Thanos AMC1.6 USB 2DOF Motion Controller

[vicpopo]

Hi Thanos,
I tested one more time the amc1.6 v1.1 version f to compare to version e.
The material and settings was the same ( that was with asynchronus + inverter video above in my last post)
The verion e hasn't reseted
The version f has twice reseted,but after this 2 resets I continued playing without resets .
The problem was when the board reseted the motor overshoot its position and made a full revolution.
The firmware f not 100% stable in this try!
Update 13-09-2012

I made yesterday a new try with amc1.6 in SFGT mode during 15min and no board reset and no motor overshooting .I seted up the % dead zone to 7%.If someone else could test the firmware amc1.6 f and post his feedback it would be great!

I didn't match having 90° for angle course, but very usefull with pot 3 for choosing what is the best angle setup fot the simulator.
One more thing I noticed.I connected a voltmeter at the setpoint output after the RC filter.I would like just knowing the setpoint range sended to the inveter .It's ok 0 10v.The only point surprising me is when you start xsim for playing the first setpoint done (I suppose for reaching the middle position of the simulator ) is with full power (10volts).
Is it really necessary at the beginning to reach the middle postioning with full power?
I remember the h-bridge I burned with dc motor!I understand too that if the h-bridge can't assume the first positionning with full power they can't either when playing!
Regards

[fooons]

Hi friends,

First of all, thanks to Thanos and vicpopo for the information about encoders. I will test the encoder with old 1.5 board with alternate firmware (I think there is a encoder version) because I have now a electro cilinder with built in encoder to test.

I want to post some pictures and video about this hardware and I am also working with a Bonfiglioli AGILE Vectron inverter and Omron J1000 inverter.

And I will connect the AMC to the inverter with an PWM to analog voltage converter 0V to 10V. I think It's a LM358N chip.

[vicpopo]

Hi fooons ,

Yes it's this chip .I give you below the files made by Thanos to convert the pwm to 0 10 v.
I ziped with Thanos pcb RC filter a pcb made by my friend to adapt the 0 5 volt P1D1 etc P2D2 output to 0 -24 volt if your inverter needs this signals for rotation directions.It needs a 24 v PSU or you can try with the inverter PSU (24 v) if it's possible.I choose a little external PSU in my case.

For the pcb I made I give any warranty because I've not yet tested this pcb.You can find a file in the archive named untituled , it's the pcb made by hand from my friend.I decided to couple the both circuit on one pcb (RC filter + adapt 0 24 ) just to simplify .

The pcb is not full tested but the solution works because all the tests I made I used the little PCB made by hand by my friend.I hope I didn't make any mistake during the compilation of both.

I put too the link for reading this pcb . If I made a mistake give me feedbak please.

Regards

10VOLT.zip

pcb files

(123.1 KiB) Downloaded 820 times

The link for reading my pcb free sotware : http://b.urbani.free.fr/pagetci/tci.htm

[vicpopo]

Hi Thanos,
I tested one more time the amc1.6 v1.1 version f to compare to version e.
The material and settings was the same ( that was with asynchronus + inverter video above in my last post)
The verion e hasn't reseted
The version f has twice reseted,but after this 2 resets I continued playing without resets .
The problem was when the board reseted the motor overshoot its position and made a full revolution.
The firmware f not 100% stable !
I didn't match having 90° for angle course, but very usefull with pot 3 for choosing what is the best angle setup fot the simulator.
One more thing I noticed.I connected a voltmeter at the setpoint output after the RC filter.I would like just knowing the setpoint range sended to the inveter .It's ok 0 10v.The only point surprising me is when you start xsim for playing the first setpoint done (I suppose for reaching the middle position of the simulator ) is with full power (10volts).
Is it really necessary at the beginning to reach the middle postioning with full power?
I remember the h-bridge I burned with dc motor!I understand too that if the h-bridge can't assume the first positionning with full power they can't either when playing!
Regards

I forgotten to say that I tested the firmware amc1.6 v1.1 f in SFGT mode and it works.I didn't test the SFGT mode in firmware version e !
I will test too !
Regards

[tronicgr]

One more thing I noticed.I connected a voltmeter at the setpoint output after the RC filter.I would like just knowing the setpoint range sended to the inveter .It's ok 0 10v.The only point surprising me is when you start xsim for playing the first setpoint done (I suppose for reaching the middle position of the simulator ) is with full power (10volts).
Is it really necessary at the beginning to reach the middle postioning with full power?
I remember the h-bridge I burned with dc motor!I understand too that if the h-bridge can't assume the first positionning with full power they can't either when playing!
Regards

The way the PID algorithm works it needs to try to find the balance point (target position) as fast is possible even if it doesn't receive any actual motion data just center position during power up.

To make it go with constant low speed to the initial position during start up requires a bit complicate driving logic that involves measuring the actual motor's speed and accurate timing. But it will not work always as it should because of all the different potentiometer or other position feedback devices.

All in all I could implement a slow ramping speed motion during startup for the motor to make his way to the center position if there are no incoming serial data available.


Thanos

[fooons]

Hi,

Thanks for the schematics vicpopo, I have been a bit occupied and I can't be able to post. Now I have some time and I want to test this voltage adapters.

I also have one layout from a friend of other forum called yobuddy. When I receive components I'll do the necesari adaptations in my hardware to know what is the best solution for the inverters I have. This is the wiring diagram for the AC inverters:

The image is based on a AMC 1.5 EZ series that yobuddy did, take care of use PWM1, PWM2 and GND pins in your boards

AC_Motor_Hookup.jpg (292.78 KiB) Viewed 5906 times

PWM to Analog~EZ.jpg (411.13 KiB) Viewed 5906 times

There is also the layout in PDF format and component list ready for digikey order.

PWM to Analog~EZ_Layout.pdf

Layouts of PCB by yobuddy

(6.91 KiB) Downloaded 800 times

PWM to Analog~EZ - BOM.csv

(965 Bytes) Downloaded 762 times

[speedy]

Hi fooons :
you have to connect an opt-coupler circuit before the PWM-analog circuit to isolate the analog side of the inverter from the digital and logical side ... because the inverter is creating frequencies that considered a high noise that could destroy your ATMEGA & firmware program .

ATmega1284p may be overclocking in-tolerate.
In that case you might have to use different crystal speed. Here is a compiled version for ATmega1284p and crystal 18.432Mhz:

FTISP for AMC1.6 1.1f atmega1284p 18.432Mhz.zip

Try it and let me *
Thanos

Hi Thanos :
Actually I'm trying to follow you on the AMC1.6 but in fact I can't find the ATMEGA644 and every thing else is ok ...I'm also not able to get the ATMEGA1284P ... can I use the ATMEGA8535 with it's latest firmware on the AMC1.6 board ? ... If not ... then what else can I use ? ... Sorry it's little complicated .
thanks in advance .

speedy

[ne02005]

Hi guys, I build my new AMC 1.6, I coud program it OK!!!! just in seconds...

BUT

When I press "START" on X-sim the board reboots and show strange characters... what can it be?

Anybody has the same problem?
I´m testing with Xsim 3 and another old Xsim.

Hi,

Your ATmega1284p may be overclocking in-tolerate.

In that case you might have to use different crystal speed. Here is a compiled version for ATmega1284p and crystal 18.432Mhz:

FTISP for AMC1.6 1.1f atmega1284p 18.432Mhz.zip

Try it and let me *

Thanos

Ok I´ll try it today!!! I didn´t see this post! Thanks!!!!!
Is driving me crazy.....

Thanks Thanos you are the best!

[tronicgr]

ATmega1284p may be overclocking in-tolerate.
In that case you might have to use different crystal speed. Here is a compiled version for ATmega1284p and crystal 18.432Mhz:

FTISP for AMC1.6 1.1f atmega1284p 18.432Mhz.zip

Try it and let me *
Thanos

Hi Thanos :
Actually I'm trying to follow you on the AMC1.6 but in fact I can't find the ATMEGA644 and every thing else is ok ...I'm also not able to get the ATMEGA1284P ... can I use the ATMEGA8535 with it's latest firmware on the AMC1.6 board ? ... If not ... then what else can I use ? ... Sorry it's little complicated .
thanks in advance .

speedy

Hi Speedy,

I checked the supported devices in the FTDI programmer and its there. You can try it I guess but you have to use the above version for 18.432Mhz and pray that the ATmega8535 you got can handle overclocking! Also mind that the firmware takes almost 80% of the ATmega8535 memory space. It may not fit future additions in the firmware...

Thanos

[tronicgr]

Hi,

I also have one layout from a friend of other forum called yobuddy. When I receive components I'll do the necesari adaptations in my hardware to know what is the best solution for the inverters I have. This is the wiring diagram for the AC inverters:

The image is based on a AMC 1.5 EZ series that yobuddy did, take care of use PWM1, PWM2 and GND pins in your boards

AC_Motor_Hookup.jpg

PWM to Analog~EZ.jpg

There is also the layout in PDF format and component list ready for digikey order.

PWM to Analog~EZ_Layout.pdf

PWM to Analog~EZ - BOM.csv

Actually that schematic is wrong. Its intended for use with Uni-directional drive modes, that the PWM signal drives both speed and direction with center being at 2.5v. Using Bi-directional control method is more safer as the motor will be 0% speed if for some reason the AC drive is disconnected from the AMC1.6.

Thanos