
Home  Forums  Rules  Articles  The diyAudio Store  Gallery  Blogs  Register  Donations  FAQ  Calendar  Search  Today's Posts  Mark Forums Read  Search 

Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving 

Thread Tools  Search this Thread 
15th September 2012, 04:56 PM  #1171 
diyAudio Member

Hi Nico, some updates on the 1DIFFQC . Thanks WSJ, Gootee et al for reviewing PCBs.
1DIFFQC amplifier
__________________
www.harrisonaudiolabs.com 
15th September 2012, 09:03 PM  #1172  
diyAudio Member

Quote:
IF you could keep all of the leads and connecting conductors completely separated (i.e. also in parallel), ALL the way to the point of load (or wherever you need the capacitance), then adding more in parallel would always continue to be an improvement. In that case, the inductances of the caps and the leads and confuctors would reduce in exactly the same way that the resistance of parallel resistors reduces. The ESRs of the caps and leads and conductors would also reduce, similarly. And the total capacitance of the capacitors would be the sum of all of the paralleled caps' capacitances. All of those numbers are heading in the right directions as we add more caps/leads/conductors in (true) parallel. The "theoretical limit" is actually just running out of space. The caps might get difficult to fit, but keeping all of the conductors separate all the way to the point where the capacitance is needed would become impractical at some point. Usually, people switch to using ground and power planes, before they get close to that point. Even with planes, the problem that can partially ruin the improvement from paralleling is when the conductors are no longer completely separate, i.e. the currents have to share some length of conductor. Then there would be MUTUAL inductance. And that wrecks the algebra somewhat so that the inductances don't fully reduce from paralleling. One other thing to realize is that IF the leads and conductors CAN also be kept in parallel and separated, then you can use much longer connections and still get a large benefit. e.g. two conductor pairs in parallel have half the inductance of one that's the same length. But they have the SAME inductance as one that's half their length. So if you had to put two capacitors twice as far away as one original one, maybe because they wouldn't fit otherwise, the lead inductances would total to the same inductance as one capacitor only half as far away, but the capacitors' total inductance would still be halved, while the total capacitance would now be double that of just one cap. Extending that, you could use n caps, n times farther away than one, and still divide their total inductance by n, and have n times the capacitance, while NOT increasing the total conductor selfinductance! And it's likely that the connections wouldn't have to be n times as long, for n capacitors. So the total connection selfinductance can often also be reduced. Cheers, Tom Last edited by gootee; 15th September 2012 at 09:18 PM. 

15th September 2012, 09:19 PM  #1173 
diyAudio Member
Join Date: Nov 2005
Location: East Coast of South Africa

Tom, I take it that you promote the idea of the caps be used as in a star ground, or am I miss understanding it. In such a case one would be saddled with several LRC circuits.
Two distinct possibilities exist in that one is of several hi Q resonators and different frequencies which may easily excite with switching transients and the other that of causing differential currents to flow from different capacitors if the LCR circuits being grossly unmatched. The first remark may have a positive effect of several narrow filters in parallel actually increases to overall bandwidth and reduces the Q thus suppresses switching transients.
__________________
Kindest regards Nico 
15th September 2012, 09:26 PM  #1175 
diyAudio Member

Attached are some data and plots for three sizes of power supplies, with an amplifier load driving maxpower square waves into an 8Ohm load.
The square waves use double the rated power of the supply and have the same PEAK voltage as a sine at the max rated power, which was found to be a good worstcase test for finding the absolute minimum usable reservoir capacitance for a maxpower sine wave that could have any phase angle and thus any timing, relative to the charging pulses. Anyway, these are all cases where the size and ratings of the transformer should NOT come into play. They are 25W, 50W, and 75W supplies, each with two 240 VA 44VCT transformer secondaries. I plotted the % Avg RMS Error (% distortion?) of the tops and bottoms of the square waves versus the reservoir capacitance, and also plotted the maximum power rail voltage dip versus the reservoir capacitance. The data tables are also included. I have marked the absoluteminimum capacitance (which depends on the amplifier circuit) with yellow highlighting. But other than that I have not tried to find any pattern or rule for the capacitance. So make of it what you can... Last edited by gootee; 15th September 2012 at 09:45 PM. 
15th September 2012, 09:29 PM  #1176  
diyAudio Member
Join Date: Nov 2005
Location: East Coast of South Africa

Quote:
Besides that I have been lurking in the dark shadows on your thread to see what is brewing, out of pure interest for where it is going and I would not miss the all revealing formula for anything in the world. I am not joking  there are those who can successfully market any product and this may be one golden method so I am not laughing at the back of your head. Not yet anyway.
__________________
Kindest regards Nico 

15th September 2012, 09:31 PM  #1177  
diyAudio Member

Quote:
Or, if I were doing it with traces or wires, I would just keep them over/under each other (on opposite sides of a PCB), or very close to each other if opposite sides of the board wasn't possible, or twisted together in the case of wires. 

15th September 2012, 09:34 PM  #1178  
diyAudio Member

Quote:
"Saddled with" should be replaced with "benefitting from". Then I can agree. The idea is that by paralleling the caps and the connecting conductors, the L and the R are both reduced and the C is increased, all by significant factors. The alternative is the original LRC, which has worse characteristics. What else are you going to do, when the equations say that one decoupling cap and its shortestpossible connections still have too much inductance? Then the only choice is to parallel several smaller ones, AND the connections, to lower the overall inductance. Cheers, Tom 

15th September 2012, 09:41 PM  #1179 
diyAudio Member
Join Date: Nov 2005
Location: East Coast of South Africa

Tom, those graphs reveal a lot of information provided that the observer realises that the output signal (driving the load is fixed) and only the error benefits can be had. Now at what point does the error represent noting more worth pursuing.
From the data you present I would select 6000uF as being overly suitable for all three applications.and will split them into 6 x 1000 uF or even 12 x 500uF, if realestate is not at premium I would go for 60 x 100uF.
__________________
Kindest regards Nico 
15th September 2012, 09:41 PM  #1180 
diyAudio Member
Join Date: May 2007

With careful wire/track positioning it might be possible to reduce the effective inductance of capacitors. Two caps in parallel have an average current path which runs between them, so if the wire runs here then the net inductance is minimised.

Thread Tools  Search this Thread 


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Valve power supply  How to size transformer?  SanderW  Power Supplies  25  4th January 2013 05:12 PM 
How do you calculate choke size in a power supply?  Original Burnedfingers  Tubes / Valves  25  5th January 2012 01:23 AM 
power supply bypass cap size  BigE  Power Supplies  11  5th July 2011 03:59 PM 
Power Supply Case Size  diymixer  Power Supplies  1  10th October 2010 06:47 AM 
What size power supply should I get for repair work?  spooney  Car Audio  3  7th December 2007 12:50 AM 
New To Site?  Need Help? 