ATtiny85 PWM on D4

I am afraid, I found a bug in ATtiny85…?!?
By happenstance I found a circuit of a cross-fade blinker, created for railroad crossing of a model railway. The producer develope and tested this on the original ATtiny85-chip. He uses D0 and D4 and this works fine.
As I reproduce the circuit in SimulIDE, D0 works perfect, but not D4…?!?
I try several actual releases, but with the same result. When I try this with D0 and D1, it works.

I think, a little mistake has slipped in?!?

Thanks, I will have a look.

Your code is working for me if I compile it (ATTiny85_pin04_no_PWM.ino).

Are you sure you are compiling it for AtTiny85?

I use a standard installation of Arduino 1.8.19 on ubuntu 18.04

my settings:
board: ATtiny25/45/85
prozessor: ATtiny85
clock: internal 1MHz

different projects run perfect with this settings…
Please correct me if I am wrong!

Weird...
If I run ATTiny85_pin04_no_PWM.sim1, which uses ATTiny85_pin04_no_PWM/ATTiny85_pin04_no_PWM.ino.tiny8.hex
OCR1C is always 0 and PWM at PORTB4 does not work.

If I compile ATTiny85_pin04_no_PWM/ATTiny85_pin04_no_PWM.ino for Attiny85 then OCR1C = 255 and PWM at PORTB4 works as expected.
Hex file attached.

here the left led seems to be continuous ON – didn't fade. Look at the screenshot.

But now, when I use your hex-file, it fades – slowly, but it fades! When I change the setting from 1MHz to 8MHz, your  hex-file works perfect!
Please excuse me! Apparently it is my mistake…?!? :-/

Now I tried to compile with ATtinyCore (8MHz/16MHz) and it works!

Did anyone know:
is it wrong to use PWM with 1MHz…???

is it wrong to use PWM with 1MHz…???

In theory you can use PWM with any frequency, but who knows what these libraries are doing.

Hello !

I'm also stuck with my pwm exploration and learning !
I cannot make even a simple sketch to work :-(

here is my minimal test code the sim file is attached.

void setup() {
   // Minimal attiny timer1 setup for PWM exploration
   // set PB1 (OC1A) as output
      DDRB |= (1 << DDB1);
      TCCR1=0;
      OCR1A = 127; // Set Duty cycle baseline half OCR1C for 50%
      OCR1C = 255; // Set Ouput frequency must be calculate as needed
      // Timer settings for PWM
      TCCR1 = (1 << PWM1A) | (1 << COM1A1) | (1 << CS10);
}

// the loop function runs over and over again forever
void loop() {
}

I use the  1.0.0-SR0 R1320 version of SimulIDE under Windows 11.
This code works perfectly on a real attiny85.

Does it possible to test PWM configuration with SimulIDE or is it really a confirmed bug ?

regards

I think you need to compile your project for ATtiny85.
The simplest way todo it is using Arduino IDE.

Or have a look at this discussion: https://simulide.forumotion.com/t323-attiny85-load-program

arcachofo wrote:I think you need to compile your project for ATtiny85.
The simplest way todo it is using Arduino IDE.

Or have a look at this discussion: https://simulide.forumotion.com/t323-attiny85-load-program

Hello !

In fact i've already tried this and used the hex file produced by arduino IDE. it gave me the same result !

But i've perhaps missed something ! So i will read the thread you suggest, give arduino IDE an other try and tell you back !

Regards

arcachofo wrote:I think you need to compile your project for ATtiny85.
The simplest way todo it is using Arduino IDE.

Or have a look at this discussion: https://simulide.forumotion.com/t323-attiny85-load-program

Hi,

I've try many ways as you suggest but with no results with 1.0.0
So i give 0.4.15 a try ... and its a bit better but not really fine  

With 1.0.0 i could setup CTC mode but no way to toggle PB1 despite the fact i set the COM1Ax and others bits correctly.

with 0.4.15 the PB1 toggle but it's not possible to set OCR1A and OCR1C. Setting differents values do not change the frequency and the duty cycle still freeze at 50%. Only the prescaler adjustement have an effect on the frequency, for sure !

I not able to determine exactly what'is going wrong.

I've notice that the SRAM adress of registers is shifted by 22(decimal) in the MCU monitor as all the other registers even if the "C:\SimulIDE_0.4.15\share\simulide\data\avr\attiny85.data" file contains   the non shifted definition  OCR1A EQU 46 -> 0x2E OCR1C EQU 45 -> 0x2D wich are the values we can found in the datasheet ...
In MCU monitor it's clearly 0x4D and 0x4E i've also wrote addresses and values of OCR1A and OCR1C in EEPROM to confirm.

EEPROM.write(0, &OCR1A);
EEPROM.write(1, OCR1A);
EEPROM.write(2, ocr1aAddress);
value = *ocr1aAddress; //0x2E
EEPROM.write(3, value);
 EEPROM.write(4, ocr1aAddress2);
value = *ocr1aAddress2; //0x4E
EEPROM.write(5, value);

EEPROM.write(6, 0xFF);
EEPROM.write(7, 0xFF);

EEPROM.write(8, &OCR1C);
EEPROM.write(9, OCR1C);
EEPROM.write(10, ocr1cAddress);
value = *ocr1cAddress; //0x2E
EEPROM.write(11, value);
EEPROM.write(12, ocr1cAddress2);
value = *ocr1cAddress2; //0x4E
EEPROM.write(13, value);



Do you think it's a direction ? IMHO there is some trouble anyway with timer1 simulation in both versions, dont you think so ?

Regards

I didn't realize you are not using the last build for 1.0.0.
Try with the last one available here:
https://simulide.forumotion.com/t390-simulide-1-0-0-tester-builds#1936

arcachofo wrote:I didn't realize you are not using the last build for 1.0.0.
Try with the last one available here:
https://simulide.forumotion.com/t390-simulide-1-0-0-tester-builds#1936

Dear arcachofo,

You are the champion

I didn't have much time to do extensive testing but, with this version, it works

The output frequencies are now accurate based on the timer, prescaler and duty cycle register settings.

Thanks a lot for your time and work!
Since I discovered SimulIDE, I have been looking forward to it.
It is a very nice tool, for education and personal research, at almost no cost. Keep pushing!

Thanks again, best regards

Glad to know that it is working now.

And sorry for not putting enough attention, I was convinced that you were using the last build.