27C400 EPROM emulators (beta) issue D technical discussion

General discussions or ideas about hardware.
nivrig
Posts: 16
Joined: Mon Jul 26, 2021 3:42 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by nivrig »

My script is just a modification of Keir's :) His original single-send version hung for me as well. I want to use Python rather than an app because I want to script uploads from makefiles and so on.

My script takes about 6s to upload a 512Kb image (or maybe that was a 256Kb image). It was able to successfully program one of my devices with four Kickstart ROMs for my Amiga 500.
nivrig
Posts: 16
Joined: Mon Jul 26, 2021 3:42 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by nivrig »

FWIW, I got a HC333 rev 2 accelerator (40Mhz 68000, 8Mb fast RAM, IDE) for the A500 6A the emulator is installed in.

With the original CPU, the machine works fine with the emulator and I can happily switch Kickstarts. With the accelerator and the original Kickstart 1.3 ROM, the machine works fine and the accelerator RAM passes memory tests.

BUT with the accelerator and emulator installed together, the machine hangs on boot with any Kickstart 1.3 - 3.2. I get some of the usual Kickstart screen colours but it never gets to the "insert disk" screen even when I left it sitting for an hour or more. Occasionally I got a yellow screen ("the CPU encountered an error"), once or twice it booted to an "Expansion Card Status" (something like that) ROM screen showing an expansion as "defective".

Did you ever get any reports like that? I've no idea what the problem might be. Some timing issue?
Are any A500 accelerators known to work with the emulator?
cmorley
Posts: 291
Joined: Tue May 28, 2019 5:46 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by cmorley »

If you're running an accelerator and the address bus is not changing regularly (you're running an accelerator) then the auto switch won't work. Run a wire to the reset pad R from the KB_RESET or RESET and change the ROM switch mode to disable "wire free" ROM switching.
http://www.boobip.com/hardware/27C400_E ... uide#opt-w

edit:
Ensure ROM switch pull up is OFF to avoid damaging the STM32 input when using a 5v signal (e.g. RESET line).
Alternatively you can add a dedicated switch for ROM switching between R-GND (turn ROM switch pull up ON).
nivrig
Posts: 16
Joined: Mon Jul 26, 2021 3:42 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by nivrig »

I disabled wire free switching and pull up. I don't have a switch connected yet so I set the inital ROM to the KS1.3 slot 0.
I set the HC333 to use it's 1.3 compatible Oktagon IDE ROM. Still hangs on boot :(

In general, does the wire free switching not work if fastram is fitted and the CPU is running from that instead of the motherboard bus?
cmorley
Posts: 291
Joined: Tue May 28, 2019 5:46 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by cmorley »

nivrig wrote: Fri Sep 24, 2021 10:21 am I disabled wire free switching and pull up. I don't have a switch connected yet so I set the inital ROM to the KS1.3 slot 0.
I set the HC333 to use it's 1.3 compatible Oktagon IDE ROM. Still hangs on boot :(
If you disable wire free ROM switch but don't connect a reset wire then you have to enable the pull up. You can't leave the input floating.

From my manual:
To disable ROM switching completely:

enable ROM switch pull up
disable wireless ROM switch
In general, does the wire free switching not work if fastram is fitted and the CPU is running from that instead of the motherboard bus?
In a word, yes. If the address pins of the ROM socket are not changing regularly the module will treat this as in RESET when wire free ROM switch is enabled. If the CPU is executing from fastram all the time on a board with buffered address lines then the motherboard address lines will no be changing... ergo the module will think the Amiga is in RESET.

If you have a non-simple setup then you need to attach the reset wire or switch.
nivrig
Posts: 16
Joined: Mon Jul 26, 2021 3:42 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by nivrig »

To disable ROM switching completely:
enable ROM switch pull up
disable wireless ROM switch
Done, and reflashed with a fresh 1.3 image just in case. Same :/
cmorley
Posts: 291
Joined: Tue May 28, 2019 5:46 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by cmorley »

Double check you are in 27C400 mode (rerun the preset wizard if necessary). Turn off autosave initial, set the initial ROM to zero, program the same ROM in all 4 banks, disable wirless ROM switch and enable ROM switch input pull up. Commit config changes to FLASH.

If that doesn't work then I don't know. ROMEN comes from Gary. Address & data are the 68000 bus. Not much to it really.

Is there any maprom function on the accelerator?

All it does is emulate a ROM... there's nothing more to it than that. So as long as it has all the inputs it needs (nothing floating/misconfigured) then it should pretty much work in anything. :|
nivrig
Posts: 16
Joined: Mon Jul 26, 2021 3:42 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by nivrig »

Did all that, and tried a different emulator board set up from new too.
Verified my rom images matched byte for byte with a known-good image someone sent me.
Used program & verify to load all slots with the same image.
Same :(

The accelerator is a HC533r2
https://bigbookofamigahardware.com/bboa ... px?id=2058

There's an IDE driver ROM on board but no maprom function mentioned.
Some mention of possibly strange AutoConfig behaviour on that link.

But yeah, it's weird. No idea why it wouldn't work when it works perfectly with an original 1.3 ROM.
It sits at a blank screen but something is sort of working because it seems to get through a few KS startup phases/colours. It varies though, so it's like some kind of random glitch.

I'll pick up an original KS2 ROM and maybe a physical ROM switcher perhaps. I'll also fire an email off to the seller and see if they've heard of anything like this.

:roll:
cmorley
Posts: 291
Joined: Tue May 28, 2019 5:46 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by cmorley »

I don't have one of those cards so can't perform any tests. My suspicion would be that the bus activity on the card is causing Gary to glitch /ROMEN... the EPROM emu is a lot faster than an EPROM and can drive more current... if /ROMEN is glitching it might be causing spurious data collisions on to the bus where an EPROM doesn't even respond.

Pure speculation without any measurements.
nivrig
Posts: 16
Joined: Mon Jul 26, 2021 3:42 pm

Re: 27C400 EPROM emulators (beta) issue D technical discussion

Post by nivrig »

I've been in touch with the designer of the HC333 and he's going to do some investigation. I have sent him one of my emulator boards to help.
Post Reply

Return to “HARDWARE DISCUSSIONS”