Aleph-X builder's thread.

Without C9/C10, C12/C13 it got significantly better.
As I removed all 4 caps at one time I can't they which ones are "bad".

We will have to wait what you find when removing C12/C13 only. This way we will know.

I'll have to give this a try. I'm not sure that I can get to this tonight, but should be able to have a look within a few days. I'll pull C12 & C13 first to see if that makes any difference in either power bandwidth, power distortion levels, or listening.
 
OK, despite the other things I needed to do tonight, I had way more fun playing with my amp, signal generator, and scope. :D

Pulling C12/C13 made no noticeable difference in output waveforms at different frequencies and at different amplitudes for either sine waves or square waves.

Pulling C9/C10, however (with C12/C13 still out) made a HUGE difference. Here are the results:

1 kHz clip onset, 3.76v peak-to-peak input, 153w output into 4 ohms (was 150w with C9/C10)
5 kHz clip onset, 3.94v p-p input, 163w into 4 ohms (was 139w with C9/C10)
10 kHz clip onset, 3.63v p-p input, 138w into 4 ohms (was 96w with C9/C10)
15 kHz clip onset, 3.67v p-p input, 140w into 4 ohms (was 80w with C9/C10)
20 kHz clip onset, 3.47v p-p input, 125w into 4 ohms (was 65w with C9/C10)

Maximum sine wave performance is MUCH improved with C9/C10 and C12/C13 removed from the circuit. Looking at the square waves, they are slightly improved from when all 4 caps were installed. There is now a bit of a small bump on the leading edges of the trace (perhaps I'll increase C2/C4 just a pinch to compensate), but overall, the vertical edges of the traces are now more vertical than they were with C9/C10 in the circuit (most noticeable at 20kHz). This, however, is a very small difference...

Listening to music on the newly modified pair of amps doesn't reveal any immediate difference at low volume levels, and the whole house is asleep, so its not the time for high-volume output listening right now :(

Conclusion: Keep C9/C10 out of the circuit!

The other conclusion, going back to the discussion from yesterday, is that I'm surprised to see the necessary input level for these amps to produce their maximum output: I'm looking at an average input level of 3.6v peak-to-peak in order to drive 150w into 4 ohms... Looks like passive pre-amps with my Aleph-X amps won't cut it for high-volume playback. Can this be manipulated by altering the input impedance of the amp? (here is where my lack of knowledge begins to show, because if you tell me "yes", I'm not sure how to proceed...)

Its time for bed.
 
Last edited:
Zen,

C9/C10 are shown in the top left and right corners of the PCB (linked on my web page):
An externally hosted image should be here but it was not working when we last tested it.


It looks like C9 sits between the middle leg of Q3, R15, and the gate pin of the Q1 output bank. Thus, it appears to be directly causing high-frequency roll off of the output signal just before it hits the output stage. C10 does the same thing on the other side.

C12 and C13 appear to be nothing more than local power supply coupling caps for the PCB (perhaps to help isolate the front end from the output stage?), so its no surprise that removing these makes no obvious difference, so I think I'll add them back in.
 
Official Court Jester
Joined 2003
Paid Member
so - these caps are already well know "oscillation cure 1nF caps " and they're needed only if your amp have some HF garbage

it's completely sane to expect HF roll-off if they're put on pcb , with reason or not

but - hey - you didn't show them on schematic , so I couldn't even think of them

:rofl:

regarding local decoupling caps - leave them on pcb
 
Zen, thanks for the description of C9/C10. I didn't understand their function when I was first building the amp and probably chose to include them simply because others had. I do remember some conversations about solving oscillation, so yes, it makes perfect sense that adding these caps in (to solve a problem) would also serve to attenuate high frequency performance of the amp. They aren't referenced in the schematic on my web page because they weren't included in Grey's original work up of the circuit

As for the on-board power supply caps, yes, I am going to reinstall them on the PCB. It makes good sense to have them around given the roar of current from the PSU to the output stage.

Thanks for the insights!
 
This has been a great discussion over the past few days: I've managed to fix a problem that I didn't even realize was a problem with my amp.

This has led to me to explore more formally something that I noticed some time ago. Driving my amps directly from my CD player produces much less output power than when I add my active preamp to the mix.

Is it worth it to attempt to modify the input sensitivity of the amp so that I can get full power out of the amps directly from my CD player? The CD player has an output impedance of 100 ohms and a maximum output of 2.5v. It seems that my amp wants to see a 3.5v signal in order to generate its maximum output.

The input network of my amps consists of R16/30 = 100k, R18/28 = 10k, and R19/29 = 10k in the schematic linked on my web page.

Any thoughts?
 
This has been a great discussion over the past few days: I've managed to fix a problem that I didn't even realize was a problem with my amp.

YES, same for me!!

As a THANK YOU! I post some distortion plot's I did with my HP8903B Analyzer.
These are made with unbalanced input as I still have to make XLR Adapters.
It is my XA80 otimized for input into app. 8 Ohms.
I also have another XA160 for 4 Ohms and if there is Interest I'll make these plots also

If you need more, tell me :)
 

Attachments

  • P_Dist_XA80_4Ohms_varF.JPG
    P_Dist_XA80_4Ohms_varF.JPG
    50.2 KB · Views: 541
  • P_Dist_XA80_8Ohms_1kHz.JPG
    P_Dist_XA80_8Ohms_1kHz.JPG
    35.9 KB · Views: 539
hi JBL4435,

To the untrained eye those distotion plots dont look right to me, any Gurus out there care to comment?

Actually those plots look absolutely right for an AX Amp.
For the non-believers I added a distortion curve of an original XA-60 made by LP-Magazin April 2007
 

Attachments

  • DistOrigXA-60.JPG
    DistOrigXA-60.JPG
    106.6 KB · Views: 529
  • P_Dist_XA80_8Ohms_1kHz_logarit.JPG
    P_Dist_XA80_8Ohms_1kHz_logarit.JPG
    61.8 KB · Views: 524
Last edited:
I also have another XA160 for 4 Ohms and if there is Interest I'll make these plots also

Thanks for the comparison! If you have the time, I'd be interested to see the 4ohm plots your XA160 (I assume this is your DIY version?).

Its a shame that the plots look nicer for 8 ohms (not a surprise at all), because I've optimized my amps for a 4 ohm load :D

Eric
 
Thanks for the compliment.

the rail voltage is +/- 15,5V 7A I use 8 FETs . It is enough, really.
CLC 100.000µF + 2,2mH (0,12 Ohm) + 100.000µF
The heatsinks are the KL-274 from Seifert Electronic with 0,1K/W each.
The additional heatsinks in the middle are for the softrecovery rectifier.
They become very hot, because the Uf (forward voltage) is very high.
You to not need more power
The sound is incredible!!
I use a pair of martin logan request and need very much current (7A).
At 20 Khz the request has 1,2 Ohm
The bass is so good, you can´t beleave it.
 
Last edited:
the rail voltage is +/- 15,5V 7A I use 8 FETs . It is enough, really.
High current is very important for the good sound of the Aleph-X

CLC 100.000µF + 2,2mH (0,12 Ohm) + 100.000µF
The heatsinks are the KL-274 from Seifert Electronic with 0,1K/W each.
The additional heatsinks in the middle are for the softrecovery rectifier.
They become very hot, because the Uf (forward voltage) is very high.

You to not need more power
The sound is incredible!!
I use a pair of martin logan request and need very much current (7A).
At 20 Khz the request has 1,2 Ohm
The bass is so good, you can´t believe it