COM ports on RB110

[goodma]

I am now trying to directly control Kondo's servos (KRS-2552HV) using the COM ports of RB110, has anyone got the idea of the following questions?

1. I've tested the TTL level of COM3, COM4 and COM5, they are 3.3v; but are they 5v input tolerant? In other word can they accept 5v logic input without any problem? For the output, we have to boost 3.3v to 5v of course to communication with a 5v divice.

2. COM6 is a 10-pin general serial port, but how can I use it as another TTL port (such as COM5) please? I know I can use a RS232-TTL converter but is there a better way, such as using a special driver from FTDI?

Cheers!

[PedroR]

Hi gooma

We've done some projects with Linux and KHR 3 here on RoboSavvy and there are a few things to be aware of:

- With regards to Signal Level conversion you can use something like a Logic Level converter http://robosavvy.co.uk/store/product_in ... cts_id/461

The Roboard should support 5V TTL logic input, but I'm not sure if the output is 5V or 3.3V. If KHR is 3.3V you may need to use this logic level converter.


In addition, and most importantly, the Logic Signals on KHR 3HV are inverted:
"invert the logic of the signal to connect the UART on your microcontroller. RCB-4 uses 5V logic, idle low"

You can read more about it here http://robosavvy.com/forum/viewtopic.php?t=6543#28111

To ease thing up what we did in the was use the Kondo USB to Serial adapter (removed the plastic case to save space) because the adapter already does all the necessary level shiftings.

Regards
Pedro.

[goodma]

Hey Pedro,
Thanks for your prompt help! I think I've understood the logic system of RCB-4, see below:

Code: Select all

[Original USB]o---1---o[RCB-4]o---2---o[SIO1-4]
                                             o ---3---o[SIO5-8]


RCB-4 has 3 serial ports. COM1 communicates with the original FDTI USB with inverted 3.3v logic. ``Hidden'' COM2 and COM3 control two groups of servos with 5v logic. We have managed to control RCB-4 without problem (using 7404). I am now thinking about a more challenging task --- to completely get rid of RCB-4 and directly control all the servos, like this:

Code: Select all

[RB110]o---COM5---o[SIO1-4]
            o---COM6---o[SIO5-8]


The Roboard should support 5V TTL logic input, but I'm not sure if the output is 5V or 3.3V.

It's output is 3.3v, if it can support 5v logic input that would be great --- I can simply use a single 7404 to boost its 3.3v output to 5v for both COM5 and COM6, so servos can understand it. Otherwise I need to use the ``Logic Level converter'' as you mentioned --- it requires more physical space as it needs both 5v and 3.3v VCC from RB110. More information about logic level conversion between 3.3v and 5v can be found in the following links.

http://focus.ti.com/lit/ds/symlink/txb0104.pdf
http://www.maxim-ic.com/app-notes/index.mvp/id/3007
http://www.microchip.com/stellent/group ... 026368.pdf
Cheers,
Jie

[PedroR]

Hi googma

We haven't gotten as far as controlling each chain of servos individually. We chose to retain the RCB 4 as it helps with playing motions and walk staibilization with the help oft he KRG and RAS sensors.

In our project we were developping a vision system for KHR 3HV so we needed Linux for Vison+Webcam. We focused on that and left the servo control to the RCB 4.

With this I mean to say I can't be of much assistance to your specific question about controlling the servo chains individually as we haven't done that.

Having said this, if you have access to a Data Analyser you could hook it up between the RCB 4 and the servo chain and see what's going on in terms of signals.

As for the Roboard taking 5V TTL input, I _think_ it can do it. It is able to interface with Bioloid servos which I believe are 5V.
They may be using the trick of putting a Diode on the RX line to lower signal level and then do the TX at 3.3V which is enough to be seen as a "1" by a 5V TTL device but I am not sure.

About this 3.3V/5V issue, I believe forum member "roboard" (which are Roboard engineers) will be the ones who can give you a definitive answer about it.

Regards
Pedro.

[i-Bot]

Sorry, I'm a bit rushed and on the move at the moment.

Suggest you post topic across under KHR too, so Chrisvo can comment.

I can help some, but not as well as Chris or Roboard.

[goodma]

Suggest you post topic across under KHR too, so Chrisvo can comment.

Thanks for your advice, I've posted just the topic under KHR too, and some updates:

I've managed to directly control a single 3HV servo without any logic conversion but the risk is more servos will pull down the 3.3v logic to about 2.5v so it may not always work while the robot is running. Another problem is if I short TX and RX from the RB110 side, it will always receieve the message from itself, I know 74125 can solve this problem but COM3-COM5 don't come with the TXDEN pins.

Jie

[roboard]

Hi, goodma,

RoBoard's COM3 & COM4 are natively 5V-tolerant (please see the COM3/COM4 circuit in RoBoard H/W introduction slide).

We have no a KONDO 3HV to try the "3.3V pulled-down to 2.5V" issue. But we have connected a Bioloid of 16 AX-12 servos to RoBoard's COM3 and the output voltage of COM3 still works normally.

When you short COM3's TX/RX, COM3 should always receive messages from itself; but you just need to first read and discard such messages and then you can still read the messages returned from the servo. (In fact, KONDO's PC Servo Manager softwares exactly do the same thing.)

If you really need TXDEN, you can employ RB-110's FTDI COM6. COM6 is a TTL port, not a RS232 port.

[goodma]

Hey roboard,

Thank you so much for your prompt help! I will try to discard the messages sent by itself and see whether the servos can response normally. The issue ''3.3v step down to 2.5'' may be because we used a 1.2A adapter for multiple servos and the board at the same time -- I will also try to use a li-po battery instead.

Cheers,
Jie