Page 51 of 66

Re: exxos's DFB1 trials

Posted: 22 Jun 2023 22:16
by exxos
I think what has happened I moved a resistor downwards at some point and the "broke" the vcc to it :roll:

This wouldn't have happened in eagle :lol: :roll:

Capture.PNG

Re: exxos's DFB1 trials

Posted: 22 Jun 2023 22:20
by exxos
Yep this probably isn't helping :lol:

IMG_0749.JPG

#Bobslaw


EDIT:

Works now :lol: :D

IMG_0750.JPG

#Idiot
#TinyBodgeWire

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 09:42
by Badwolf
Oh wow! The top address line was floating?

Amazing anything worked!
exxos wrote: 22 Jun 2023 22:16 This wouldn't have happened in eagle :lol: :roll:
Someone didn't run their DRC check. ;)

BW

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 12:36
by exxos
Badwolf wrote: 23 Jun 2023 09:42 Oh wow! The top address line was floating?
Yep!
Amazing anything worked!
Yeah, I assume on power up it was probably coupling to some other pin forcing it high. After 15 or so second it likely drifted back to zero and started working.

Someone didn't run their DRC check. ;)
Yeah I did, then probably moved a bunch of stuff afterwards :lol:

It's my main annoyance in KiCad actually. Moving stuff breaks its connection :roll:

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 12:52
by Badwolf
exxos wrote: 23 Jun 2023 12:36 Its my main annoyance in KiCad actually. Moving stuff breaks its connection :roll:
I've never had it any other way, so I wouldn't miss it. I used DOS OrCAD for a morning back in 1995 and another package extremely breifly the same day and then KiCAD a couple of years back. None let me move stuff and not break traces. :)

BW

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 12:57
by exxos
Badwolf wrote: 23 Jun 2023 12:52 I've never had it any other way, so I wouldn't miss it. I used DOS OrCAD for a morning back in 1995 and another package extremely breifly the same day and then KiCAD a couple of years back. None let me move stuff and not break traces. :)
Yep I have had to do that for the past 30 odd years. But Eagle9 has the tool which is a must when you have RSI. Not much works in Eagle 9, but the moving tool, DRC, straighten tool are the few things which almost work right in it. I still use Eagle 7 for everything else.

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 14:17
by exxos
Just been tidying up all the code and gone back to flashing the firmware now :) Everything behaves as expected now :)

I do think it will need some sort of cooling, even though it seems to run all day fine, it does get rather hot. Either by sticky carbon sheet to the metal shielding, or a small fan fitted *somehow*. Though could be tricky with the keyboard :shrug:

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 14:47
by stephen_usher
Some sort of metal block which would bridge between the CPU and the metal shield and then use that as a heatsink, in a similar manner to the C64, may be a possibility.

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 15:06
by exxos
stephen_usher wrote: 23 Jun 2023 14:47 Some sort of metal block which would bridge between the CPU and the metal shield and then use that as a heatsink, in a similar manner to the C64, may be a possibility.
Indeed.

I don't think I have the metalwork to test anything, will have to look when I get chance. . Though getting blocks cut to size could end up expensive over hacking on a fan somehow.

I would probably use the carbon sheet and see if its long enough to connect to the bottom shielding :shrug:

Re: exxos's DFB1 trials

Posted: 23 Jun 2023 15:31
by Badwolf
1) I'm not sure it'd fit with top shielding in place anyway.

2) the little Exxos 030 heatsinks do just clear the keyboard IIRC.

3) mine does get jolly hot. Not helped by the old and manky Atari PSU right next to it getting hot too, but with the internal fan running at reduced speed I've not had an overheating event yet with quite a lot of agressive soaking. Fully buttoned up, I compiled EmuTOS twice on it, running from a ramdisc (so virtually all at 50MHz) and can run the Quake loop (at about 0,5fps) which uses the FPU for hours.

Plastics at 40MHz might be worse than the ceramics at 50.

I did bluetac a thermocouple onto the heatsink at one point. I can't remmeber the figures, but I don't think it got to untouchably hot (~70C).

BW