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

[guido]

AuroraB

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.
http://www.mhennessy.f9.co.uk/ looks like a good piece of work.

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

Greetings,

Guido

[Miles Teg]

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...

[Miles Teg]

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.

[AuroraB]

Guido...

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...

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..??

[guido]

Miles,

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.

AuroraB,

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.

Greetings,

[Miles Teg]

LOL

But do you really that it is very important to muzzle the clock ?
At this frequency, any distubrance can reach the source signal ?

[guido]

Miles,

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.

Greetings,