Page 4 of 4

Re: Oric's 8912 volume scale

Posted: Tue Apr 27, 2021 11:47 am
by jbperin
Sorry, there's a mistake in the table above. I used a wrong rounding method.

The correct table is:

Code: Select all

      0   1   2   3   4   5   6   7   8   9  10  11  12  13  14  15
   | --------------------------------------------------------------
0  |  1   2   2   3   4   5   6   9  12  17  23  33  46  64  91 128
1  |  2   2   2   3   4   5   7   9  12  17  24  33  46  65  91 128
2  |  2   2   3   3   4   5   7   9  13  17  24  33  46  65  92 129
3  |  3   3   3   4   5   6   8  10  13  18  25  34  47  66  92 129
4  |  4   4   4   5   6   7   8  11  14  19  25  35  48  67  93 130
5  |  5   5   5   6   7   8  10  12  15  20  27  36  49  68  94 131
6  |  6   7   7   8   8  10  11  14  17  22  28  38  51  69  96 133
7  |  9   9   9  10  11  12  14  16  19  24  31  40  53  72  98 135
8  | 12  12  13  13  14  15  17  19  23  27  34  43  56  75 101 139
9  | 17  17  17  18  19  20  22  24  27  32  38  48  61  80 106 143
10 | 23  24  24  25  25  27  28  31  34  38  45  54  68  86 113 150
11 | 33  33  33  34  35  36  38  40  43  48  54  64  77  96 122 159
12 | 46  46  46  47  48  49  51  53  56  61  68  77  90 109 135 173
13 | 64  65  65  66  67  68  69  72  75  80  86  96 109 128 154 191
14 | 91  91  92  92  93  94  96  98 101 106 113 122 135 154 180 218
15 |128 128 129 129 130 131 133 135 139 143 150 159 173 191 218 255

Re: Oric's 8912 volume scale

Posted: Tue Apr 27, 2021 11:40 pm
by iss
@jbperin: I've just realized that there is something suspicious in my previous captures. I'll re-do them again, together with all requested modes in your other thread (i.e. ayramp).

Meanwhile I've created a simple playground for you :) :
play-with-ay.png
You can start experimenting with AY instantly HERE !

What we (or at-least me) don't know is the AY chip internals - especially the schematic of the 3 channel outputs.
We need this to be able to include the output impedance in your equations.
I'll try to measure it by disconnecting the 3 outputs and using different loads.

Re: Oric's 8912 volume scale

Posted: Fri Apr 30, 2021 9:02 am
by jbperin
iss wrote:
Tue Apr 27, 2021 11:40 pm
You can start experimenting with AY instantly HERE !
Hey .. this is cool !!! :D
thank you for the link.
iss wrote:
Tue Apr 27, 2021 11:40 pm
What we (or at-least me) don't know is the AY chip internals - especially the schematic of the 3 channel outputs.
The same here ..
There's something else I find difficult to understand. It is the difference of phase between channels.
Is there any logic behind or is it just the time at which the register is programmed that trigger the signal .. regardless signals already running on other channels.
iss wrote:
Tue Apr 27, 2021 11:40 pm
We need this to be able to include the output impedance in your equations.
I'll try to measure it by disconnecting the 3 outputs and using different loads.
Sorry for I don't master electronic very well ..
which output impedance is it ?
Is it the impedance of analog channels output ?
When one channel is at high level and that the two other channels are low, a part of the current outputted from the high level goes back to the 8912 through the pins that are at 0V potential.
I must admit that I considered it negligible.

Re: Oric's 8912 volume scale

Posted: Fri Apr 30, 2021 7:51 pm
by iss
jbperin wrote:
Fri Apr 30, 2021 9:02 am
Is it the impedance of analog channels output ?
When one channel is at high level and that the two other channels are low, a part of the current outputted from the high level goes back to the 8912 through the pins that are at 0V potential.
I must admit that I considered it negligible.
Here is very simplified "model" of the outputs:
Oric-amp.jpg


Rx-int should be included in your Ohms Law equation :) but we don't know their value(s).
Actually they can be even not linear (i.e. U/R ratio can be parabola for instance :) ).

EDIT: We KNOW what's inside the chip - now we need ONLY to understand it! :lol: @mikeb: what do you think ?
8192.png

Re: Oric's 8912 volume scale

Posted: Fri Apr 30, 2021 9:49 pm
by jbperin
iss wrote:
Fri Apr 30, 2021 7:51 pm
Rx-int should be included in your Ohms Law equation :) but we don't know their value(s).
Actually they can be even not linear (i.e. U/R ratio can be parabola for instance :) ).
The Ohms Law in my equation was just here to help me approx the value you had measured with what's specified in the datasheet.

I totally agree with you that it doesn't take into account the Rx-int.

The thing is: I do hope that the sample you're going to record with oscillo will allow us to build a mathematical model of the behaviour.
And this model will integrate these output impedance because these impedance will impact the voltage you're going to measure.
Even though we don't know how it is build inside the chip, we can build an accurate enough model to build upon.

By the way, if it is more convenient for you, instead of recording voltage on pin 1, 4, 5 of AY, maybe you can record the line out signal which is outputted on the DIN connector.
At the frequency we are working at, the low pass filter (C5) should not perturb too much the measure.