SOLVED trump-card V1 filled with ram chips only sees 256 kb

[techy]

okay , as promised an update ;

eprom 1.37 vs 1.21 ; NO difference
Pal 2v2 changed for known good one ; NO difference
Verified ras 0-2 from 2v2 to all memory chip sockets ; OK
all memory chips tested ; OK

Minerva did not detect any bad chips , apparently does not see more than the 256kb the trumpcard reports

im an at the end of my ideas , anyone more tips ?

[Pr0f]

It would suggest to me a problem with the A19 address line, as Minerva will see the trump card ROMS at the extra addresses that it scans - if the fault is on the card itself, then it may not work in a JS or JM QL - as that would need to rely on the trick of seeing the Trump card ROM at 0C0000h first (where both A19 and A18 are high) before remapping the ROMS to the final locations.

Incidentally - A19 is not used in a standard QL - so any fault with that line will not show up until you start working with expansion cards.

[Pr0f]

Do you have any other plug in cards you can use to test if A19 is working from the computer?

[techy]

i have 2 trump cards myself , and they work fine on the same ql .

[Pr0f]

so the issue is specific to this trump card then...

It will have something to do with A19 I would think - it's not seeing the signal from this line, so consequently only seeing 256K of the RAM - the other 512K of RAM where A19 would be high is not being seen by the CPU.

[techy]

i am currently testing all memory chips to be sure .

And will continue searching A19

[Pr0f]

From the design of the Trump card 2 - the PAL's won't be hugely different in both version 1 and version 2 of the card - the main difference was the reduction in RAM chip's. A19 is used by one of the PAL's to decode the different chip selects - and also the line that is used to shift the location of the ROM in the trump card. With Minerva that circuit is basically rendered redundant as the ROM scanning includes the final locations of the Trump card ROM (it's basically seen as 2 adjancent ROMS at 010000h and 014000h)

The other PAL is responsible for the RAS and CAS lines - and these also need A19 to work

If I get chance tomorrow I will try and read the PAL's in the Trump card I have and see if I can convert back to equations.

I have the trump card 2 equations for comparison

[techy]

got 512 kb !! 640k total

1 defective 41256

2 banks working , now to get the last bank to work

[techy]

and 3 banks working ,

another failed 41256 .



THANK YOU ALL !! for supporting me in finding the solution .

[Pr0f]

Well done

Glad you were able to get it working!