Breaking all the rules, a "custom" bridged story

[Redshift187]

Here's the back panel (the lid is off still).

[Redshift187]

I decided to put the 1000 uF caps on one channel, that way I can do a direct comparison. Got the amp set up and tested DC offset. Looks like about 3 mV on the channel with the 1000 uF caps and 4 mV on the other. Tomorrow I will test with my test speaker, then if all goes well I'll start the burn in.

[Redshift187]

After burning in the small cap rectifier board and listening to it, I replaced with the large snubberized cap rectifier board, burned it in and listened to it as well. My test speakers are 150W Pioneer 3 way floor speakers. Keep in mind I've only ever owned a Pioneer combo receiver/tuner/eq/cassette setup. My source was a Pioneer DV 353 dvd player, and the material was David Gilmour Live In Concert and David Gilmour Live At The Royal Albert Hall.

Different setups:
1. 2,200 uF per rail on rectifier board and 10.1 uF per rail on amp board. Great mid and high range clarity (3 out of 4). Almost no bass below 150 Hz (1 out of 4).

2. 2,200 uF per rail on rectifier board and 1010.1 uF per rail on amp board. Best mid and high range clarity (4 out of 4). Very little bass below 100 Hz (2 out of 4).

3. 15,000 uF per rail snubberized on rectifier board and 10.1 uF per rail on amp board. Decent mid and high range clarity (1 out of 4). Good bass down to around 30 or 40 Hz (3 out of 4).

4. 15,000 uF per rail snubberized on rectifier board and 1010.1 uF per rail on amp board. Good mid and high range clarity (2 out of 4). Excellent bass down to around 20 Hz (4 out of 4).

The difference in mid and high range between the first two and the last two was not major. A tiny bit of the "magic" went away in the last two, but it's still amazingly clear. The texture of vocals is the best I've ever heard. The difference in mid and high range with or without the 2,000 uF on the amp boards was minimal. I did detect that the side with the extra capacitance seemed clearer in the mid frequencies.

The difference in bass between the 4,400 and 30,000 uF on the rectifier board was amazing (albeit expected). 30,000 uF has just the right amount of bass in my setup. The difference with and without the 2,000 uF on the amp board was once again minimal.

[samsagaz]

What About this Setup ?

15,000 uF per rail on rectifier board and 1010.1 uF per rail on amp board?

[Redshift187]

I didn't try the 15,000 uF per rail without the snubbers, as it seems to be a general consensus that large capacitor values should have snubbers.

[Redshift187]

Here is the final schematic of the PSU that I'm using.

TR1 is a Plitron 300VA 18V dual secondaries
B1 and B2 are made with MUR860 diodes
R1 and R2 are 1,500R 2W generic carbon film resistors
C1 and C2 are 15,000uF 50V Panasonic TS-HA capacitors
C3 and C4 are 0.1uF 50V Kemet Golden Max capacitors
R3 and R4 are 1R 0.5W Vishay NFR25 metal film resistors
C5 and C6 are 0.1uF 63V Vishay MKP capacitors

[Redshift187]

As it turns out, this is very much like the chipamp.com snubber PSU design, although I had to use a ceramic cap for the small bypass, as the pin spacing wouldn't allow a polypropylene one. Also, the main caps are a little larger at 15mF as opposed 10mF.

[AndrewT]

Originally posted by Redshift187
Different setups:
1. 2,200 uF per rail on rectifier board and 10.1 uF per rail on amp board. Great mid and high range clarity (3 out of 4). Almost no bass below 150 Hz (1 out of 4).

2. 2,200 uF per rail on rectifier board and 1010.1 uF per rail on amp board. Best mid and high range clarity (4 out of 4). Very little bass below 100 Hz (2 out of 4).

3. 15,000 uF per rail snubberized on rectifier board and 10.1 uF per rail on amp board. Decent mid and high range clarity (1 out of 4). Good bass down to around 30 or 40 Hz (3 out of 4).

4. 15,000 uF per rail snubberized on rectifier board and 1010.1 uF per rail on amp board. Good mid and high range clarity (2 out of 4). Excellent bass down to around 20 Hz (4 out of 4).


The difference in bass between the 4,400 and 30,000 uF on the rectifier board was amazing (albeit expected). 30,000 uF has just the right amount of bass in my setup.

if you believe many of the designer/builders on this Forum then it's your imagination telling you that the High Cap produces more/better bass.
"albeit expected" why?
Because it does, but few want to believe us.

But you are confirming Peter Daniel's results that zero rectifier capacitance and high values of decoupling capacitance at the chipamp do give the best midrange.

Seems an excellent case for bi-amplifying all two/three way speakers.
Use the High cap (ver 4) for the bass and the zero cap (ver 2) for the mid and treble. This is NOT an active speaker set up.

[Redshift187]

Does anyone (AndrewT ) know how to calculate the time period for the R/C network of the PSU when using a bridged amp?

The amp schematic I'm using is:

[AndrewT]

for an 8ohm load each half sees a 4om load and the R of RC is 4ohm.

But each amp draws from the alternative rail and this may relax the requirement for the amps to be fed from separate PSUs. If the PSUs were separated then the total capacitance is 160mF (2off @ +-40mF) to get 4Hz @ -1dB into your 8ohm load, (F-3dB=1Hz on PSU when +-40mF drives 4r0).
But the pair drawing alternately on a common PSU may get away with 80mF in total for the same frequency response.


But, I would never bridge a pair of amps that were not designed from the beginning as a bridged (balanced drive) pair.