Suggested Mod For Some Baycom Type Modems ----------------------------------------- Originally I found on both my DX2 and P120 PCs that the Baycom driver worked a lot better on some ports than on others. It just didn't seem to want to receive on some ports and all the ports that gave problems had 16550 (buffered) UARTs. I had seen comments from other people about problems with Baycom modems and 16550s, but I just could not understand why there should be such a problem. Eventually I found the solution to my own problems - it might not work for everyone, but if it works on two of my PCs then I'm sure there will be a lot of other people who will find that it works! Both my Baycom modems are built on boards from a well known UK supplier of PCBs and kits (Badger Boards). They drive the CTS output direct from the 74HC04, and the low voltage on the output is about +0.5V - it never swings negative. That may be in spec for TTL, but it's way out of spec for RS232. I have another genuine Baycom board, which I have never got round to building, that drives the CTS output via a PNP transistor which ensures that the output voltage can swing negative. I modded both my boards to the Baycom design and the driver then worked perfectly! The implication of this is that the problem is not caused by the 16550 or its emulation, but by the line receiver arrangements being more in spec than those employed with older UARTs. Here's the mod - I'm not too good at ASCII schematics, so I hope it makes sense! (The original circuit has a 2k2 from pin 6 on the 74HC04 to pin 8 on the 9 pin 'D'). "+5V" is the output from the regulator, which goes to pin 1 on the TCM3105, pin 14 on the 74HC04, etc. RTS (pin 7 on 9 pin 'D') | < < 10k < |--/\/\/\--- CTS (pin 8 on 9 pin 'D') C| 2k2 10k / 74HC04 pin 6 ---/\/\/\--|B BC213L or similar *PNP* transistor \ E| | +5V Roger Barker, G4IDE June 1997