USB2Dynamixel with no CM5 not working
USB2Dynamixel with no CM5 not working
by clusher
Mon Oct 18, 2010 10:38 pm
Hi all,
It's been a while since my last posts, I've been lost in simulations. However, I am now getting back to the Bioloid and I need to remove the CM5 from the loop, since my laptop is now running the controller and the CM5+batteries only add weight.
I have 3 Lipos (to a total of 12V, more or less) and the USB2Dynamixel. Following the manuals, all I have to do is:
1. Connect the DATA from USB2D to the DATA of AX-12
2. Connect the + from the batteries to the VCC of AX-12 (middle pin)
3. Connect the GND of the USB2D to the - of the batteries and to the GND of the AX-12
With all these connections in place, the AX12 led turns on and off (indicating it has power) and when I sent WRITE commands, everything goes ok (I can move the servo, turn the LED on and off, etc). The problem is that when I do a PING or a READ, I can't get any response.
I checked the wiring and if all the pins are connected where they should be. The voltage between DATA and GND is steady at 5V and between VCC and GND is at 12,3V. I would guess that the problem is somehow connected to the 74HC126D responsible for the TxD of the Ax12, but I have no way to know. There is also the possibility that a pull-up resistance at the DATA line is needed since the CM5 had one and I'm not sure if USB2Dynamixel has one (I couldn't find any schematic on it), although the manuals fail to mention any of this.
Has anyone had this problem or have any idea on what's happening? Thanks
It's been a while since my last posts, I've been lost in simulations. However, I am now getting back to the Bioloid and I need to remove the CM5 from the loop, since my laptop is now running the controller and the CM5+batteries only add weight.
I have 3 Lipos (to a total of 12V, more or less) and the USB2Dynamixel. Following the manuals, all I have to do is:
1. Connect the DATA from USB2D to the DATA of AX-12
2. Connect the + from the batteries to the VCC of AX-12 (middle pin)
3. Connect the GND of the USB2D to the - of the batteries and to the GND of the AX-12
With all these connections in place, the AX12 led turns on and off (indicating it has power) and when I sent WRITE commands, everything goes ok (I can move the servo, turn the LED on and off, etc). The problem is that when I do a PING or a READ, I can't get any response.
I checked the wiring and if all the pins are connected where they should be. The voltage between DATA and GND is steady at 5V and between VCC and GND is at 12,3V. I would guess that the problem is somehow connected to the 74HC126D responsible for the TxD of the Ax12, but I have no way to know. There is also the possibility that a pull-up resistance at the DATA line is needed since the CM5 had one and I'm not sure if USB2Dynamixel has one (I couldn't find any schematic on it), although the manuals fail to mention any of this.
Has anyone had this problem or have any idea on what's happening? Thanks
by i-Bot
Tue Oct 19, 2010 10:56 am
The USB2Dynamixel TTL interface uses this chip:
http://www.fairchildsemi.com/ds/NC%2FNC7WZ241.pdf
The enables are driven from the FTDI TxEn, and the receive passes through the selector switch. Protection diodes and the 47R resistor are used. Also a 10K pullup on the TTL Bus.
I never had problems driving servos direct from USB2Dynamixel.
How long do you wait for the response ? What is your FTDI latency set to ?
http://www.fairchildsemi.com/ds/NC%2FNC7WZ241.pdf
The enables are driven from the FTDI TxEn, and the receive passes through the selector switch. Protection diodes and the 47R resistor are used. Also a 10K pullup on the TTL Bus.
I never had problems driving servos direct from USB2Dynamixel.
How long do you wait for the response ? What is your FTDI latency set to ?
by clusher
Tue Oct 19, 2010 11:10 am
I have no idea about the FTDI latency... actually, I have no idea what that is.
The controller is written in Python and it waits up to 1 second before calling a timeout (using the CM5 I had it at 300ms and I increased it to see if the response was taking too long). The port is configured for 1Mbps, just as it was when the power came from the CM5.
The controller is written in Python and it waits up to 1 second before calling a timeout (using the CM5 I had it at 300ms and I increased it to see if the response was taking too long). The port is configured for 1Mbps, just as it was when the power came from the CM5.
by clusher
Tue Oct 19, 2010 11:30 am
Update: I installed the RoboPlus and ran the Dynamixel manager. I used the Advanced Search to search all baudrates, and it came back empty, no Dynamixels found.
@i-Bot: do you know where I can find the USB2Dynamixel circuit diagram? I need to show it to my supervisor so he sees the pullup resistance...
@i-Bot: do you know where I can find the USB2Dynamixel circuit diagram? I need to show it to my supervisor so he sees the pullup resistance...
by i-Bot
Tue Oct 19, 2010 12:01 pm
The FTDI Latency is usually set to either 1 or 16 millsec, either by software, or in the advanced driver config settings. Should not be a problem if you are waiting longer than this.
You might check the switch is working OK. In TTL mode pin 1 of switch U3 should connect to pin 3 of switch U3.
You might check the switch is working OK. In TTL mode pin 1 of switch U3 should connect to pin 3 of switch U3.
by clusher
Tue Oct 19, 2010 12:16 pm
Just after checking if the 1 and 3 pins of U3 are connected (they are), I decided to go back a step, and I reassembled the Bioloid, connecting the CM5 to the bus just like usual and using the USB2Dynamixel for communication. It didn't work, it happens exactly the same thing: the AX12 obey commands but aren't answering (or the USB2D isn't getting the answers...). I guess all the other posts are now meaningless...
So, like House would say: "New symptom. Differential?". It seems I blew something up in the USB2D? Or can this be something else?
So, like House would say: "New symptom. Differential?". It seems I blew something up in the USB2D? Or can this be something else?
by clusher
Tue Oct 19, 2010 12:50 pm
By now I have tried the following:
-Using the CM5 for both communication and power, everything works fine, all servos are detected and move. So the servos should be alright.
-Using the USB2D for communication and either the CM5 or batteries for power, the servos move but they don't respond. This happens with one servo or multiple servos in the chain.
-Using the CM5 for both communication and power, everything works fine, all servos are detected and move. So the servos should be alright.
-Using the USB2D for communication and either the CM5 or batteries for power, the servos move but they don't respond. This happens with one servo or multiple servos in the chain.
by clusher
Tue Oct 19, 2010 2:59 pm
Well, this RoboPlus software is really practical, I hadn't tried it yet.
As for your questions:
1. By using the search option of the Dynamixel Manager, I only see the RxD led flashing, the TxD doesn't blink. This is done with all the 18 servos and powered by the CM5.
2. By using the RoboPlus Manager, and connecting the USB2D to the CM5 and setting it to use RS232, everything works fine, all devices are recognized and both LEDs are lit.
As for your questions:
1. By using the search option of the Dynamixel Manager, I only see the RxD led flashing, the TxD doesn't blink. This is done with all the 18 servos and powered by the CM5.
2. By using the RoboPlus Manager, and connecting the USB2D to the CM5 and setting it to use RS232, everything works fine, all devices are recognized and both LEDs are lit.