bridge is not."
In case that output current from + and - rail is not equal; for example 2A from +, and 1A from -
in single bridge BOTH secondaries are loaded with 2A and 1A alternating (changing direction)
in dual bridge ONE of secondaries is loaded with 2A, other with 1A all the time.
"If voltage of secondaries is not equal the better performance have single bridge because both output voltage will be equal"
In case that secondaries voltage are not equal; for example one is 20Vpeak, second is 25Vpeak
in single bridge, + and - caps is loaded with BOTH voltage alternating and resulting voltage is 25-0.7V without load
in double bridge, one cap is loaded at 20-2*0.7V, second at 25-2*0.7V ---- 5V is unbalance
""I still maintain that a dual bridge is superior to a single bridge in every aspect except cost and the slghtly lower DC voltages from the extra diode drops.""
I agree with you, but I am referring to TheFettlers ideas with wich I disagree.
Sorry on my English
Regards
Fred Dieckmann said:
Hi,
" In single bridge loading secondaries wire warming is equal in this condition, in dual bridge is not."
In case that output current from + and - rail is not equal; for example 2A from +, and 1A from -
in single bridge BOTH secondaries are loaded with 2A and 1A alternating (changing direction)
in dual bridge ONE of secondaries is loaded with 2A, other with 1A all the time.
"If voltage of secondaries is not equal the better performance have single bridge because both output voltage will be equal"
In case that secondaries voltage are not equal; for example one is 20Vpeak, second is 25Vpeak
in single bridge, + and - caps is loaded with BOTH voltage alternating and resulting voltage is 25-0.7V without load
in double bridge, one cap is loaded at 20-2*0.7V, second at 25-2*0.7V ---- 5V is unbalance
""I still maintain that a dual bridge is superior to a single bridge in every aspect except cost and the slghtly lower DC voltages from the extra diode drops.""
I agree with you, but I am referring to TheFettlers ideas with wich I disagree.
Sorry on my English
Regards
Hi,
By rectifying the "+" side and connecting this side to ground at the amp of course. How else?
Sayonara
Mad_K said:
By rectifying the "+" side and connecting this side to ground at the amp of course. How else?
Sayonara
Peter Daniel said:
Farnell has Motorola MUR860's. They cost e 72,30 for 100.
Maybe I will try them once...
Fedde
I agree with you, but I am referring to TheFettlers ideas with wich I disagree.
thats fine, having looked a little closer at the schematic I can see that I was wrong with my first thought
however, with one rail loaded heavily there still seems to be an imbalance with one bridge - load down the +ve rail and the -ve rail will start to sag.
with 2 bridges the other rail will keep it's regulation better.
BUT, this implies that with a single bridge BOTH secondaries are trying to preserve the loaded rail , rather than just one for a dual bridge.
what would this mean in real life, I don't know.
😉
ray
thats fine, having looked a little closer at the schematic I can see that I was wrong with my first thought
however, with one rail loaded heavily there still seems to be an imbalance with one bridge - load down the +ve rail and the -ve rail will start to sag.
with 2 bridges the other rail will keep it's regulation better.
BUT, this implies that with a single bridge BOTH secondaries are trying to preserve the loaded rail , rather than just one for a dual bridge.
what would this mean in real life, I don't know.
😉
ray
Hi,
I don't see bigger unbalance in power supply (single or double) in strictly audio apps (amps, preamps, etc). I have problems with unbalance in (for example) power supply for some audio consoles. There currents from rails can be very different.
Regards
I don't see bigger unbalance in power supply (single or double) in strictly audio apps (amps, preamps, etc). I have problems with unbalance in (for example) power supply for some audio consoles. There currents from rails can be very different.
Regards
Hi T.
Excellent idea, thanks for the tip.
Have you tried this 4-transformer 4-double diode rectifier with a Gain-clone or other chip-based amp ? Is there any special recommendation for the transformer construction ? Why only 80VA ?
TIA, v.
Excellent idea, thanks for the tip.
Have you tried this 4-transformer 4-double diode rectifier with a Gain-clone or other chip-based amp ? Is there any special recommendation for the transformer construction ? Why only 80VA ?
TIA, v.
I am fairly interested in interesting things, but don't see something interesting here 😉. Could anybody provide some clarification here. It was my guess that if only one trafo is used for 2 channels, a set of two bridges per channel (4 total) would be helpfull, each channel has it's 'own' filter caps whereas the shared bridge configuration has by definition also shared filter caps. Is this correct?
Hi,
No. Never. This little trick is out of my older high power PA Amplifier designs. It allowed/allowes the use of Diodes otherwise problematic to handle or expensive and ensures a well defined current path. Dual Schottkies are very cheap and readily available and having just one 3-Pin diode appeals to me.
In the old days I had "press-fit" metal Diodes that needed to connect directly into the heatsink and one connection had to be taken from the heatsink. Making several bridges that way gets messy fast. So we used this scheme with more transformer windings and more transformer and less (then) difficult to get and expensive silicon.
Later when I could get big rectifier bridges we found that scheme with a dual winding and a dual diodes with a transformer per voltage sounded notably better and that in a PA Amp!
I could take some guesses here but will refain.
The point about the ground impedances Fred Dieckmann makes is very valid of course, yet this is not all, or going single winding & bridge should have sounded the same.
The individual transformer per voltage may sound better simply because if the signal is asymmetric (which music always is) the less loaded rail will retain a larger voltage, leading to asymmetric clipping with higher peak voltage possible on the unloaded rail for short durations.
There is more, but as said, that would be guesses, yet remember that a transformer transforms into all directions, even between the various secondary windings.
Double C Core with a small airgap and an electrostatic screen.
The 80VA torroids are recommended for ready availability.
A "Humpty" uses a 170VA cut core transformer (fancy name for double C- or R-Core) and two times 80VA = 160VA. You can scale up if you like.
Sayonara
vladimir said:
No. Never. This little trick is out of my older high power PA Amplifier designs. It allowed/allowes the use of Diodes otherwise problematic to handle or expensive and ensures a well defined current path. Dual Schottkies are very cheap and readily available and having just one 3-Pin diode appeals to me.
In the old days I had "press-fit" metal Diodes that needed to connect directly into the heatsink and one connection had to be taken from the heatsink. Making several bridges that way gets messy fast. So we used this scheme with more transformer windings and more transformer and less (then) difficult to get and expensive silicon.
Later when I could get big rectifier bridges we found that scheme with a dual winding and a dual diodes with a transformer per voltage sounded notably better and that in a PA Amp!
I could take some guesses here but will refain.
The point about the ground impedances Fred Dieckmann makes is very valid of course, yet this is not all, or going single winding & bridge should have sounded the same.
The individual transformer per voltage may sound better simply because if the signal is asymmetric (which music always is) the less loaded rail will retain a larger voltage, leading to asymmetric clipping with higher peak voltage possible on the unloaded rail for short durations.
There is more, but as said, that would be guesses, yet remember that a transformer transforms into all directions, even between the various secondary windings.
Double C Core with a small airgap and an electrostatic screen.
The 80VA torroids are recommended for ready availability.
A "Humpty" uses a 170VA cut core transformer (fancy name for double C- or R-Core) and two times 80VA = 160VA. You can scale up if you like.
Sayonara
- Status
- Not open for further replies.