Need Help regarding BlueSmirf and RCB3.

[ryann2k1]

It seems, my BlueSmirf just works for a while. After several motions, it does not work anymore. Even changing to the sleep mode.
I wonder to know why? I have changed the chip (7414/schmit trigger).
Anyone has the same problem?
My project is stuck again.. oh..no..


Cheers

ryann2k1

[PaulP]

H2H relies on responses being sent back in a very specific time.

If you get a slight delay in the transmission from the bot to the PC, it will stop sending. It can happen early or late in the sequence, anywhere in fact.

[PaulP]

What do you think this board which I found through google?

http://www.bdmicro.com/mavric-iib/

That is an extremely large board as its meant for Development

I've gone for this one

http://www.siphec.com/microcontroller/i ... m128-TB1.2

It'll need a few bits built but its extremely small for an Atmega128

[ryann2k1]

H2H relies on responses being sent back in a very specific time.

If you get a slight delay in the transmission from the bot to the PC, it will stop sending. It can happen early or late in the sequence, anywhere in fact.

Thank you PaulP.
I have set up manually the baudrate of my computer com port from 9600 to 115200. Would it be this causes the problem?
How do you manage this problem and does your robot perform the motions you send properly?

And about the ATMega board, it seems really light and handy.

Cheers

ryann2k1

[Kochiro]

Paul,
I've been reading this post with much interest as I was thinking about doing the same sort of thing. I'm wondering why you didn't go with something like this though.

http://www.lynxmotion.com/Product.aspx?productID=252&CategoryID=66

It's the Mini Atom Bot Board. With it you can control up to 20 servos as well as getting onboard programming of either a Stamp or Atom chip as well as other options like the PS2 controller port.

Thanks,
Jay

[shsan]

Just a quick check at the documentation and it says that the board cannot accept more than 7.2V for the servos (So won't do for the KHR2/1-HV) same limitation regarding the other input max 9V.

So unless someone wants to bypass the power supply for the servo and add an extra voltage regulator the board won't do.

shsan

[PaulP]

Unfortunately you are going to have to regulate the voltage on most MCU/CPU you use. Fortunately the SSC-32 has a 5V output you can use.

The Atom board is good but has no ADC channels. It is limited to Atom Basic as well. I need at least 8 ADC inputs which will not tie up outputs.
I also need 2 UARTS so that I can receive data from the PC, make decisions and then pass commands to the SSC. Also the SSC runs 32 servos.

The Atmega128 has all that and more. Ity is also cheaper and can be used with many programming languages & IDEs including Atmels own FREE studio.

With the upgrades that Laurent has now to the SSC. Sequences stored in onboard EEPROM, socket control and other things the SSC is a hard board to beat. It also has 4 ADC channels.

The Atom board has a true RS232 level UART so it would need a maxim 232 to convert to TTL level whereas both the ATMEL and SSC can plug into the BlueSmirf without any mods or level shifting.

And as Shsan says, you can power the whole thing off 10.8V. The SSC will power the servos at 10.8, its own logic at 5V and still give you 5V output to use on your BlueSmirf. I've not checked but I may be able to run the Atmel off it as well.

[Kochiro]

Thanks shsan and Paul. I hadn't really delved into the specs of the board yet, I was just looking at it's general features.

The points you brought up all make sense, but there are a few that I'm going to have to read up on before I fully understand them.

Thanks,
Jay

[PaulP]

I know what I have written may appear that I am biased towards the particular choices I have made but don't misunderstand. There are hundreds of ways to achieve it. What I have done is a reasonable approach that hasnt been without problems. This thread for one, the servo's failing and other minor issues.

The Atmega is a good choice but I only chose it because it was such an obvious option. Originally I had intended to go with one of the RabbitSemiconductors product but my confidenc in my ability to program in C and the fact that no-one else has yet gone that one slightly put me off. Other very good options are PIC based.

Its a bit of a cop-out but it worked for me.

If you have the nerve to experiment and the resources then some of the options Ive read about are amazing and I only aspire to achieve what these guys are doing. Mini ITX, Linux on board, Derek Zahns thread is stunning material to read

And as to wireless control, there are now so many options available that its almost impossible to keep up with the speed at which they change. I wanted to go Serial-Lan with mine but I dont know enough about I.P. and the fact that so few applications can interface to it meant I would be missing out on stuff.

So please dont interpret what Im saying as 'this is the way to do it'. Its an option that works.

Kochiro you are doing exactly the right thing. I nearly made an expensive dratic mistake. I was looking at one board and it seemed to be perfect. I had an order in place and everything. Then I noticed the size of the thing. It was way too big. Massive in fact. Luckily the company was very understanding and cancelled the order at no charge.

Wow thats a long waffle...

[shsan]

I must be getting old quicker than I realized.

Today after reading some docs about the RM-BOARD which can be connected to the RCB3, I suddenly realized that the RCB3 serial connection was in fact RS232C compatible.

Which means, that a simple MAX232 chip will do for the convertion or even better a DS275 (which doesnt need the capacitors).

I had a HIN232 already mounted on a board and ready to use from an old experiment I made, so I soldered the cables and tried it with the RCB3 and BlueSmirf.

And as I thought, it worked straight away. It DOES NOT solve the timing issues but it's working...

Anyway I'll reuse that for my next experiment.

[Bob]

I have been looking for a wireless solution for a while. I tried the Blue Smirf, but as everyone discovered eventually bluetooth won't work. I am now in the process of trying it with zigbee, using the XBee board. So far I have found a nice USB Serial dongle here: http://www.newmicros.com
and a nice breakout board from SFE, and I have the two communicating at 115000bps.

I have also found a nice 3.3 v switching regulator here: http://www.hvwtech.com/products_view.asp?ProductID=583

So far so good. Now, does anybody know if the serial levels need to be inverted like they do on the blue smirf? And if they do, does anybody have a nice miniture signal inverter that can be integrated into the package easily?