Wrong ADC definition at-tiny-24

[djfd]

Hi,

playing with my firmware I noticed that ADC definition of tiny24 is in contradiction with at-tiny-24a datasheet (pp.145-146)

Currently only 4 pins of ADMUX are used, while datasheet uses all six bits, and consequently modes themselves described not in full and simply wrong.

Well, the software model files are for models without A suffix, so I was curious if 24A version differs from pure 24, and compared it with at-tiny-24 datasheet briefly. It seems there are not any differences. Or am I wrong, or maybe missing something?

Proposed fix (as a patch, pls do not ask for PR, sorry), which works for me (I was able to read the chip temperature sensor)

--- a/simavr/simavr/cores/sim_tinyx4.h	2019-04-15 21:54:36.054957167 +1000
+++ b/simavr/simavr/cores/sim_tinyx4.h	2019-04-15 21:53:54.761157331 +1000
@@ -116,7 +116,8 @@
     .adc = {
         .r_admux = ADMUX,
         .mux = { AVR_IO_REGBIT(ADMUX, MUX0), AVR_IO_REGBIT(ADMUX, MUX1),
-                    AVR_IO_REGBIT(ADMUX, MUX2), AVR_IO_REGBIT(ADMUX, MUX3),},
+                    AVR_IO_REGBIT(ADMUX, MUX2), AVR_IO_REGBIT(ADMUX, MUX3),
+                    AVR_IO_REGBIT(ADMUX, MUX4), AVR_IO_REGBIT(ADMUX, MUX5),},
         .ref = { AVR_IO_REGBIT(ADMUX, REFS0), AVR_IO_REGBIT(ADMUX, REFS1), },
         .ref_values = {
                 [0] = ADC_VREF_VCC, [1] = ADC_VREF_AREF,
@@ -147,17 +148,43 @@
         },
         .bin = AVR_IO_REGBIT(ADCSRB, BIN),
 
-        .muxmode = {
+        .muxmode = { /* 64 entries, including offset calibration */
             [0] = AVR_ADC_SINGLE(0), [1] = AVR_ADC_SINGLE(1),
             [2] = AVR_ADC_SINGLE(2), [3] = AVR_ADC_SINGLE(3),
+            [4] = AVR_ADC_SINGLE(4), [5] = AVR_ADC_SINGLE(5),
+            [6] = AVR_ADC_SINGLE(6), [7] = AVR_ADC_SINGLE(7),
 
-            [ 4] = AVR_ADC_DIFF(2, 2,   1), [ 5] = AVR_ADC_DIFF(2, 2,  20),
-            [ 6] = AVR_ADC_DIFF(2, 3,   1), [ 7] = AVR_ADC_DIFF(2, 3,  20),
-            [ 8] = AVR_ADC_DIFF(0, 0,   1), [ 9] = AVR_ADC_DIFF(0, 0,  20),
-            [10] = AVR_ADC_DIFF(0, 1,   1), [11] = AVR_ADC_DIFF(0, 1,  20),
-            [12] = AVR_ADC_REF(1100),    // Vbg
-            [13] = AVR_ADC_REF(0),        // GND
-            [15] = AVR_ADC_TEMP(),
+            /* no ofs.calibration PA0-PA0, x1 */ [0b100011] = AVR_ADC_DIFF (0, 0, 20), /* ofs. calibration */
+            [0b001000] = AVR_ADC_DIFF (0, 1, 1), [0b001001] = AVR_ADC_DIFF (0, 1, 20),
+            [0b001010] = AVR_ADC_DIFF (0, 3, 1), [0b001011] = AVR_ADC_DIFF (0, 3, 20),
+            [0b101000] = AVR_ADC_DIFF (1, 0, 1), [0b101001] = AVR_ADC_DIFF (1, 0, 20),
+            [0b001100] = AVR_ADC_DIFF (1, 2, 1), [0b001101] = AVR_ADC_DIFF (1, 2, 20),
+            [0b001110] = AVR_ADC_DIFF (1, 3, 1), [0b001111] = AVR_ADC_DIFF (1, 3, 20),
+            [0b101100] = AVR_ADC_DIFF (2, 1, 1), [0b101101] = AVR_ADC_DIFF (2, 1, 20),
+            [0b010000] = AVR_ADC_DIFF (2, 3, 1), [0b010001] = AVR_ADC_DIFF (2, 3, 20),
+            [0b101010] = AVR_ADC_DIFF (3, 0, 1), [0b101011] = AVR_ADC_DIFF (3, 0, 20),
+            [0b101110] = AVR_ADC_DIFF (3, 1, 1), [0b101111] = AVR_ADC_DIFF (3, 1, 20),
+            [0b110000] = AVR_ADC_DIFF (3, 2, 1), [0b110001] = AVR_ADC_DIFF (3, 2, 20),
+            [0b100100] = AVR_ADC_DIFF (3, 3, 1), [0b100101] = AVR_ADC_DIFF (3, 3, 20), /* ofs. calibration */
+            [0b010010] = AVR_ADC_DIFF (3, 4, 1), [0b010011] = AVR_ADC_DIFF (3, 4, 20),
+            [0b010100] = AVR_ADC_DIFF (3, 5, 1), [0b010101] = AVR_ADC_DIFF (3, 5, 20),
+            [0b010110] = AVR_ADC_DIFF (3, 6, 1), [0b010111] = AVR_ADC_DIFF (3, 6, 20),
+            [0b011000] = AVR_ADC_DIFF (3, 7, 1), [0b011001] = AVR_ADC_DIFF (3, 7, 20),
+            [0b110010] = AVR_ADC_DIFF (4, 3, 1), [0b110011] = AVR_ADC_DIFF (4, 3, 20),
+            [0b011010] = AVR_ADC_DIFF (4, 5, 1), [0b011011] = AVR_ADC_DIFF (4, 5, 20),
+            [0b110100] = AVR_ADC_DIFF (5, 3, 1), [0b110101] = AVR_ADC_DIFF (5, 3, 20),
+            [0b111010] = AVR_ADC_DIFF (5, 4, 1), [0b111011] = AVR_ADC_DIFF (5, 4, 20),
+            [0b011100] = AVR_ADC_DIFF (5, 6, 1), [0b011101] = AVR_ADC_DIFF (5, 6, 20),
+            [0b110110] = AVR_ADC_DIFF (6, 3, 1), [0b110111] = AVR_ADC_DIFF (6, 3, 20),
+            [0b111100] = AVR_ADC_DIFF (6, 5, 1), [0b111101] = AVR_ADC_DIFF (6, 5, 20),
+            [0b011110] = AVR_ADC_DIFF (6, 7, 1), [0b011111] = AVR_ADC_DIFF (6, 7, 20),
+            [0b111000] = AVR_ADC_DIFF (7, 3, 1), [0b111001] = AVR_ADC_DIFF (7, 3, 20),
+            [0b111110] = AVR_ADC_DIFF (7, 6, 1), [0b111111] = AVR_ADC_DIFF (7, 6, 20),
+            [0b100110] = AVR_ADC_DIFF (7, 7, 1), [0b100111] = AVR_ADC_DIFF (7, 7, 20), /* ofs. calibration */
+
+            [32] = AVR_ADC_REF(0),       // 0V AGND
+            [33] = AVR_ADC_REF(1100),    // 1.1V internal Vref
+            [34] = AVR_ADC_TEMP(),
         },
 
         .adc = {
@@ -190,7 +217,7 @@
             [AVR_TIMER_COMPA] = {
                 .r_ocr = OCR0A,
                 .com = AVR_IO_REGBITS(TCCR0A, COM0A0, 0x3),
-                .com_pin = AVR_IO_REGBIT(PORTB, 0),
+                .com_pin = AVR_IO_REGBIT(PORTB, 2), /* p.64 */
                 .interrupt = {
                     .enable = AVR_IO_REGBIT(TIMSK0, OCIE0A),
                     .raised = AVR_IO_REGBIT(TIFR0, OCF0A),
@@ -200,7 +227,7 @@
             [AVR_TIMER_COMPB] = {
                 .r_ocr = OCR0B,
                 .com = AVR_IO_REGBITS(TCCR0A, COM0B0, 0x3),
-                .com_pin = AVR_IO_REGBIT(PORTB, 1),
+                .com_pin = AVR_IO_REGBIT(PORTA, 7), /* p.60 */
                 .interrupt = {
                     .enable = AVR_IO_REGBIT(TIMSK0, OCIE0B),
                     .raised = AVR_IO_REGBIT(TIFR0, OCF0B),
@@ -234,7 +261,7 @@
         .r_icrh = ICR1H,
 
         .ices = AVR_IO_REGBIT(TCCR1B, ICES1),
-        .icp = AVR_IO_REGBIT(PORTB, 0),
+        .icp = AVR_IO_REGBIT(PORTA, 7), /* p.62 */
 
         .overflow = {
             .enable = AVR_IO_REGBIT(TIMSK1, TOIE1),
@@ -251,7 +278,7 @@
                 .r_ocr = OCR1AL,
                 .r_ocrh = OCR1AH,    // 16 bits timers have two bytes of it
                 .com = AVR_IO_REGBITS(TCCR1A, COM1A0, 0x3),
-                .com_pin = AVR_IO_REGBIT(PORTB, 1),
+                .com_pin = AVR_IO_REGBIT(PORTA, 6), /* p.62 */
                 .interrupt = {
                     .enable = AVR_IO_REGBIT(TIMSK1, OCIE1A),
                     .raised = AVR_IO_REGBIT(TIFR1, OCF1A),
@@ -262,7 +289,7 @@
                 .r_ocr = OCR1BL,
                 .r_ocrh = OCR1BH,
                 .com = AVR_IO_REGBITS(TCCR1A, COM1B0, 0x3),
-                .com_pin = AVR_IO_REGBIT(PORTB, 2),
+                .com_pin = AVR_IO_REGBIT(PORTA, 5), /* p.61 */
                 .interrupt = {
                     .enable = AVR_IO_REGBIT(TIMSK1, OCIE1B),
                     .raised = AVR_IO_REGBIT(TIFR1, OCF1B),

UPD Yet another issue with at-tiny-24 definition: timers' pins assignment is wrong (compared against both datasheets, for 24th and 24Ath versions). Updated the patch (every correction has related page number of datasheet).

Thanks

[edgargrimberg]

why not a PR?

[djfd]

@edgargrimberg Because never did it/have no time to learn the procedure. Sorry.

UPD But I believe that one day there will be a guy, brave enough to turn the patch into PR ))

[JarrettBillingsley]

I really wish GH would let you turn a patch into a PR without having to fork, pull, apply patch, push, blah blah blah blaaaaaaah

[gatk555]

... a patch into a PR ... PR 439. Trying to learn git, and beginning to like it.

[gatk555]

Nice patch BTW. References to datasheet pages? Whatever next!

[djfd]

@gatk555 Thank you so much, my friend!

What do you mean asking for " References to datasheet pages?", BTW? There is a couple of references for both Attiny24/24A at the very first post, is it not enough? Just checked, they are still working.

Thanks again for your efforts!

[gatk555]

You are welcome! It seems a nifty bit of software, but a pity that it is now maintained so little.

I do not understand your comment beginning "What do you mean by asking ...", but I think it does not refer to anything I did or wrote. All I did was apply the patch, build, commit, push, and then go to the site and make a pull request. It was the quoted page numbers in your patch that made me think I was looking at something good, as there seems to be some rubbish among the orphaned suggested changes.

On Mon, 22 Mar 2021 at 22:11, djfd @.***> wrote:

@gatk555 https://github.com/gatk555 Thank you so much, my friend!

What do you mean asking for " References to datasheet pages?", BTW? There is a couple of references for both Attiny24/24A at the very first post, is it not enough? Just checked, they are still working.

Thanks again for your efforts!

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/buserror/simavr/issues/339#issuecomment-804429548, or unsubscribe https://github.com/notifications/unsubscribe-auth/APAYEBAW7AK3NKWIRDOBQULTE66CNANCNFSM4HF7NKUA .

[gatk555]

But looking again here, I see that I did say that. It was intended as approval of their inclusion in the patch. Further proof that irony does not work on the internet, if any was needed.

[gatk555]

The patch is now integrated, so this issue should be closed.