Diagnostics/ram check for U1MB?

[svhovater]

Trying to track an issue. 

Diagnostics for U1MB, anyone?

 

thanks

Steve

[flashjazzcat]

If you want to test the extended RAM, XRAM will do. Or is something else acting up?

[svhovater]

Hi Jon

Xram ? Where?

 

as to the problem - I put the U1MB in my 65XE, and that and the Side2 were magnificent.

Unfortunately, the power LED doesn't work on that machine, and I neglected to power the machine down

before I removed the Side2. Now, the Side2 mostly works - but the real problem comes in when I operate

spartados X - from the U1MB. I have the settings to use the 1024 Rambo handler for the ram disk, but after awhile

I type "DIR", and get a ton of "symbol undefined" errors. 

I noticed that when I use a different handler for the U1MB ram - say, the 256 or 320k handlers, I don't see

the problem. Or haven't since. My working theory is that the side2 insertion/removal with power on did something

to the memory that spartados is bankswitching. But I could just be totally full of beans, too.

So - my first step is to "formally" test the extended ram. I'll hunt around for XRAM. Thanks for the tip.

 

[svhovater]

found xram 0.21

 

[flashjazzcat]

OK: let us know what happens.

 

[_The Doctor__]

really shouldn't damage anything by having a cart come out... unless the cart card edge doesn't have the fingers with differing length to prevent power issues on hot insert/pull of the cart.

 

if the data were damaged you'd think it would fail immediately.. if it's happening after time passes... well then there may be something going thermal.

 

You could get lucky though, a little carbon built up on the cart port or the sides fingers... give it a good cleaning and see if that helps.

 

only other thing would be to put it in another computer and flash/reprogram them while cold or on an exernal programmer...

 

Can't hurt to try them on another machine to make sure where the damage is for the individual parts of this equation... Atari/side/u1m

Edited March 6, 2021 by _The Doctor__

[svhovater]

Finally got back to this - I ran the xram test, and I didn't see any errors.

Just for fun, I reflashed the U1MB to 3.10, and haven't seen the problem recur yet.

Seems like a Schroedinger thing. When I am *trying* to reproduce the problem, it's

rock solid. Sigh.

 

[_The Doctor__]

This is a slight pattern as a handful have re-flashed it and it starts to work again...

Edited March 13, 2021 by _The Doctor__