My X5000/40 has started to switch itself off every now and then. Within a year there has been at least 7 cases. The standby leds (green and red) stay on and the MCU Error led is blinking when this happens.
What is causing this issue? Can it be fixed somehow?
I have a top-quality PSU, and have checked the power cabling. CPU is not overheated, the temperature stays under 65°C.
My X5000/20 has had this problem over the years, and it just started getting worse and worse.
I have the MCU connected to serial so I can monitor the error and see the output. I used the 1 second output option and watched the vdd_3v3_xorro drop from 3.21 to 2.95 over the period of about 15 minutes.
My X5000/20 has had this problem over the years, and it just started getting worse and worse.
I have the MCU connected to serial so I can monitor the error and see the output. I used the 1 second output option and watched the vdd_3v3_xorro drop from 3.21 to 2.95 over the period of about 15 minutes.
My X5000/20 has had this problem over the years, and it just started getting worse and worse.
Thanks for your answer, Bill! I will check that cable you gave the link for...
But is there any way to fix that kind of problem? It is ironical that Xorro is not used anyhow by most users... Could you just 'eliminate' some component's related to Xorro, to get rid of that issue?
My third reply. I keep quitting without pressing "Submit". Apparently I'm losing my mind.
There is software to read your voltages from a shell, without needing an MCU cable. It's free, it's boring, and I can't test it right now, but if Bill tells me it's working, I'll put it up on the Depot.
Anyone sending an email to "lylehaze at gmail" can have it as soon as I reply to your mail. Just ask for "MCU"
!! NOTE There's already tools to do this on the Depot. Just search "MCU" !! Yup, I'm losing my mind.
LyleHaze wrote: There is software to read your voltages from a shell, without needing an MCU cable. It's free, it's boring, and I can't test it right now, but if Bill tells me it's working, I'll put it up on the Depot.
Anyone sending an email to "lylehaze at gmail" can have it as soon as I reply to your mail. Just ask for "MCU"
!! NOTE There's already tools to do this on the Depot. Just search "MCU" !! Yup, I'm losing my mind.
Thanks... But I'am already using 'MCUinfo.docky', which shows the same info. But as the shutdown happens suddenly, you do not have time to see which the voltages has dropped! So I think the MCU cable is the only way to record that info...?
@Gregor the serial cable displays on a second computer, so the evidence remains after a power down.. true, and advantage.
using a software method can't do that, but it can let you monitor gradual changes before the fatal ending. and it has the advantage of being free, requires no additional computer or cable.. you can have it right now.
I know nothing of the stuff on the depot. my own simple program can show all data every run, or it can show only a specific voltage of your choosing, in case you want to gather data in a script..
none of this is rocket surgery, I just wanted to offer what I have on hand in case it's useful. I don't have a freescale based Amiga any more, so I had to ask Bill if he would verify the program first.
And you asked about replacing parts.. low voltages can be caused by the supply (regulator) or by excessive load. chips that draw too much power get noticeably warm.. but history puts the odds on a flakey regulator, which may not get warm at all.
LyleHaze wrote: using a software method can't do that, but it can let you monitor gradual changes before the fatal ending. and it has the advantage of being free, requires no additional computer or cable.. you can have it right now.
That's true... When I have followed the voltages with MCUinfo, the values have been very stable. There are only four values in which there is a constant 10mv 'flip-flop':
And you asked about replacing parts.. low voltages can be caused by the supply (regulator) or by excessive load. chips that draw too much power get noticeably warm.. but history puts the odds on a flakey regulator, which may not get warm at all.
These shutdowns has appeared totally randomly, both after the computer has been on for a long time, and just for a few minutes. That suggests there is no temperature issue, at least. Also, I have had both AmigaOS and Linux running when the issue appears, so the OS does not seem to affect it directly.
The manual has the pins. Let me know if you need those as well.
I borrowed a similar type of adapter cable (ADA954 3.3V TTL serial) which should do the same, but I cannot get any data out with it. I connected it to the MCU debug connector like the one you used (pin1 - black, pin4 - white and pin5 - green).
The connection itself seems to work as there is no error message if I run "sudo screen /dev/ttyUSB0 38400" on terminal . There appears just a white blinking rectangular (prompt?) but trying to input and run commands like 'help' or 'v' does not show anything on the window.
Trying to use 'ttyUSB1' which you had in your example gives just an error: non existent.
I tested also with 'Cutecom' but no output from the MCU, either. Connection itself seems to be ok.
Any idea what is wrong...? Is there some jumper on the motherboard which should be opened/closed so that the debug connector would work?
Gregor wrote: I borrowed a similar type of adapter cable (ADA954 3.3V TTL serial) which should do the same, but I cannot get any data out with it. I connected it to the MCU debug connector like the one you used (pin1 - black, pin4 - white and pin5 - green).
I found the problem... The white and green cables had to be the opposite way!
Now about trying to catch the switch-off issue... Do you have to have the continuous 'printing' on? Or does the MCU send out the error message anyway?