Aleph-X PS question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I was playing around with Duncan's PSU Designer II and found that if I entered the following:

12.5VAC Xformer -> Bridge -> 230000uF -> 0.47R (50W) -> 10000uF -> all into a load of 5k Ohms

This produces ripple (albeit tiny).

1) What is the input impedance of the Aleph-X?

2) What are the effects of raising / lowering the resistance of the PI-Filter>

Thanks,
Gaz
 
Hi Rarkov,

I didn't read the whole thread but I understand it's about a PSU design for a poweramp.

I thougth a PushPull Class A circuit present a load of about 2x the speaker load it is intended to drive, for a bridge topology like the AlephX this is about 1x the intended speaker load.... am I right? So that would be 8 Ohms. :att'n:

Put this in the PSU simulator and see what happens....

Regards,
Thijs
 
The resistor cause a voltage drop which increases with resistive value and Iq... that the downside...
but it also form a Lowpass filter with the cap after the resistor, filtering high frequency component of the rimple... ehhhh 0.47Ohm and 10mF has a Fc of about 34Hz. The fundamental frequency of the supply rimple is 120Hz or 100Hz :scratch: .. but the spikes can be much higher ....


I hope this helps a bit,


Regards
Thijs

edit:
lowering the value of the resistor lowers the voltage drop (and heat dissipation) but also make the Fc higher...so it's a trade of.. that's why inductors are so ideal in place of the resistor...
 
Hi Rarkov,

7 volts drop sounds a bit high.. I'm gonna simmulate some PSU's too a see what happens.. Can you tell me the Iq and rimple through the 8R load.

I calculated the filter cut-off frequency as 1/(2*pi*R*C). The R being 0.47 and the C being 10e-3... But I think with that low a resistor the Caps own equivilent series resistance plays a significant role...

I really think an inductor or Capacitor-multiplier (BJT or MOSFET) is a much mucg better option.


What do you all think?

Regards,
Thijs
 
:dead: :dead: :dead: whatttttttt... .. 5A from 12 Volt .... what are we driving here?

It is a PushPull amplifier (in effect at least) and it is suposed to drive 8 Ohm right? That'll be 4 ohm in bridge .. let's hope for 3 ohms minimum....So Ipeak will be 12/3 = 4 A since it is pushpull 2A Iq should do it... that's per amplifier


hmmmmm

I see ... 2 amplifier draw 4A ... that's close to 5 A .. That's a lot of current to rectify.. What was Nelson thinking....... :bawling:


The good new is that I've got some sims ready with supprising(or not ;) ) results...


Regards,
Thijs
 
OK.. the results

AC 50Hz, 12V AC, 01R series resistance from tranny, diode bridge, Load was always 8R which reflects one perfect PP-amp in 4 Ohms... here goes

C
20mF:
Vmin=10.05 Volt, rimple= 488mV, very high inrush current :att'n:

CRC
10mF, 0R47, 10mF:
Vmin=9.49 Volt, rimple= 215mV, more sine like rimple, less High freq

CLC
10mF, 2mH (0R2 in series) , 10mF:
Vmin=9.9Volt, rimple=110mV, 1mH was just as bad as 0.47R, a 4mH inductor would be far better..


CTC MOSFET
10mF, IRFP140 (100K/1uF to gate), 10mF
Vmin=6.885 Volt :bawling: , rimple = 5mV :clown: The voltage drop along the MOSFET is toooooooo high...it start dissipating a lot og heat..



:nod: www.sound.au.com

CTC
10mF, BC141/2N3055, 10mF
Vmin= 9.0 Volt, rimple .... tada .... 0.75mV :clown: :clown: :clown:

Regards,
Thijs
 
Grataku has a good point regards the ripple if C1 in your CRC PS.

Mr Pass in a previous search of PS threads stated the caps after the rectifer should be rated for at least 3 times the ripple current of the intended current draw.

This means the C1 capacitor must be rated at 13.5 amps at least for reliability/ long life. I image that 1 x 10Mf would not have this rating, and would recommend 2 or 3 10Mf in parrellel, the 2nd bank of caps should be equally large to represent a low ERS etc to the amps

macka
 
Member
Joined 2002
Paid Member
PSU...

I think this will work....
For improvement, try more C2, and better quality (lower ESR).

Arne K
 

Attachments

  • untitled1.gif
    untitled1.gif
    35 KB · Views: 334
Ohhhhhh no...... you love firework don't you...

Somebody thinks kicking 250,000uF in a PSU wil do the job.... (I take it you mean 20mF not 20uF or did you?)

What did your simulator tell you ... ?


Just for fun I adjusted my values to you circuit... sparks flying out my computer I can tell you.. :clown:

Rimple = 43mV which is more than 57 times higher than a far cheaper Cap multiplier..

Vminimal = 9.0 Volt .. no improvement here either...

I inrush wil be 150 Ampere :att'n: , I don't think that's practically possible but what just to give you a feeling for the amount of firework you can expect.

I peaks wil be around 25A ..... .. the tranny will hum like a .. well it hum a bit ..:clown:

Please post you're results...

Regards,
Thijs
 
Member
Joined 2002
Paid Member
Re: PSU

Hmmm, I do not see the problem, :scratch:
-and ofcourse, there need to be inrush current limiter. (KCC CL-60)?.
And the values I tried was: C1: 230 000uF, C2: (should be): 20 000uF, and adjust ESR values to typical specs.

Arne K

And if/when budget allows, increase C2 (or add additional).
(I can send you a "spark-arrest" for your computer)...:clown:
 
Thjs,
I can confirm that there is NO problem with Cobra supply idea.

My PS is actually 250,000uF-0R15-250,000uF with dual 30A bridge. Four caps and two resistors was pretty simple to build.
I would have done it if I had to start paralleling caps.

I have a thermistor on the primary, don't remember the type (some CL-XXX) and nothing else really. Don't think the transient is too bad with 15V rails. I wouldn't try removing the thermistor though! ;)
 
Ok.. that's allright then .. I guess a thermistor works very good to limmit the firework.. I would think twice before doing it that way .. then again.. Nelson uses 200.000uF in the XA200.. so it should be possible... Still.. somehow ..(not just to win some argument) .. I find a BJT-cap multiplier more elegant than 2500.000uF of capacitance that only comes almost close to the rimple rejection of the a BJT-cap multiplier ....



How about some more info/pics/etc about DIY PSU inductors

Regards,
Thijs
 
tschrama said:
Ok.. that's allright then .. I guess a thermistor works very good to limmit the firework.. I would think twice before doing it that way .. then again.. Nelson uses 200.000uF in the XA200.. so it should be possible... Still.. somehow ..(not just to win some argument) .. I find a BJT-cap multiplier more elegant than 2500.000uF of capacitance that only comes almost close to the rimple rejection of the a BJT-cap multiplier ....



How about some more info/pics/etc about DIY PSU inductors

Regards,
Thijs

I'll grant you that I went overboard with the caps and that on paper the ripple figure of the regulator look good, Nelson also uses the regulator for the zenV4.
However, real energy is stored in real caps, virtual caps are bringing down the ripple by dropping the voltage just like any other follower regulator. For double the current the X configuration requires I find I sleep better knowing a wirewound resistor is on the job rather than thinking about 5 amps going through a mosfet.
Then you have the separate board, heatsink etc. It comes down to preferences.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.