Aleph 3, P 1.0 , P 1.7, help for newbies (and confirmed…)

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here, we can found some questions and answers (I hope…) for the people who is not habit with DIY audio parts.

First for the Aleph 3:
-It’s really recommended to build separated alimentation for each canal ?
-For this (or these) alimentation we need a toroïdal transformator, how many VA for this ? (in minimum because everybody know that it’s better when the transfo is bigger or overkill…)
-How many is the power dissipation for each transistor of the Schematic ? (very useful for choosing heatsinks)
-The equivalent for the IRF9610 are : 9510, 9620, 9520, 9110, 9210 right ?
for the IRF 244 are : IRF140, 240 , P140, P240 isn’t it ?
- In a lot of Pass project we can see the use of thermistor for grounding and thermic protection : what is the species and reference ? The principle of the use of a thermistor for grounding ?
- It was recommended to make the bridge diode (monoblock) with diodes alone ? The species and ref of the diodes that you use ? It’s recommended to put on parallel of each diodes a caps ? How many µF the caps ?
-I have heard that we can use two diode bridge rectifier per channel. How it works ? It’s really recommended ?

And for the poor young electronic man I am, what does it mean “Front end” (I’m a little shame, I’m sure I know it under a French name).

For my part I draw PCB with good PC program (P-CAD 2000) and I learn to use it in school, so send me the maximum informations you can for the drawing of Aleph 3 and P 1.0 1.7 hybrid.

I’m a little afraid with the volume selector of the P 1.7, I think too it was overkill to use a microcontroller. I already worked for optical interrupters (like Three channel Optical Incremental Encoder Modules HEDS-9040) coupled with Quadrature Decoder/Counter interface IC’s (HCTL-2000 series) who is providing binary code directly (that we can use for command a lot of relays).
That is a hope ? For made it I need to know :
- What it’s is the precise reference or species for the optical interrupters used in Pass P 1.7

For this decoder I need to build a 14 MHz clock generator, I think it will be easy.
I think we can abort the use of the IR telecommand…
For the channel select a simple commutator will be more indicated (with or without relays, cheaper…).

For the others can we say about the things to know when we put a heatsink on a FET. What is the best way to isolate ? to thermic conduct ? more and more…

I notice one thing. We talk never enough about the simple electronic, it’s was we are expect to use in the Nelson Pass philosophy, isn’t…?

I wish this thread became a reference for who is going to build is Aleph project. Today I ask a lot of questions, but for the further time I will answer to others.
I’m going to holiday for the next two week, so make good response, not only for me (the half of this question I’ve got the answer) but for make a center of the Aleph build.

Take good time, I stay on beach for two weeks and I will be back for finish the PCB I have promise.

I you have other questions take place bottom, it was make for this…
Absolute value encoder

Consider using an alps ACE -- absolute contacting gray code encoder -- 128 steps. You need a converter to binary (and logarithm unless you factor the log into resistor values).

At least you don't have to worry about quadrature encoders + you get absolute position (powers up at same volume as when you turned it off).

I think I have seen one with true binary direct out, but have not found it since, so that may have been a misunderstanding.

Good luck.

using ACE


I am using an ACE in my preamp to read the position of a normal ALPS pot. Check out the website if you are interested. If it still interesting drop me a mail.

PS, The output of the ACE is not 'logical'. You need a table (in eprom or in the code) to convert it into the something usefull.

Also saw on the net that there is a 'formula' to calculate it.
But that takes time and code-space in a controller.



From the link above:

The classical way to convert Gray to Binary is as follows:

Feed the Gray MSB directly through to be your Binary MSB.
Then make an XOR of the the Gray MSB and 2.MSB. Output will be your Binay 2.MSB.
Then XOR this output with your Gray 3rd. MSB.
Output is now your 3. MSB Binary.
Proceed with XORing each output with your next following unused Gray bit until your all done.
That's all there is to it.
On the other hand, - generating a LUT in a PROM might do it in one chip,
as 8 bits will need two XOR DIPs. One DIP will suffice up to 7 Gray bits...........

But the output is not "normal" gray code:

Gray Code 7 6 5 4 3 2 1 0
0 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 1
2 0 0 0 0 0 0 1 1
3 0 0 0 0 0 0 1 0
4 0 0 0 0 0 1 1 0

0 0 1 1 1 1 1 1 1
1 0 0 1 1 1 1 1 1
2 0 0 1 1 1 1 1 0
3 0 0 1 1 1 0 1 0
4 0 0 1 1 1 0 0 0

Since it is possible to 'compute' the binary (with a microprocessor), so i agree it should also be possible
with logic. But i have the routines at home somewhere (for a pic16f84) and they are many lines...
I guess this 'translates' into much logic.
A table (eprom or inside a microprocessor) seems easier.

Or am i (and the person who did those routines) wrong ?
Sorry,--I was quite in a hurrry yesterday...

Of course code conversion can be done in LUTs, - be it bare EPROM or with processors, but somehow there is a resistance within the audio community to use processors within amplifiers, so the X-Or converter is fully asyncronous. Many people also don't have the tools for programming PROMs or CPUs, and only need the Gray-to-Binary convertion to drive e.g. a relay based attenuator.
The beauty of the microcontroller i of course that it gives the possibility of remote control as well, without any extra counters etc.

Another thread under Solid state amplifiers set out to make a "universal amplifier", and the different parts were divided into sub-threads, but the thread for the control system seems to be dead.....??

I am working on the ideas for a microcontroller based control system for volume, input selection etc. but my spare time is farly limited, and will be for some time to come. I have even launched the idea into a BSc project at a local college, but there will be no results from here until june, - if at all.

What i meant to say with my post is that it is not possible to convert the ACE output with only XOR.
The tables in my post were not really clear, tabs dont work with posting.

What i posted is that the graycode for pos 0 should be (in 8 bit): 0000 0000, converting into binary
with XOR then also gives 0000 0000, etc.

The ACE gives 0111 1111 as output for pos 0, 0011 1111 for pos 1 etc. So what bourns means with gray
code is that only one bit changes when the position changes one step. But it is not convertable with
XOR like 'normal' gray. So i think if you use the ACE, it can only be done with EPROM or processor
(in a practical way).

It is possible to calculate the binary from the ACE output, but it takes many lines of code in a
processor or much processing time with less lines. Again, i 'translate' this in much logic if you
want to use logic chips. The EPROM or table in a processor is easiest, calculating with a processor
is also an option.

I am using a micro with sleep mode. If you don't push a button on the remote it is sleeping.
The clock is dead in sleepmode, no problems with interference. Click on my website and you will find
a preamp with remote control, ACE, microprocessor etc. Not an idea, but a working piece (actuall there
are two now, one with tubes and one with opamps, used by a friend of mine). If you find that there might
be something usefull, give me a mail. There are others here to be found, e.g. looks like a good piece of work.

Hope the ACE thing is clear now. Miles i don't want to steal this threat !!


Great your project is a very good base for my own preamp project.
I willl try to replace your PGA circuit with the preamp Aleph Class-a.
And for the logic instead of a Pic I wil use a 68HC11 (I learn about it a long time in school) or the big DS89C420.

A step later I wll try to make my own N/A converter in order to have less loss on the signal source. For this the DS89C420 would manage it too.

And the idea of the vaccum display is very good, I will try it.
But instead of a button for each source I will take a simple encoder (like the volum but with less resolution) and make the display of the source on the vaccum too.

Very interesting, there I have a lot of work here... :cool:
Oh, one other question Guido.
Are you sure that the frequency of your PIC (generate by a quartz isn't it) make some disturbance on your analog circuits ?
I don't remember but its frequency reach the order of Mhz (a multiple of 4Mhz I se) and it will not fuzz your source signal, and if you separate the power supply of the PIC or plug near of it some little cap (it's already done I see). Why ?

Or why not? If the sleep mode was integrated on the pic this application look to be simple. And very interesting for the learning point of view.

Maybe I will take some interrest to study a little the pics circuit.

In your side are you interested by the 68hc11 or the ds89c420?
A conjoint work will be very interesting.

Your right...with me not reading the specsheet in the first place.... this one seems easiest to convert in a LUT..
However..- there's something inherently strange in using 8 bits to code 7 bits...:xeye:

Just having printed the specsheet, I will study it carefully.. I think by shuffling the bits there must be a standard gray code in there somwhere.....??

BTW- I have been using shaft encoders, mostly optical, for many years, but usually with a lot higher resolution, and these are quite expensive little buggers...anyone knows the price for this one..??

I think you mistake mark's preamp (the link in the post) with mine...
I don't use a vacuum display (don't want it, has a clock which is always 'on').
It looks fancy for sure, but do you really need it ??
Also don't have buttons for each source (i have remote control only).

-> To see (some info on) my preamp, click on WWW under my name!

The 68HC11 has no stopping of the clock, as far as i know. Try to use a processor
that does. Maybe a newer Motorola chip, so the assembly still looks familiar.

The quartz of the clock on a pic16f84 is DEAD in sleepmode.
So no disturbance. Don't know if Mark stops his clock though.


The ACE has indeed only 128 positions, while 8 bits could code 256.
Has probably something to do with the internal build and the fact that
only one bit is allowed to change for 1 step.

I tried the shuffling. Good luck..! If you do find it, please let me know.
It would save me the usage of a 24C02 EEPROM (that's where my table is).

Farnell NL has them for E9.21 ex VAT. Type EAW0J-B24-AE0128, ordercode 7005258.
Not too much in my opinion.


About disturbance, well i don't hear it when it is on. But the chip can do this,
so i use it. If the high freq get on the audio signal, it might cause distortion
in the poweramp (audioanalyse from '89, if you are french, you might know it..)

I also choose it because it is in-circuit programmable, has EEPROM and free development
tools, including simulation. And the most important: tons of code on the internet to use,
e.g. I2C etc, etc. Only the RC5 i wrote myself, cause i liked the state-machine way and what i
found on the net so far was not really good. I can say, the RC5 part works excellent.

And yes, there are probably better processors out there (now). But it does the job nicely
and it is small.

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.