Aleph-X: High-Power Version

[Blitz]

Ok, to give this threat again some life and having played around now with Pspice for some time, here it is.

Ratings:

voltage 22,7 volts actual rail voltage at the fets
bias 6,4 amps total bias for one channel
ac current gain 50%
number of fets 12 total number of fets for one channel
Peak current 6,4 amps maximum output current
power 8 Ohms 107,1 watts
power 6 Ohms 122,9 watts
power 5 Ohms 102,4 watts
power 4 Ohms 81,9 watts
power 2 Ohms 41,0 watts

Dissipation 290,56 watts
Dissipation per fet 24,21333333 watts
Transformer secondaries 17,46 volts for 1.3
18,92 volts for 1.2
20,64 volts for 1.1
22,70 volts for 1

Transformer wattage 581,12 watts for factor 2
871,68 watts for factor 3
Source-Resistor 0,47

Frontend with jfets for better sound. I have simulate the frontend with the 9610 and with 2sj109 in comparison and if you get them to the same squarewave-response, the jfet has even a better high-frequency response. In reality the value of the small caps need to be varied, but it shows that the jfet-front-end should work nicely as well.

The change of the feedback-resistor gives as well better high-frequency-response and the change of Ian's resistors from 4,7k to 1k gives better squarewave-response.

Any comments on this (especially from someone who has built it), is highly appreciated.

Best Regards

[carpenter]

Your schematic drawing is beautiful.

Why do the wattage values drop when the speaker impedance is lowered below 6 ohms? I'm planning on building 4 LAB sub horns per channel and they are driven by two 4 ohm speakers each. Because of the acoustic loading on the driver the ohm value will be higher, but I'm still going to need an amp that can play into less than 1 ohm load.

Crazy huh?

LAB horn project webpage

[roddyama]

Hi John,

If you go to page 3 of this thread, you’ll find a post by William (Wuffwaff) with an attachment. The attachment is an Excel spreadsheet that allows you to easily change the voltage, current, and number of output devices and see the results in Watts for 4, 6, and 8 Ohms (I also added 16ohms), as well as other loading results. It really is an eye opener.

For example, for a particular rail voltage, if you increase the bias current 1 amp at a time, you find a point where the Watt output becomes Voltage limited (clipping) no matter how much you increase the current beyond that point. The lower the load impedance, the more current you need to output a given Wattage.

Rodd Yamashita

[carpenter]

and I don't have Excel. Bummer.

Would you mind posting a spread sheet with some figures?

[jh6you]

Originally posted by Blitz
ac current gain 50%

Based on my experiments, I think that R47 and 49 are to be 450-500.

JH

[Blitz]

Why is that ? It compromised the bass-response if changed to 500.

Best Regards

[jh6you]

I did not mean any sound character, but only technical specification. In the specification of the amp, the ac-current rating is indicated as 50% and powers are estimated based on it. Therefore, I just thought about how to get the 50% value.

Someone could measure the rating after construction of the amp.

JH

[grataku]

Blitz,
the values of the 'macmillian resistors' at 1k seems on the low end. Nelson might have characterized the effects of these resistors on feedback since I remember mentioning something negative about using such low values.

[roddyama]

Originally posted by carpenter
and I don't have Excel. Bummer.

Would you mind posting a spread sheet with some figures?

Hi John,

I have attached a zip file with 2 CSV spreadsheets. These spreadsheets are the power maps for the Aleph X according to William's calculations for 4 and 8 ohm loads. The maps covers rail voltages from 10 to 60 Volts and bias currents from 1 to 20 Amps.

Enjoy,
Rodd Yamashita

[BrianGT]

I started a version of the Aleph-X pcb a while back with 2 or 3 devices per bank, as shown in the above schematic. Here is my current board layout with the 12 total devices. If anyone wants my board layout files (I did it in Protel), I can post them. They are done with the schematic editor, and then routed by hand using their utilities, so they can easily be modified to accomidate other's needs.

I was going to make a high powered version, but since playing with lowered power designs, I found out that my speakers don't need more then 50W per channel to be happy.

Here is a picture of my last layout. I have more pictures in my gallery at:

http://brian.darg.net/aleph-x

I made some board for the 8 devices per channel, but never finished it.

--
Brian