PCB EAGLE Project for Aleph30/Aleph5 Hybrid available
Hi all,
I have gotten my boards from a protofab in Bulgaria (run.to/pcb). They look pretty good and I was able to get 4 channels done for US$120. They measure 12.6"x2.3" and I laid it out like the aleph30 production model, but for TO-247 packages. I have the EAGLE projects for the left and right channels if anyone is interested or can't wait that long for the passdiy.com version
Now, to populate it and power it up !
Hi all,
I have gotten my boards from a protofab in Bulgaria (run.to/pcb). They look pretty good and I was able to get 4 channels done for US$120. They measure 12.6"x2.3" and I laid it out like the aleph30 production model, but for TO-247 packages. I have the EAGLE projects for the left and right channels if anyone is interested or can't wait that long for the passdiy.com version
Now, to populate it and power it up !
questions for Tortello
Hi Tortello,
the idea of an adjustable bias sound good and I would like to build my Aleph 5 like that. A few thing are still not quite clear to me so I would be pleased if you could answer my questions.
1. If you adjust R19 (Aleph5), do you also change the value for R21? If so how and if not, what value do you use?
2. The negative clipper you talk about is triggered when the voltage over R16 is about 0,6V. Right?
3. I already bought induction free 1 Ohm resistors for Q6-8 and Q18-20 so I would like to keep them. Does this mean that the minimum value of the bias get´s bigger (no 1.5A possible)?
4.C7 is connected from collector to emitter with the Aleph5 and from collector to base with the Aleph60 and not there in the Aleph3. Do you know the reason for this and wich configuration do you use.
5.C8 (10pF) is used in all Alephs except Aleph3. Is it right that this reduces gain at very high frequencies to keep the amp stable? What kind of capacitor is used here?
cheers,
william
Hi Tortello,
the idea of an adjustable bias sound good and I would like to build my Aleph 5 like that. A few thing are still not quite clear to me so I would be pleased if you could answer my questions.
1. If you adjust R19 (Aleph5), do you also change the value for R21? If so how and if not, what value do you use?
2. The negative clipper you talk about is triggered when the voltage over R16 is about 0,6V. Right?
3. I already bought induction free 1 Ohm resistors for Q6-8 and Q18-20 so I would like to keep them. Does this mean that the minimum value of the bias get´s bigger (no 1.5A possible)?
4.C7 is connected from collector to emitter with the Aleph5 and from collector to base with the Aleph60 and not there in the Aleph3. Do you know the reason for this and wich configuration do you use.
5.C8 (10pF) is used in all Alephs except Aleph3. Is it right that this reduces gain at very high frequencies to keep the amp stable? What kind of capacitor is used here?
cheers,
william
Hello Wuffwaff, I'll try to answer to your questions:
1. The R19 value has no pratical effect on the simmetry of the two halves, so you can leave it as is.
I use a trimmer [... another trimmer, I know :-( ] to set a precise simmetry of the "two currents", I don't recall the precise value, but if you want I can open the amplifier and see it asap.
2. Yes, but the effects of the clipper begin with a bit less than 0.6V => use a wide margin (the IR Hexfet are very powerful, they allow this).
3. With 1 Ohm of source resistors, the bias current can't be set higher than the original Aleph5's, because the "contribution due" to R19 value is marginal: the whole contribution is set by the 1 Ohm resistors.
If we suppose about 2A of bias current - more or less the A5 original choice - it means 660mA per Hexfet, or 0.66 Volts between the Q5 base and emitter: about the whole contribution for the Q5 biasing: you can increase the R19, but the current doesn't change substantially.
4. The C7 purpose is to avoid an auto-oscillation of the dynamic current source.
His value depends on number of Hexfet of the o/p stage, the position of the capacitor - A5 vs. A60 position - and on the PCB layout, at least I think: I use the A5 layout for layout convenience, and with my board I set it 0.022uF, or 22nF.
Note that the schematic reports 47nF, but the PCB layout scheme reports 10nF: with this value (10nF), in my Alephs I noted an HF oscillation, that disappeared with 0.022uF.
I wrote time ago to Nelson Pass, and he, very kindly, anwered me that this is not a big issue, "whatever works", he wrote.
5. Yes, that's right: you may verify this with a square-waves generator and an oscilloscope.
I think that an high-quality capacitor is needed here: I use silver-mica caps for their stability.
Best regards,
Tortello
1. The R19 value has no pratical effect on the simmetry of the two halves, so you can leave it as is.
I use a trimmer [... another trimmer, I know :-( ] to set a precise simmetry of the "two currents", I don't recall the precise value, but if you want I can open the amplifier and see it asap.
2. Yes, but the effects of the clipper begin with a bit less than 0.6V => use a wide margin (the IR Hexfet are very powerful, they allow this).
3. With 1 Ohm of source resistors, the bias current can't be set higher than the original Aleph5's, because the "contribution due" to R19 value is marginal: the whole contribution is set by the 1 Ohm resistors.
If we suppose about 2A of bias current - more or less the A5 original choice - it means 660mA per Hexfet, or 0.66 Volts between the Q5 base and emitter: about the whole contribution for the Q5 biasing: you can increase the R19, but the current doesn't change substantially.
4. The C7 purpose is to avoid an auto-oscillation of the dynamic current source.
His value depends on number of Hexfet of the o/p stage, the position of the capacitor - A5 vs. A60 position - and on the PCB layout, at least I think: I use the A5 layout for layout convenience, and with my board I set it 0.022uF, or 22nF.
Note that the schematic reports 47nF, but the PCB layout scheme reports 10nF: with this value (10nF), in my Alephs I noted an HF oscillation, that disappeared with 0.022uF.
I wrote time ago to Nelson Pass, and he, very kindly, anwered me that this is not a big issue, "whatever works", he wrote.
5. Yes, that's right: you may verify this with a square-waves generator and an oscilloscope.
I think that an high-quality capacitor is needed here: I use silver-mica caps for their stability.
Best regards,
Tortello
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.