Ok I found the problem, fixed it and got them working. Now I like these modules!
The fault was this. When I programmed the first module, I put it in a socket, programmed it and took it out. Then I realised I needed to program the second module. So I wanted to ensure the settings were the same. So I put the first module back in and hit "Read", assuming this would read back the data. Then I programmed the second module with this data.
The problem is that there appears to be a bug when you hit "Read" it reads back incorrect data, specifically it reads back the frequency as 315Mhz, no matter what is actually in the module. But, the *second* time you hit "Read", it reads back the correct data.
So to see what the module is programmed with, hit "Read" twice.
Of course, what had happened is that I had programmed the first module at 434Mhz, and the second module was programmed with the erroneous data from the first readback at 315Mhz. The capacitor values for the antenna also read back incorrectly on the first read.
Range is good. Modules work right next to each other and I got 250 metres which is what the data sheet says. And I was sending a serout with both a qualifier "ABC" and a checksum, so this is a real reliable data range.
At Last!! A module that goes as far as the spec sheet says.
So I'm giving these the picaxe tick of approval
Manuka - a suggestion and a question. My modules were the RS232 ones (there is a max232 on board) so could be programmed directly from an RS232 cable, so one could imagine a little programming board with a D9 female socket and a 6 hole socket and an onboard 5V reg. And a question - if you have the TTL module and you program it from an RS232 connection, will the -12V zap the unit?