Posts by Jens

Again, the eMail address you're using here in the forum is not registered on the shop system. What's the user name you have there?

Jens

Nothing written, no. Implementation is really dirty, as you need to give /Slave before /AS is valid. That's all there is to it - the shortened access is then handled by the Z4 board.

Jens

We're currently not planning to add such things to the store. I could think of something simple that could be made from sheet metal. Mounting on the Docking Station will be easy, as the base is wood, where screws can easily be inserted.

Jens

Right, your 10-year old Full-HD TV may either do nasty scaling, or it has "movie enhancements" switched on in it's settings. So you should check both, the Indivisin AGA MK3 settings AND the TV's settings.

Jens

Quote from Rms68-2

is the shop a different account than forum?

Yes, shop and forum are separated. Unfortunately, the shop requires you to remember your username. I do not see a shop member using the same eMail address as the one you've used for the forum, so you should be able to just make a new account.

Jens

We're still short-handed, many employees on sick leave. I hope we'll have time for a few returns this week.

There is not much new in this version - basically a source clean-up that works the same as before, but testing can never hurt.

We've also looked at the NetBSD sources, and they also only check for the PCI ID, no other magic before or after that. So it really boils down to "use the latest version, make sure you have mmulib installed". The underlying assumption is that doraemonppc neither uses the original P5 processor libraries (which specifically open the PCI address space), nor Thor's MMUlibs. The latter are called by the driver to make sure that PCI space is open, and no bus error is thrown.

Jens

Since the original thread starter didn't respond in over a week, I take it that the problem is solved.

Quote from Rms68-2

I have a 4000/040

If it's the original 3640, then it's not exactly fast. Only accelerators with local fastmem can give you good performance. Still, the experience you're describing indicates some "handbreak" that needs to be released.

Quote from Rms68-2

I could not get your AmiTCP to work

With the little interaction our automatic installer requires, that's hard to achieve. Please try again and open a support case if it fails. If you're a recent customer of the X-Surf-100, then the AmiTCP install disk is part of the package, just go to your order history and download the file.

Jens

I'm pretty sure it came with 25MHz CPU clock and 50MHz RAM clock. I asked you to take a closeup picture with macro functionfor a reason - I want to see the solder spots. However, your last picture still doesn't have the required detail. Please use more light - ideally daylight.

Jens

This week's three-day meeting with Peter revealed that the 68040-40 CPU violates it's datasheet in terms of "when does it take over data". As a result, we've decided to re-factor the one thing that's at the core of the design: The SD-Ram controller. Tweaking the card for top performance is currently too tedious, and going beyond 40MHz for the 68040 might stretch that "violating" timing even more, requiring modifications on SD-Ram timing that are currently too difficult to make.

Don't worry, this is not "back to the drawing board" - but it's definitely a delay in releasing the cards to beta testers. May still happen this year, but it may also be January 2024.

Jens

This indicates that @doraemonppc 's hardware is different. Since this is built on Cypress CPLDs (programmable parts), it's going to be lots of trial and error to find out how it's initialized. To my knowledge, the contents of these CPLDs haven't surface anywhere (yet?).

Jens

Sorry, no timeline. I am not taking any money if I can't name a specific delivery date - you may know this policy, and the general warning to look really close if something crowd-funding-like is offered. Too many people in the Amiga/Retro market have lost money to false promises. I will do everything I can to not associate such an experience with iComp.

Jens

That means that file size wouldn't be an issue. We are currently on the software for the new Buddha Plus One, and Timm will soon move from work on the ROM firmware to doing tweaks on the DOM installer. While in that process, he may be able to upload the CD installer in a different format to the WIki.

This might take a few days. Please understand that this is a bit outside the scope of normal support, as you don't have everything that was shipped with the product. We try to make products as complete as possible, so you don' thave to buy any additional gear for using them, but if parts get taken out of the equation without iComp being at fault, we can't handle such a case at top priority. There is a solution, and we'll help - it might just take a bit longer.

Jens

Quote from graysters

This one was probably my fault for using the FPU socket as a handy place to poke wires into so I could hook it up to my scope while I was debugging the failed CPLD..

Most sockets have a spec for max. 0.9mm thickness of the pins. The 68030, 68040 and 68060 CPUs have exactly that maximum size, so they are difficult to insert and remove from any socket. I stopped using sockets for the 68030 after the ACA1230, because it is a) too expensive and b) direct soldering helps with heat dissipation. Only on the ACA1240/1260, I'm going back to sockets, because that one has an active cooling system, and people might have a legit reason to change to a different CPU.

Congrats on this resurrection

Jens

hawkman, please mention the order ID that you've bought the Buddha controller with. We should try to improve your situation with a new flash chip.

Jens

What methods do you have for file transfer to your Amiga?

Quote from Leone

I only wish Jens would consider a "limited edition" 56 MHz version.

Sorry, that's not going to happen for very basic practical reasons: The CPUs that were stable at 55.555MHz were selected from a large batch of 50MHz CPUs in a long and tedious process. The amount of CPUs that can't run at this pace for extended periods of timeis very low, and I have had lots of 50MHz CPUs sorted out ever since my very first ACA1230 accelerator has hit the market. I am not going to do this again, as it would make both the 50MHz version and a faster version significantly more expensive.

If you want more speed, the ACA1240 will provide that.

Jens

Quote from hawkman

I'd like to add a note to this thread - the Buddha card must be in the furthest right slot (next to the processor expansion slot) for this to work. Otherwise the machine will lock up. I assume this is because the card has not been autoconfigured.

That's not correct. We use that exact same procedure for initial bring-up of the cards after production, and we use any slot of an A2000. The computer is booted from a floppy with the two files (flash tool and binary), along with the command line in startup-sequence. I developed that procedure some time in 2017 and specifically wrote into the instructions that during this mass-bringup, all slots should be used for some wear levelling.

The best explanation I have for your observation is that there was some unconfigured card before the Buddha (closer to the CPU slot), or there was another Zorro card after it that you needed in order to bring the computer up. TBH, this is such a rare event that it's not worth investigating any further.

Jens

I did understand correct, and your question was already answered.

The marked pins are colour output pins from the ECS flicker fixer (or Denise), and they are not used on the A604n. Indivision ECS will detect that it's mounted on an A600 and go into "Denise shadow mode" where it sniffs the chip bus for data that is sent to Denise. The picture will be replicated in the flickerfixer, then written to the frame buffer and output at higher frame rate, just like all flicker fixers do.

The special thing about this is that you don't even have any direct picture data on the trapdoor bus. You only have the chipset bus, but that's enough to replicate the picture that Denise outputs. Once you've understood the concept, you should understand that the colour output pins on the Denise socket are not used for anything in this configuration.

Jens