Banked Video RAM?

[zhulien]

CPC 6320 and CPC 6512. 
Why don't you say that history knows such cases.
ps. zhulien 512K VRAM possible. I'm dying laughing

https://www.cpcwiki.eu/index.php/CPC_6320_/_CPC_6512_-_internal_320K_/_512K_for_CPC_6128

This?  I haven't seen it before - so will have a read.

I just read it - interesting indeed, but a bit different to what I suggested.  I am surprised the different logic for the 192k extra ram vs the 256k (silicon disc) in that - was it cheaper for DkTronics to just include that same circuitry and ram in both units than just 192k in the 256k ram expansion?  I actually thought the 256k ram expansion disabled the internal ram- it does right? but it behaves like the internal ram?.

[McArti0]

New PAL cpc

on GreenPAK SLG46533v-DIP



Only We need after upload CPLD change DIP upside down and cross by wire PIN 2 and PIN 8.

When the courier brings the kits, I will test and let you know.

[McArti0]

Link to simulator HOW NewPAL WORK...

[GUNHED]

Would be nice to see a video of the extended PAL in work on a real CPC 

[McArti0]

I got a programmer SLG4DVKDIP - GreenPAK DIP Development Board ... and Soft does not see the programmer and the chip inserted into it. It won't be easy...   

[Bread80]

I've used that range of chips before (but not that exact chip) and the tools usually work pretty smoothly. Make sure you have the right development board selected ('Platform') and the *exact* chip model specified in the file - even the 'same' chip in a different package will require a different file.

I'll probably get one of the chips the next time I'm doing a Mouser shop but that won't be for several weeks.

My comment on this: the chip they've chosen is only available in QFN packages. The DIP variants appear to only be available in small quantities (at least from Mouser) and they're relatively expensive. If they'd developed it on chip (also) available in a TSSOP package it would be a lot more practical.

[McArti0]

Thanks for the answer Bread80. Your site has been very helpful.

https://bread80.com/2021/06/03/understanding-the-amstrad-cpc-video-ram-and-gate-array-subsystem/

btw. I wouldn't have thought that with the rising edge of the nCPU, nCAS could be low.

The problem is that Soft does not detect the Programmer. In System Info Board is N/A. As if the drivers are not installed or the device is broken. Only the blue LED is on. I can't access the GreenPAK forum, because this site does not respond to login, something is broken.

[McArti0]

100 Euro in shit.   Progammer and 11 chips.

[GUNHED]

Sorry to read 

[McArti0]

the complaint is pending. www.futureelectronics.com has sold.

This programmer is cheaper and no longer produced. "I saved"

[robcfg]

Could you provide us with the models of the programmer and board you have?

It would be nice to know which OS and machine are you using for that, what type os, I guess USB cable, you use to connect it, if you're using a USB hub, etc...

If the programmer is somewhat older, maybe you could try an older version of the software. Also try different USB cables, and such to rule out problems.

[McArti0]

https://www.renesas.com/us/en/products/programmable-mixed-signal-asic-ip-products/greenpak-programmable-mixed-signal-products/analogpak/slg4dvkdip-greenpak-dip-development-board

Full legal:
Windows 10 64-bit
original wire USB
Last version soft support programmer and Chip SLG46533v

Tested on two other computers.

https://www.renesas.com/us/en/document/mat/greenpak-dip-development-platform-user-guide?r=1580801

vol. 23 of 32

"Connect USB cable to GreenPAK DIP Development Board, if everything is correct, you will see two blinks of green LED, and then LED becomes red."

I have blue LED.

[McArti0]

... and I suspect that it does not program OTP memory at all, but has only two modes of emulation and testing.

Such a non-programming programmer/ 

[robcfg]

Hmmmmm....

So you have the "programmer" which should be the GreenPak DIP Development Board and then the SLG46533v which you have to put on top of it, is that right?

Which one shows a blue LED? I cannot find on the docs any reference to a blue led anywhere and the SLG46533v only has one led for power and maybe an optional one, so I guess if there's any error, it's on the programmer's side.

Now, I don't know if you should be tinkering with OTP as it's One-Time Programmable.

Did you put the SLG46533v in the right position?

For what I've read in the manual, is there any indication on the GreenPak Designer software that either the programmer or the the SLG46533v are connected?

Pictures of the stuff would be welcome.

[McArti0]

Shining STATUS. Blue light. Not Shining POWER.

[robcfg]

If you power up the dev board without the SLG46533v attached and it still doesn't show the Power led, I'd say the dev board is broken and you should return it or get a refund.

Sadly, the documentation says absolutely nothing about the status led and its meaning.

[McArti0]

https://lang-ship.com/blog/work/greenpak/

here...   https://lang-ship.com/blog/wp-content/uploads/2020/11/image-79.png

is bottom text: GreenPAK Development Board [ID: 0x101] in use

and here... https://lang-ship.com/blog/wp-content/uploads/2020/11/IMG_7084.jpg

shinig both LED blue and RED

[robcfg]

As a last test, you could try to bridge manually, with a piece of wire or similar, the + side of the status led and the + side of the power led and see if it lights up, just to check that the led is not broken, or just check the voltage across the led.

[McArti0]

They gave me money for a broken programmer.
I bought a second one from Mouser.
I programmed the SLG chip in 1 second. Soft said correct.
I connected SLG instead of PAL on long 10 cm wires.
CPC showed a white screen with flashing black short lines.

it switches something but differently than in the simulation.

I am concerned about the length of the reset.
SLG needs to load the configuration from NVMemory to the RAM of SLG.

It still won't be simple  .

[Bread80]

My experience with the PicoGarry is that the CPC's reset pulse is more than long enough for it to startup, load the code from ROM and initialise everything before the pulse ends.

To test if it's the startup time or the actual operation causing the issue, install a reset switch somewhere. That way you can reset the CPC /after/ you know the chip has initialised.

[Bread80]

Looking at your simulation I notice sometimes it's generating CAS1 whilst CPU is high. When CPU is high the CRTC is addressing RAM and these accesses should *always* be to bank 0. I.e. a CAS0 signal.

Looking at your design this appears to be an issue with 3-L4. If I'm understanding correctly this  outputs either 0 or 1 to indicate CAS0 or CAS1?

Also, I can't make out what the circuit at the bottom around CNT3/DLY3 is doing. It's being triggered by an IO write to the config latches (Q0,Q1,Q2) from the ADR LUT. The ADR LUT should only be used to latch the config state.

I suspect you can fix this by setting the B input to 3-L3 permanently high? Possibly you need to be feeding CPU directly into it? Maybe latching CPU during a CAS?

[McArti0]

Dear Bread80. Read this topic.

My NewPal is supposed to be able to switch VRAM to Bank1 if Out (&7FFF),&11xx xxxx occurs twice in a row.

CNT3/DLY3 is a counter that detects whether the second OUT occurred quickly. I assume that no previous program code sends OUT to PAL twice.

Currently, after my tests, I have an unstable PAL socket. It only works when the original PAL is inserted a little crookedly.
Mechanics and contacts blocked my work.

Idea is here: https://www.cpcwiki.eu/forum/amstrad-cpc-hardware/banked-video-ram/msg231087/#msg231087

[Bread80]

My NewPal is supposed to be able to switch VRAM to Bank1 if Out (&7FFF),&11xx xxxx occurs twice in a row.

Okay. That makes sense if this is new functionality.

[McArti0]

DONE IT !!!

128kB VRAM for CPC6128. First time in This World.

Simple toggler CRTC for BANK0 and BANK1.

SuperHiperVRAM NewPAL <>

[SerErris]

Looks great, now you can run interlaced 640x400 ... wohoo...