Hi,
connect the load to D & 4, to minimse the inductance of the cabling.
Twist the cables from transformer to rectifier.
Twist the cables ab with 12
Twist the cables from smoothing caps to load.
Is the load an amplifier?
Then run the power supply traces right next to each other as much as possible and keep the power ground trace as near as possible.
Look at
http://www.diyaudio.com/forums/showthread.php?postid=1890160#post1890160
connect the load to D & 4, to minimse the inductance of the cabling.
Twist the cables from transformer to rectifier.
Twist the cables ab with 12
Twist the cables from smoothing caps to load.
Is the load an amplifier?
Then run the power supply traces right next to each other as much as possible and keep the power ground trace as near as possible.
Look at
http://www.diyaudio.com/forums/showthread.php?postid=1890160#post1890160
Is the additional impedance from D to E and 4 to 5 significant? And if so, would it not improve the filtering, when it is before the last pair of capacitors and therefore be a good thing? Or if it is bad, would the consequence be to connect the load right in the middle between C - D and 3 - 4? And should A and 1 also be connected to that same point?AndrewT said:connect the load to D & 4, to minimse the inductance of the cabling.
According to capacitor manufacturers the right way to do it is disconnect 1 from 2, connect 1 to 5 and connect the load between E and 2, provided all capacitors are the same size.
my reasoning follows along the rCrC philosophy.
the first C in the group is the first C in the rCrC.
This is then followed by the longest (=highest impedance) route to the output.
The other three have two equal length wires and one zero length wires. This combination have the lowest impedance connections possible if one does not stack the caps on top of each other.
The r in this arrangement is tiny but it must have some effect and it cannot do any harm.
A reminder of what I said before. The first Cap in an RCRC or LCLC smoothing arrangement suffers high ripple.
the first C in the group is the first C in the rCrC.
This is then followed by the longest (=highest impedance) route to the output.
The other three have two equal length wires and one zero length wires. This combination have the lowest impedance connections possible if one does not stack the caps on top of each other.
The r in this arrangement is tiny but it must have some effect and it cannot do any harm.
A reminder of what I said before. The first Cap in an RCRC or LCLC smoothing arrangement suffers high ripple.
The question is, whether the r between the capacitors is significant. There were no values in the first post. The r could be so small that the set-up should be treated as parallel configuration of capacitors, i. e. rC instead of rCrC.
And should therefore have low ESR, the lowest possible voltage rating and the biggest possible physical size. Plus sufficient airflow around it, which many people also like to forget.AndrewT said:The first Cap in an RCRC or LCLC smoothing arrangement suffers high ripple.
No, we are trying to learn things from each other that one might know and the other not.
I have never built SMPS, nevertheless I learned it the same way like you (see post #2). Still I am always open for new insights, so I take Andrew's posts serious, because they often are very informative. Therefore I ask about things I have learned differently sometimes in a provocative manner. That often brings up more detailed explanations, which help me, and hopefully other Members too, to decide whether to take that information or leave it.
I have never built SMPS, nevertheless I learned it the same way like you (see post #2). Still I am always open for new insights, so I take Andrew's posts serious, because they often are very informative. Therefore I ask about things I have learned differently sometimes in a provocative manner. That often brings up more detailed explanations, which help me, and hopefully other Members too, to decide whether to take that information or leave it.
That's OK just teasing you guys and the text books really
Pacific I didn't understand your post #2 then, and not too much discussion on where to pull the load... s/b as far away from any rectifier charging spikes on any one conductor even if you have to notch out the traces to maintain seperation on wide traces everywhere.
I understand where Andrew was coming from but he was starting to wander off a bit.
Pacific I didn't understand your post #2 then, and not too much discussion on where to pull the load... s/b as far away from any rectifier charging spikes on any one conductor even if you have to notch out the traces to maintain seperation on wide traces everywhere.
I understand where Andrew was coming from but he was starting to wander off a bit.
be specific, I want to learn too.he was starting to wander off a bit
oh not too complicated just best to keep esr and esl totals lowest rather than cascading sections with thin traces and resistors and such. You also want to keep high di/dt in a localized small loop and if it starts getting complicated with staggered sections you will soon reach a point of limited returns.
I consider neither to be right. Both have significant variations from good practice.stormsonic said:
me too
me too in none of my amplifiers i ever use the 0V of my amplifier connected to the case or to the electrical ground of the 220/110 line
tha can cause a million complications
stormsonic said:
me too in none of my amplifiers i ever use the 0V of my amplifier connected to the case or to the electrical ground of the 220/110 line
tha can cause a million complications
Re: me too
I would like to see both of your preferred diagrams.
..Todd
AndrewT said:
I consider neither to be right. Both have significant variations from good practice.
sakis said:
me too in none of my amplifiers i ever use the 0V of my amplifier connected to the case or to the electrical ground of the 220/110 line
tha can cause a million complications
I would like to see both of your preferred diagrams.
..Todd
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Solid State
- How to do the best of connection in multi-CAP power supply