Physical Memory Cards Not Working On Latest PSX Core

[cloudyng44]

my physical memory cards can not work on lasted PSX core (20220701) through snac port, but they can work on last month‘s PSX core (20220601) can somebody help me solve the issue？

[cloudyng44]

I used this design of snac https://github.com/blue212/SNAC-PSX
It works well on last month‘s PSX core(20220601)
but the lasted psx core can not identify physical memory cards

[FPGAzumSpass]

Will ask Blue1, who made the SNAC implementation if there is a possible reason.

[cloudyng44]

I think Jun 24 update "Pocketstation SNAC“ cause the problem

after that update，the physical cards can not work any longer

@blue212

[FPGAzumSpass]

It still works for Blue1.

Which game do you have problems with?

[cloudyng44]

every game⋯⋯ the lasted public（20220702 ）psx core can not find physical cards

[FPGAzumSpass]

Can i assume that you use a multitap to connect the cards?

Blue1 verified it works with newest unstable builds and Porkchop also did.

please also report back:
- what card exactly (original memcard or 3rd party)
- which game as example

[venice]

I am running the PSX Core 2022-07-02
without Problems using physical MemCards.

Card 1 is an original Sony,
Card 2 is a PicoMemcard found here https://github.com/dangiu/PicoMemcard

Many thanks to FPGAzumSpass for his great Core
and to blue212 for his PSX Snac Adapter.

[cloudyng44]

Can i assume that you use a multitap to connect the cards?

Blue1 verified it works with newest unstable builds and Porkchop also did.

please also report back:
- what card exactly (original memcard or 3rd party)
- which game as example

I used the original memcards （scph-1020 made in japan），and
play《Castlevania: Symphony of the Night（usa）》
im sure i dont use a multitap

[cloudyng44]

I am running the PSX Core 2022-07-02
without Problems using physical MemCards.

Card 1 is an original Sony,
Card 2 is a PicoMemcard found here https://github.com/dangiu/PicoMemcard

Many thanks to FPGAzumSpass for his great Core
and to blue212 for his PSX Snac Adapter.

do u use north america‘s memcards？maybe my card which is made in japan cause the problem？

[FPGAzumSpass]

Blue1 only tested memcards with multitap until now.
Maybe the issue is related to you not using one. It's hard to say it this point.
I guess you have to wait for him to be able to test without multitap.

[venice]

do u use north america‘s memcards？maybe my card which is made in japan cause the problem？

My one is a „SCPH-1020 Made in Japan“.

[cloudyng44]

do u use north america‘s memcards？maybe my card which is made in japan cause the problem？

My one is a „SCPH-1020 Made in Japan“.

I use snac and insert 2 cards and 2 oringinal controller.how about your input device？different from mine？

[venice]

I am using a Dualshock 2, SCPH-10010.
The NeGcon works as well.
Never tried two Controllers at the same time.
And I only have one original MemCard.

[nonete]

Today two more memory cards have arrived to my home, I had two that work well for me both in the mister's snac and in the console and I wanted another two, two for the mister and two for my real psx console.

The four are scph-1020 model, the first two that I already had work fine for me on the snac, but the two new ones don't work for me in the mister, but all the four work on the original console.

Strange, I've been doing tests, copying things from the real console, from the cards working in the mister to the don´t working, i put in the snac and nothing happens.tested port 1 and port 2, working memory in port 1 not working in port 2, i try all, the two news memorys don´t work in the snac.I've decided to open them to see inside.

Although they are all of the scph-1020 model, NONE of them are the same inside, they all have different chips, they do not resemble each other. The two that work for me, one is a Japanese model and the other is a transparent blue one, all original from sony .

In the end I have given up and I have left the two that work for me in the mister and the other two that do not work in mister for the original psx. I do not know if they have to be formatted or according to the type of chip they have, they work or No, I mention it in case it can be of help.

If i can help, i can do test.Don´t know why some memory cards works and others no, maybe diferent chips inside...

Excuse my english....

[nonete]

Also, I have the original multitap, I just tried and the same thing, two work and two don't. Tomorrow I will do more tests to see if I can get all four to work

[cloudyng44]

could you show me the back of your cards which can work well through snac？

[venice]

I found somewhere that some counterfeit MemCards use the 7.6v (converted to 5v) from the rumble system.
Depending on the PowerRegulator on the MemCard the 5v from the SNAC Port can be enough or not if the voltage is not “stepped up” to 7.6v.

Maybe that helps

[nonete]

could you show me the back of your cards which can work well through snac？

A photo is not going to help you, they are all of the same model, and apparently the same on the outside, 2 work and two don't, the interior changes, with an external photo you won't know which ones work and which ones don't.

There is a blue transparent working, otiginal from sony, but not 100% sure if all blue transparents works.

[cloudyng44]

I tested every psx core between date 2022-6-18 and 2022-6-26
I found the problem first appeared on 2022-6-20 core

at last I solve the problem by editing the file psx.sv and use quartus to generate the rbf file again.open the psx.sv file and I edited the line 406 code
‘TURBO’ back to ‘Data Cache’

then i use the edited rbf，i can find the physical memcards again.

[FPGAzumSpass]

Please check in OSD->Misc if you datacache/turbo setting is set to "ON(U)" and set it to off.

Renaming the option does nothing, but maybe quartus will ignore your setting now as it doesn't find the port name.

[venice]

Off at my side and never switched on.

[nonete]

Off here too, if you share the modified rbf core I can test if the four memory cards work for me with that modification

[cloudyng44]

Please check in OSD->Misc if you datacache/turbo setting is set to "ON(U)" and set it to off.

Renaming the option does nothing, but maybe quartus will ignore your setting now as it doesn't find the port name.

Im sure I never turn it on

[FPGAzumSpass]

Then your change has changed absolutly nothing for you and cannot fix the issue...

- The OSD text change(line 406) does nothing at all, it just changes the visible text
- the port rename(line 912) will make the input port absent, likely leading to forcing it to zero = off

[pgimeno]

I wonder if this can be the kind of issue where different compilations with different seeds can make the difference between working and not working. That happened to me with the Pong core and with the SAM Coupé core.

[aberu]

I will revalidate the status bit addressing changes I made in PR #143 just to be sure, but I doubt that is the issue as well since the OSD doesn't seem to be associated with this.

That commit you are referencing @cloudyng44, did more than just rename the OSD variables:

https://github.com/MiSTer-devel/PSX_MiS ... 8ce47ba4f3

The GTE Timing was significantly reworked, which probably wouldn't be a problem, but that's beyond my abilities to determine.

EDIT: Revalidated my commit, it's fine:

validateosd.jpg (119.19 KiB) Viewed 3839 times

[FPGAzumSpass]

IF the GTE timing would be an issue, blue1 and others would also see this, but for them it still works fine.

So it is probably something with either:
- OSD settings
- special card
- connection situation(no multitap)

Build timing could also be an issue in case there is some metastability issue.
I will add 2 stages for the input pins in the next build to rule that out.

[venice]

I have not used a multitap.

[FPGAzumSpass]

ok, so this one is also out.