Sorry if I make this place start sounding like a broken record, but at least my situation adds a new twist:
I have an Atom Pro 28 module, in a Mini-ABB.
I am unable to get the module to properly identify or program from my notebook computer using a USB-to-serial adapter…
BUT!
…I am able to get a response and program it properly from my desktop machine, using a ‘real’ RS-232 serial port.
Sounds like a problem with the USB-to-serial adapter.
That was my thought as well, but despite attempts with three different units, updating, uninstalling, and reinstalling the drivers, etc., still no joy. Changing the ‘reset hold time’ variable did nothing to rectify the situation, either. The three USB-to-serial adapters used were a BAFO BF-810, a Belkin F5U409, and some other brand which I can’t remember and have since loaned out to a friend, all of which work just fine for programming a BASIC stamp on the same notebook computer. (Edit: The third adapter was from Dynex - still no joy.)
As stated above, I can still program the module properly from my desktop machine, but not being able to blow code into the Atom Pro on my lunch hour at work is seriously slowing down my development time.
I have also had problems getting different USB to serial cables working. I am an FTDI based one right now. I don’t remember the brand, but I know I posted it up here earlier. It has Blue tips and the wire has a clear plastic on the outside where you can see the shielding below… It works resonably well, although very slowly. My home PC with a hardware serial port reprograms the Atom or PRO probably 2 to 3 times faster. (Note this USB serial device can have baud rates up to 8 time faster then the BAFO)
Earlier when I was having problems, I also purchased a BAFO 810. It arrived and it also did not work. My guess is that some earlier USB to serial device that was also Prolific based may have corrupted something. I would be interesting to see what AcidTechs driver details shows for his device.
My shows SER2PL.SYS v 2.0.21 - Prolific, Microsoft hardware comp…
SERENUM.SYS v5.1.2600.2180(…) Microsoft
Hey Kurte, if you are certain it is an FTDI based converter you may want to check out their website and see about adjusting a setting called latency. Depending on the protocol used for the download, especially if there are lots of small packets of only a few bytes like a handshake, this could make it take longer with the converter than a hardware port. Well, it’s always going to be slower due to the nature of the beast so it is more a question of how much slower. Anyway, if I remember correctly the adjustment is on the advanced settings tab of the com port in device manager. It might be different if you have a different driver version (or I might just not remember it correctly too, heh) but I am pretty certain it’s talked about on their website ftdichip.com
EB
:mrgreen:
Hey thanks,
I started playing around with the latency and buffer size. When I used the defaults a download of a 15.8K program was taking about 2.5 minutes. When I decreased the buffer size to 1k and latency to 3, I now have the download taking about 1 minute
Thanks again
I have left all settings, other than Reset Hold Time, alone.
I have tried values for Reset Hold Time from 0 to 75 in increments of 5, with no luck.
When scanning the COM ports for the Atom Pro-28, the program in the module stops running when it gets to checking COM4, which is the port assigned to the BAFO adapter. After a pause, a window comes up with a message that reads: “Could not find AtomPro”. When I hit “OK” to clear that window, the program in the ATOM Pro restarts.
The settings above match those on my desktop, which has no problems programming the Atom Pro-28 module using the same ABB carrier board, through a “true” RS-232 port.
My guess is that AcidTech’s setting or drivers are different than the ones that we have tried.
As I have stated in a few different posts, my FTDI based USB to serial converter is working fine now. I think I still get better throughput on my desktop machine with a hardwired serial port, but it is working well enough.
I retried my BAFO this morning, with the same bad results as before. I tried changing the settings for the prolific driver with no improvements. Note I have had other USB to serial cables in the past that used prolific drivers. My first one may have been a clone that may not have been 100% compatible. I since junked that cable… So I wonder if maybe this cable left some bad settings somewhere…
My BAFO is using the latest driver off of the Prolfic site. It would be interesting to see more details from AcidTech’s machine that is working properly. Things like Driver Details, and the advanced settings for this device as well as his settings within the Atom Pro software.
Just adding some info on the Bafo. According to another user the BAFO only worked for him when it was on a powered hub. Most laptop USB connections are not powered. This could explain why it works for me but not for you(my laptops usb ports are powered).
I too was having problems getting my Parallax USB to RS-232 converter to talk to an Atom 28 on an ABB.
The Parallax USB to RS-232 converter worked fine with any flavor of BS2 on a Parallax BOE.
Since the Parallax USB to RS-232 converter uses the FTDI chip, I tried reducing the latency to 3 as suggested, but left the buffer size at 4K.
The IDE will now talk to and program the Atom 28 on the ABB.
However, I will have to remember to change the latency back to the default of 16 when I use the Parallax USB to RS-232 converter with a BS2 on the BOE.
I don’t think you’ll need to change back when you use the BS2. I also would suggest changing the Latency timer to 1 instead of 3. This should improve the overall speed when programming. I just got the um232-r 24pin module for testing with our devices and found 1 was a bit faster than 3.
