DigicoolThings Forum

Thanks David,

I do have some alternate serial devices on order which I'll try once they arrive. It just seems strange that it doesn't work without hardware hand-shaking i.e. I can load software via the monitor with no problems. Of course, there might be extra overheads involved with XMODEM (I'm not really familiar with the protocol) but I would've still thought there should be enough grunt to handle it at low BAUD rates). There also seems to be a rush of characters sent to the terminal after the transfer times-out which I also find unusual - almost like it is not even checking the status of the serial port. I'll see if I can find a version of kermit somewhere and try that. It's not anything critical - just something that I found a bit annoying as it seems like it should be something that just works. I notice that there is also a DOWNLOAD.COM but that seems a bit clunky as you need to convert your file into another format before transferring - I guess I should try that as well but it's not my ideal choice. Anyway, I have a few things to try - I was more curious as to whether I was just doing something completely wrong.

Probably not much help but XM is working for me. I'm using a FTDI USB to 5v adapter with CoolTerm on a M4 Mac. I selected CTS handshaking in coolterm. baudrate is 38400 bps.
I think Ive had problems in the past where transfering to ram disk was more reliable than using IDE devices.
hth David.

edit: just checked, its an FTDI serial adapter

Thanks for running that test. Just to check was this running on the MECB Z80 with the 6850 serial connection or on the RC2014? I just tried via a Raspberry Pi and had the same issue. I'll try and get CoolTerm just to ensure it isn't some weirdness with minicom.

Ha, looks like it is something weird with minicom - I just tried it with CoolTerm and it worked fine even without any flow-control.

Thanks for helping me debug this - at least I know it is not something weird with the MECB hardware or software now

Yes XM was tested with MECB 6850. It didn't work for me without cts flow control at 38400 bps. Glad it's working for you now, David

Hi David,

Perhaps the faster 10 MHz Z80 I have on mine helped to push it over the line for 38400 usage?

By the way, I connected up to the Fujinet discord group. It looks like there is still quite a bit of discussion with respect to how to implement the interface to the ESP32 device (which does all the grunt work) on the RC2014. i.e. whether it be a parallel connection via an 82C55 (as in the H8/H89) ... which in itself seems to be taking a long time to develop (over four years, I would've naively thought this to be a relatively simple alternative), using the SPI interface (for which they are debating various implementations), or going via a UART. I can definitely expect a great deal of effort required on the software side (on both sides of the interface) but I'd have to admit that I'm a bit surprised that the interface itself seems to have been a sticking point for so long (over four years for the RC2014 as well).

Emil.

I had a look on the Fujinet discord, the RC2014 version seems to be very much in flux with parallel, serial, and spi contenders.
I have an Apple II here for repair, I'm tempted to try and get it working with Fujinet - need to fix it first though.

I have not yet managed to get a CF interface woorking, Your boards arrived yesterday and will try one shortly, but need to wait for some connectors to arrive first.

A post I made yesterday has vanished, the jist of it was:
T-Net gives you access to files on your PC via a serial port.
T-Net is easy to implement just needing a spare serial port (do you have one?)
The CP/M drivers already exist and work.
The source code and lots more too is available on John Hanson's website:
https://www.cccplusplus.co.uk/xbeaver/x ... vault.html
source files exist for various UARTs, I highly recommend and shall dig it out when I get the CF interface working.

Best regards, David.