Aleph-x. Help me starting it up.

[anders.a]

Help me fire up my Aleph-X. This is how I think it should be done. But I’m not sure
because the information is scattered all over Diyaudio. Most are collected from the wiki, various posts and this link:
http://www.penguinlovers.net/audio/alephx_proj.html
Have I missed something? Is there a better way? Have I completely misunderstood parts or all of it? How should you do it?
PLEASE make your comments.


1) Check an recheck that you have mounted the right components on the right spot.
• Right place
• Right values
• Not mounted them backwards.

2) Install the output boards onto the case. Mount the mosfets with mica insulator and thermal paste (or better). If the mosfet's mounting hole is not insulated, then one needs to insulate the screw to the heat sink. I prefer not initially to mount them on the board when testing, but do some wiring between mosfets and the board. It's easier to change them if you break one of them.

3) Install the main board and connect to the rail. For initial connection, add a 0.5 ohm/2W
resistor on both rails. This will serves as power protector if something goes wrong!
(That means, if something goes wrong, the resistor will burn out first, but not damaging
your circuit!).

4) Connect a variac and slowly raise voltage. You may want to install everything except Q1, Q2, Q10 and Q11, then power it up to see that the diff pair and it's current source work properly, and that there are no surprises with the rest of the circuitry. This could save you some expensive matched power transistors.

What should I expect here? What to expect from current source, diffpair and rest of
the circuitry if everything is ok? What and where exactly do I measure?

5) Connect all mosfets

Now its time to adjust the AC

6) Put a 1.5k pot in place of R12/R34. Connect an 8-ohm ‘dummy’ load, or even a
real speaker to the output. Feed the amplifier input with a 60Hz sine wave (or perhaps
1kHz if you’re using an oscilloscope - 60Hz is suitable for most multimeters).
Measure the AC voltage across the paralleled R2/R3 output resistors, and across R5 (a big advantage is two voltmeters).
Trim R12 until the voltage over R5 is equal to the voltage over R2/R3.
Now your Current source is set to 50%.
Repeat the same procedure for the R34-side.
Measure the value of the 1.5K trimpots and replace them with fixed resistors.
Aleph-X builder's thread.


Then its time to adjust DC

7) For DC offset adjustment, there is the Absolute DC offset and the Differential
DC offset. Set V1 and V3 initially to R=0 and V2 to 0
For this part, use Grey Rollins steps:
Assuming that you use the output current adjustments at all. Set them such that you have the same voltage drop across the source resistors on the output MOSFETs comparing one side to the other (that is by adjusting VR1 and VR3 and check each side's voltage).

8) Set the front end current source adjustment (by adjusting VR2) so that the
absolute DC offset (DC offset measured from the outputs to ground) is as
close to zero as possible (if not, an Absolute DC offset as high to 1 ~ 2V is still OK!).
As long as you're close, you're okay. The speaker does not see the absolute
DC offset. The only reason you fiddle with it at all is that it will cut into your
potential output voltage swing to either positive or negative, causing premature
clipping. It will drift a bit, particularly with heat variations. Don't worry about it!

9) Let the amp cook for a while and readjust if necessary. Thirty minutes to an hour should be enough to get the amp nice and warm. The output current adjustments can help with small relative DC adjustments at the output.

10) Differential DC offset is important, as this is what speaker will see and too high will definitely hurt your speaker. By design, as long as it is lower than 100mv, you are OK!

11) Now check the bias voltage across (all) the R5 resistors to be 0.47V. If not, adjust the VR1 to let close to 0.47V. This translate into 0.47V/0.47ohm = 1.0A. Therefore, the bias on one side is 1.0A*3=3.0A. Do the same for R40 with VR2.

12) Now measure the voltage across R5 and R2/R3. Make sure their voltage is the same. If not, adjust the R? pot, so that they are matched as the same voltage. With this, we have the AC gain:
0.47V / ( 0.47 / 2 ) = 0.50A
0.47V / 0.47 = 1.0A
(1 - (1.0 / 0.50)) * 100 = 50%

13) Now adjust VR2 again, so that the absolute DC offset is as nearly to 0V as possible! (or refer to notes above).

14) Connect input from RCA and XLR. Some plan to add a toggle switch for toggle between unbalanced RCA input and full balanced XLR input.

15) Connect the output to the speaker binding post.

16) This completes one channel. Repeat for the 2nd channel mono block.

17) Enjoy the music

[wuffwaff]

Hi,

most of it is OK. Just a few remarks:

3. 0,5Ohm/ 2 watt will dissipate 4.5watts at 3A bias and 18watts at 6A bias (meaning smoke without something being wrong) It is better to monitor one source resistor and watch for a too high current.

5.5 ac-current-gain can be set much much later. I would leave the resistors/pots out until everything else works

6. depends on the source resistors and output resistors.

11. 0,47V? depends on the chosen bias and the value of the source resistors. so 6A bias with 12 fets and 0,33R source resistors would mean 0,33volt

12 this is the same as 5.5


If you use a variac you will see that bias goes down if you up the voltage (this is normal)
You will need to check the value of the caps in the feedback loop and (later) in the active current source by looking at square waves and looking for oscilation or overshoot. Istarted without caps and added them with the help of a 10kHz square wave and 8 ohm load.

hope this helps,

william

[anders.a]

And if you use Kristijan Kljucarics board?

[wuffwaff]

Hi,

I don´t think the board changes much of the behaviour of the amp except maybe capacitor values for feedback and current source.

William

[steenoe]

I don´t think the board changes much of the behaviour of the amp except maybe capacitor values for feedback and current source.

Speaking of capacitor values! Why is it allways 220uF, wouldnt it be possible to use like 330uF??

Steen

[wuffwaff]

Hi,

330uF would be possible but unneccessary.

William