[zhiwarwadi]

I calculate motor tuning more time


Sent from my iPhone using Tapatalk

[zhiwarwadi]

I changed to 25kh to 35 to all but same issues


Sent from my iPhone using Tapatalk

[fernando123]

Check the feed rate for the material you are trying to cut. It's a little complex to explain. But tomorrow I will have more time and I will try to help. But you probably trying to cut to fast through the material. Must check the tool bit, the feed rate, the material , the motors torque etc..

Enviado de meu SM-G530H usando Tapatalk

[zhiwarwadi]

this issues before cutting just in tast


Sent from my iPhone using Tapatalk

[zhiwarwadi]

thanks


Sent from my iPhone using Tapatalk

[mcm_xyz]

Boys...

You are totally out of topic here... Please post in this thread only related things... You will make this thread hard to read for those interested in this project...

Thanks
Mihai

[nitrochopper2]

Boys...

You are totally out of topic here... Please post in this thread only related things... You will make this thread hard to read for those interested in this project...

Thanks
Mihai

Thank you so much for sharing Mihai. I have 5 boards in progress based on the 644p. It would be awesome if somebody with some time and mikropascal could compile the code for the 328p and list the corresponding pinout. Could result in more interest...

[Khalid]

I am also interested in 328P version of ATMEGA.

Thank you so much for sharing Mihai. I have 5 boards in progress based on the 644p. It would be awesome if somebody with some time and mikropascal could compile the code for the 328p and list the corresponding pinout. Could result in more interest...

[mcm_xyz]

I am also interested in 328P version of ATMEGA.

I've looked a little over the datasheet for the 328P MCU. It's ok for the design. The only problem I see is the low memory (2K SRAM) that will not allow a very big commutation table to be allocated. (The table storing the commutation shift for each encoder step). This being said the circuit will only allow lets say 1024PPR max encoders, or if bigger used, a precision decimation is needed. Or we could drop the commutation table, and use in realtime calculation of the commutation, sacrificing in performance...

Thus being said, what we would need is:

1. at least 2 external interupts
2. at least 16K of Flash
3. at least 2K of RAM
4. at least 6 PWM outputs (it would be nice to have them all 16 bit)
5. at least 1 ADC input
6. hardware UART is nice but not really a must
7. lowest price.

ATMEGA 328 has all of these... I could try and modify the design and code for it but it would be later this year, in autumn I hope...

Mihai

PS: I just discovered that int 1 is shared with a PWM Output (PCINT19/OC2B/INT1) PD3... This is not so good for our design, We will have to use a PCINT for the second interupt = performance issues...

[nitrochopper2]

I've looked a little over the datasheet for the 328P MCU. It's ok for the design. The only problem I see is the low memory (2K SRAM) that will not allow a very big commutation table to be allocated. (The table storing the commutation shift for each encoder step). This being said the circuit will only allow lets say 1024PPR max encoders, or if bigger used, a precision decimation is needed. Or we could drop the commutation table, and use in realtime calculation of the commutation, sacrificing in performance...

Thus being said, what we would need is:

1. at least 2 external interupts
2. at least 16K of Flash
3. at least 2K of RAM
4. at least 6 PWM outputs (it would be nice to have them all 16 bit)
5. at least 1 ADC input
6. hardware UART is nice but not really a must
7. lowest price.

ATMEGA 328 has all of these... I could try and modify the design and code for it but it would be later this year, in autumn I hope...

Mihai

PS: I just discovered that int 1 is shared with a PWM Output (PCINT19/OC2B/INT1) PD3... This is not so good for our design, We will have to use a PCINT for the second interupt = performance issues...

Thank you for catching that..saved me some time. I will stick w the 644p. They are on the way already.

Hmmm...how about code for a 1284p? More common probably. Wait..or is the code already compatible w the 1284p? I have one of those.

[mcm_xyz]

Thank you for catching that..saved me some time. I will stick w the 644p. They are on the way already.

Hmmm...how about code for a 1284p? More common probably. Wait..or is the code already compatible w the 1284p? I have one of those.

Hello.
From what I know so far ATMega1284 has 2 counters on 16 bit so a very little code modification is requred (I think) for PWM initliazition. That should be all. I also have a few of those and I might give it a try.

Mihai

PS. 1284 has also a great bunch of SRAM(16K) so further code optimizations I think are also possible.
PS. Code is fully compliant with 1284. 1284 has another counter (OCR3) that is on 16 bit but it's not used. (all you have to do is open the project and change MCU type and then hit compile)

[nitrochopper2]

Hello.
From what I know so far ATMega1284 has 2 counters on 16 bit so a very little code modification is requred (I think) for PWM initliazition. That should be all. I also have a few of those and I might give it a try.

Mihai

PS. 1284 has also a great bunch of SRAM(16K) so further code optimizations I think are also possible.
PS. Code is fully compliant with 1284. 1284 has another counter (OCR3) that is on 16 bit but it's not used. (all you have to do is open the project and change MCU type and then hit compile)

The code is too big for the free version of Mikropascal..and i dont want to spring for a license yet. I might though if i can get these drives to work on some bigger motors.

[Khalid]

Thanks for detail reply: We have an option of Arduino Mega chip...
https://www.arduino.cc/en/Main/arduinoBoardMega2560

Thanks

[mcm_xyz]

First cut with cooling liquid!!!

I can tell you that the servos work perfectly, movement is very precise and accurate. What you see in the movie is a cut with 1000mm/minute feedrate, depth per pass 0.5mm, spindle speed 15000RPM, single flute spindle, 6082 alu.

https://www.youtube.com/watch?v=_mpDGQWrO0w

[rene-dev]

Hi,
I am working on a similar project, but targeted at 400V servos with resolvers. But sin/cos, incremental and absolute encoders are also supported.
https://github.com/rene-dev/stmbl

Rene

[fernando123]

Hi,
I am working on a similar project, but targeted at 400V servos with resolvers. But sin/cos, incremental and absolute encoders are also supported.
https://github.com/rene-dev/stmbl

Rene

I have 3 danfoss servo drives . It is also 400Vdc. Any ideia how to control them.?

Enviado de meu SM-G530H usando Tapatalk

[mcm_xyz]

Hi,
I am working on a similar project, but targeted at 400V servos with resolvers. But sin/cos, incremental and absolute encoders are also supported.
https://github.com/rene-dev/stmbl

Rene

Excellent work! Finally someone is doing something for the community. You project has a bright future, I'm sure. People have been asking me about changing MCU with a STM32, but I have no experience with those... My servo is at a very hobbist level. Your looks very close to commercial ones... Keep up the good work and thanks for sharing...

Mihai

[punisher454]

People have been asking me about changing MCU with a STM32, but I have no experience with those... My servo is at a very hobbist level.

I just recently bought some little STM32 boards, for around 5 bucks each on Ebay. There is a group porting Arduino to STM32.

Arduino for STM32 - Index page

I'm just getting started with it, but its not too much more complicated to set up than the regular AVR based Arduino. I bought 2 programmers, one was a genuine STLink for around $20.00 and the other was a little thumbdrive style STLink knock-off for about 5 or 6 Dollars I think. Both work Fine.
Basically ARM development is now at least as easy and inexpensive as AVR development was just a few years ago.

[rene-dev]

You should seriously have a look at arm micro controllers, life is too short to struggle with too little memory or flash, we are living in 2015, 8 bit CPUs are 80s technology.
Still good for controlling a washing machine, but not for serious projects like yours.
The stm32f4 has a 32 bit cortex m4 core, with FPU and DSP, is running at 168 Mhz.
And you geht a load of periphery(16 timers, 2 DMA controllers, 9 UARTs, USB, ...) which takes care of all the realtime stuff, like reading the encoder/resolver/uart, counting clk/dir.
I do not see much point in porting any Arduino stuff, because in my experience from AVRs it only makes life more difficult in larger projects.

I think I will open a thread for my project as well, so people can discuss stuff and make suggestions.
At the moment there is only the stuff on Github, and a thread in a german cnc forum.

Rene

[Dragonfly]

Hello Mihai, from the near neighborhood
I am in Bulgaria.
Haven't been in a while on the zone ahd have missed this excellent thread. I've been thinking recently about using a DC brushed motor as a servo motor and I saw some comments from you how it can be done. I am working with Atmel uCU's for my different projects but my favorite environment is AVR Studio + WinAVR and writing the code in C. Once upon a time in the eighties and nineties of the past century Pascal was my favorite language but haven't used it for a very long time.
Is there some new development with your project? I'd very much like to build one driver even if I have to buy a BLDC motor.
Regards,
Ivan