What does "Bad Checksum" mean?

Started by nathanchronister January 10, 2005

Hello. I'm hoping you may be able to help me with this. When running
a program on the BX24, I keep getting a message in the Downloader
Status Window that says "Bad Checksum". From time to time, I also
see, very briefly, the message "Stopping Basic X". The program
appears to keep running or at least resumes quickly.

Currently I'm trying to get the BX24 to work with the Al Williams PAK-
VII coprocessor (see below). The PAK gets its power from the BX24
regulator and they reside on a board that I constructed. The sample
communications program seems to work most of the time, but the "Bad
Checksum" message is there almost the whole time and I thought it
might account for some problems I'm having. For example, I haven't
been able to read a pulse duration from the PAK register. (I've seen
the "Bad Checksum" message in the past, when running other programs,
but not so frequently.) Do you know what this message indicates, or
what might be causing the problem?

Thanks for your help,
Nathan

Al Williams PAK-VII coprocessor for PWM input:
http://www.al-williams.com/pak7.htm

Sample program for reading PAK timer output:
http://www.al-williams.com/bx24.htm





--- In , "nathanchronister"
<nathanchronister@y...> wrote:
>
> [...] The PAK gets its power from the BX24 regulator and they
> reside on a board that I constructed. [...]

This is generally not a good idea unless you have determined that the
power consumption is absolutely below the BX's current capabilities.
Add an external regulator and run the BX-24 and other circuitry off
of its output.

The occurence of checksum error indicates that some number bits were
received in error during communication. Low supply voltage (due to
excess loading), external noise, bad connections, alpha particle
storms and other factors can cause bit reception errors. The message
is an indicator that something is going awry.



Hi Nathan,

The BX-24 hardware manual cites 17 mA to 25 mA for the BX-24. The onboard regulator is rated 100 mA max.
The PAK-VII lists 77 mA. This suggests that some of the time it is exceeding max specs by 2 mA (or more).
Add an external 5 VDC power supply with a least a 100% safety factor for current rating over temperature.
Add more heatsinking if necessary.

Best Regards, Eric
----- Original Message -----
From: nathanchronister
To:
Sent: Monday, January 10, 2005 11:23 AM
Subject: [BasicX] What does "Bad Checksum" mean?
Hello. I'm hoping you may be able to help me with this. When running
a program on the BX24, I keep getting a message in the Downloader
Status Window that says "Bad Checksum". From time to time, I also
see, very briefly, the message "Stopping Basic X". The program
appears to keep running or at least resumes quickly.

Currently I'm trying to get the BX24 to work with the Al Williams PAK-
VII coprocessor (see below). The PAK gets its power from the BX24
regulator and they reside on a board that I constructed. The sample
communications program seems to work most of the time, but the "Bad
Checksum" message is there almost the whole time and I thought it
might account for some problems I'm having. For example, I haven't
been able to read a pulse duration from the PAK register. (I've seen
the "Bad Checksum" message in the past, when running other programs,
but not so frequently.) Do you know what this message indicates, or
what might be causing the problem?

Thanks for your help,
Nathan

Al Williams PAK-VII coprocessor for PWM input:
http://www.al-williams.com/pak7.htm

Sample program for reading PAK timer output:
http://www.al-williams.com/bx24.htm ------
Yahoo! Groups Links

a.. To




Thanks for your help on the question I posted earlier. In addition to
sharing the power supply, is it possible this could be caused by
noise in the serial cable from the BX to the computer? I'm using a 4-
conductor, 10 foot ribbon cable for this purpose. I wanted a
lightweight, long cable to free up my project, and I didn't realize
noise or signal loss might be a factor.

I'll try running some BX programs without the PAK to see if it still
has problems. If it works then I can assume the power supply is to
blame. If it still doesn't work I won't know if it's the cable or the
supply or both!

Thanks again!
Nathan





--- In , "nathanchronister"
<nathanchronister@y...> wrote:
> [...] I'm using a 4-conductor, 10 foot ribbon cable [...]

I wouldn't hesitate to use a cable like this one or two feet long but
10 feet seems too long to me; even more so if you're using it in a
noisy environment.

The BX-24's serial interface doesn't use drivers that meet the
specifications for RS-232; they only swing from 0 to 5 volts while
the minimum RS-232 swing is -3 to +3 (if I recall correctly). While
this does work in many/most cases, the longer the cable and/or the
noiser the environment the more likely it is that this will be a
problem.

I am currently using two 6 foot shielded cables (commercially
available) connected together to make a 12 foot cable (in an
office/lab environment) and have not noticed any problems at all.
These cables are quite flexible and don't pose a problem for me on
the test bench. I would, however, recommend a single cable of the
required length as there is less of a chance of problems.

Don




--- In , "nathanchronister"
<nathanchronister@y...> wrote:
>
>
> Thanks for your help on the question I posted earlier. In addition
to
> sharing the power supply, is it possible this could be caused by
> noise in the serial cable from the BX to the computer? I'm using a
4-
> conductor, 10 foot ribbon cable for this purpose. I wanted a
> lightweight, long cable to free up my project, and I didn't
realize
> noise or signal loss might be a factor.
>
> I'll try running some BX programs without the PAK to see if it
still
> has problems. If it works then I can assume the power supply is to
> blame. If it still doesn't work I won't know if it's the cable or
the
> supply or both!
>
> Thanks again!
> Nathan

Ribbon cable is more sensitive than twisted cable but I have
successfully used a 100ft UTP-cable between the PC and BX24 (also
for downloading new code).

Ingvar