Doesn't look like a software problem at all. The Keil board has the usual arrangement of transistors joining RTS to /RST and DTR to P0.14. With the jumpers removed to disconnect the transistors from /RST and P0.14, all works fine. And even when I use my RS232 breakout box to disconnect DTR and RTS from the host, during high activity on the RS232 TD or RD lines, the outputs of the two transistors for /RST and P0.14 show noise at about 10KHz (unsurprising from 9600 baud comms) of about 2v (in the case of the P0.14 line, with a 22K pullup) and 1v (in the case of /RST, with a 10K pullup). Alas, I can't too easily get scope probes onto the gates of the transistors on the board because they're tiny surface mount ones, but I'll try investigating those next to see what's happening there. Interestingly, with RTS and DTR disconnected from the host, a scope put on thise lines at my breakout box is showing noise similar to the noise on the outputs of the transistors - so I suspect the noise is coming into the lines on the gate circuit somewhere. Now, the schematic for my MCB2100 board shows pretty much the exact same circuit as listed under RTS_DTR.DOC in the files section. I doubt it's an actual circuit problem, since there's no electrical connection between the DTR/RTS stuff and the RD/TD stuff, so I guess it's probably a problem with the layout of the board or something - but if anyone else has a board that does the DTR/RTS thing, could they possibly try maxing out the RS232 port while watching the actual /RST and P0.14 lines with a scope, to see what they can find? Anyway, with the jumpers pulled on the board so the lines on the CPU are only connected to the pushbuttons, it works fine :-) I'm planning on building a board with an RS232 transceiver and DTR/RTS ISP logic (to meet the fine new Standard ISP Header spec that Rickman has been helping with on the Wiki!) - I think I'll build it on breadboard first and experiment to make sure it doesn't end up exhibiting the same problem. Thanks, ABS
Message
My problems with my LPC2106 crashing under high serial throughput...
2004-02-28 by Alaric B Snell
Attachments
- No local attachments were found for this message.