1233n to 1234 upgrade

Caution: Non registered users only see threads and messages in the currently selected language, which is determined by their browser. Please create an account and log in to see all content by default. This is a limitation of the forum software.


Also users that are not logged in can not create new threads. This is a, unfortunately needed, counter measure against spam. Please create an account and log in to start new threads.

Don't Panic. Please wash hands.
  • No worries at all, that's why I didn't link them in the original post. Please feel free to delete the others too if it avoids the possibility of generating further grief for anyone.

    I am happy to make stuff if there's nothing else available but if there might be at some point I am also happy to wait. I've made a couple of those open-source unbuffered interfaces in the past, because there was nothing else available, and they have worked okay for me but I know not for everyone. I haven't the skill to design a buffered one myself, and no time to learn how, so I'm more than happy to try and help support those who do.


    ...In fact thinking about how that type of device is attached straight to the data bus, usually with a wonky, cheap CF or CF adapter plugged into them, that might well explain why a lot of real Amiga 1200s glitch out randomly every now and then, my own included. The same software image run on WinUAE is totally stable, as is my ACA500+ since I switched to using good quality real CF cards in it.

  • Not the "express" version (as the IDE speeders on the accelerators are faster), but a "2021" version with a CF card slot is likely. That's the "todo list" hint earlier in this thread :-)

  • Not the "express" version (as the IDE speeders on the accelerators are faster), but a "2021" version with a CF card slot is likely. That's the "todo list" hint earlier in this thread :-)

    I vote for that option and am interested in one. :) Especially if ACA1240 and ACA1260 will not have IDE port on them. If I remember in one of your comments somewhere, you have said that they will not. Not yet sure if I'm interested in 040/060 route or should I stick with my B1230 but in any case, some stable buffered IDE option is needed in any case.

  • I vote for that option and am interested in one. :) Especially if ACA1240 and ACA1260 will not have IDE port on them. If I remember in one of your comments somewhere, you have said that they will not. Not yet sure if I'm interested in 040/060 route or should I stick with my B1230 but in any case, some stable buffered IDE option is needed in any case.

    you never know, looking at the surprise CF card slot in the ACA1234 maybe the same suprise will be on the ACA1240/60?


    CF slot is not the same as IDE header and would solve the “wobbly contraption” issue that Jens has with an IDE port on an accelerator 🙂

  • maybe the same suprise will be on the ACA1240/60?

    Nope - there's just no space for a CF card slot on the ACA1240/1260.


    CF slot is not the same as IDE header and would solve the “wobbly contraption” issue that Jens has with an IDE port on an accelerator 🙂

    I'm thinking of making a rigid connection to the 44-pin A1200 IDE, not a cable, so "wobbly" is not what I have in mind :-)

  • Hi,


    spent some time on this today and ordered a sample PCB, which is scheduled to be here in early August. So.. I can "almost" tick that off the todo list, and hope that it actually improves IDE speeder success for some of you.

    This is a proof-of-concept. The design is taken straight from my older (and very successful) buffered IDE interfaces, so I expect it to work right away. However, we need people who have a setup that corrupts the CF card if the IDE speeder is switched on, who can test and confirm that a buffered adapter will really improve the situation.


    If it does, I'll do some changes to the board to make it mass-producable. It currently has SMD components on top and the 44-pin IDE connector on the bottom, which is fairly expensive to make. OK for a prototype, but not for mass-production.

  • I am willing to try and experiment. Iwill be able to try CF, SSD and HDD latter being SATA working over streight adapter but showing the same behaviour: i.e. 2,6MBPS in pio 0 trouble free and coruption in higher PIO settings.

  • The adapter I've shown in the post above is only for CF. The other mass-storage devices you're mentioning will have to connect through a different buffered adapter.

  • Still willing to try and experiment ^^.

    I have ordered buffered ide interface as per your advice (more chips). I realy want to see what IDE soeeder cand do. I am curious because fast IDE adapters are not realy well designed and overlap with Indivision which I do not want to remove.

  • On a sidenote, if this buffered interface works as expected and layout of the circuit board is slightly changed to bring CF card over PCMCIA port (access from outside as provided in new A1200 cases) it would be THE BEST solution in the market by far (there are CF adapter which provide such lication but they are just eliminating cable and fix LED problem - no buffering)

  • it would be THE BEST solution in the market by far (there are CF adapter which provide such lication but they are just eliminating cable and fix LED problem - no buffering)

    I don't understand - you want access to a non-removable card from the outside to do what? I mean, yes, a CF card looks "removable", but in this application it is like the boot card of the ACA500plus: It should not be removed. You should consider it a harddrive - you don't change that on a regular basis either.

  • I "play" with various setups quite a lot. I have IDE to USB 3, SATA to USB 3 adapters and CF reader on my PC. Advantage is that I make fast backups on PC, I can write image files on PC or use WinUAE and most often I copy large amount of data to my Amiga (e.g. WHDLoad libraries - to update I download all few times a year, ADF libraries, CD32 and CDTV ISO images, other CD ISO files etc.). Connecting CF to PC is by far the most convenient and fastest way. A lot of users do it. Some users run long IDE cables to CF customised cover under a FDD next to mouse porr but most prefer shorter connection. End goal is to have convenience of acces without necessity to open the case. Not just CF, even HDD is removable 😁.


    My original 80MB HDD was 3.5" connected via handmade cable and it was always outside the case back in 1995. I would cooy loads of stuff on it on my friends A4000 and deliver graphics to him (we used to make Amiga games back at the day).

  • Been there didn't work for me. Firstly it needs working AmigaOS, it doesn't beat recovering large drives from backup *.img files or WinUAE folder copy. USB 3 + SATA HDD is sooo fast in comparison, especially when cooying hundreds or thousands of small files. On the other hand a lot of users do not have a desire, need ir skill to go down that road....


    At the end of the day.... not necessary if you don't like but I see growing popularity of it (A1200.net has this as a feature if their new cases together with USB port side mounting)

  • Been there didn't work for me. Firstly it needs working AmigaOS, it doesn't beat recovering large drives from backup *.img files or WinUAE folder copy. USB 3 + SATA HDD is sooo fast in comparison, especially when cooying hundreds or thousands of small files.

    I'm not talking about copying - that will be slow for the usual 10MBit card that you plug into the PCMCIA slot. However, if you mount a NAS share, you only transfer the files that are *currently* needed. And with WHDload loading everything with preload, shutting down the network within the WHDload script is the most obvious thing to do - if you want to avoid copying that amount of data over multiple hops.


    (A1200.net has this as a feature if their new cases together with USB port side mounting)

    If they had an option for a break-out *above* the PCMCIA for a CF card, I'd think about an extension to this product with a second CF card slot - again, similar to the ACA500plus, allowing software to switch power to the external CF card slot and having real hot-plug capability.

  • Quote

    If they had an option for a break-out *above* the PCMCIA for a CF card,

    This is exactly what they have. It creates convenient way to remove CF card, plug into PC, mount in WinUAE and bang on some fast transfers. Fir occasional transfers of small files I use CF to PCMCIA adapter and another (smaller) CF card. I am aware of a risk of removal of these cards. That is what I use as a hotplug transfer media.

    Quote

    However, if you mount a NAS share, you only transfer the files that are *currently* needed. And with WHDload loading everything with preload, shutting down the network within the WHDload script is the most obvious thing to do - if you want to avoid copying that amount of data over multiple hops.

    Oh, Jens. I don't see average user doing it. Most of them just stick CF card on the back using long IDE cable or break-out above PCMCIA even DIY on old cases. Cheapest and easiest. Buffered version baired with ACA accelerators.... now, that is something to look for.

  • Hi Jens.


    Thank you for an advice. I have acquired one if new buffered IDE interfaces (one chip though). I have 2.5" SATA drive with adapter. Out of the box speed has increased to 2.2MBs but after installing included drivers bumped to 8.8MBs. PIO increased in ACA Tool not IDE interface software. Next step is to replace HDD with M2 SSD (less power and less heating).


    I nedd to resolve PCMCIA not being recognised but it seems to be software issue because PrepCard does see it.

  • I have acquired one if new buffered IDE interfaces (one chip though).

    If it's one chip only, it is not buffered. You need at least three chips: Two for the 16-bit data bus, and either one programmable, or two discrete logic chips. Here's a first glance at the prototype we're currently working on:

    As you can see, I opted for the discrete-logic approach, because GAL chips are too hard to come by these days.


    We have a considerably large set of CF cards that don't work with an unbuffered interface, but do work with this one. And we're currently on tiny details, such as LED appearance, possible violation of the specs by the CF cards and the like. Our early tests show that there's quite a range of CF cards that cause trouble on unbuffered adapters, and there are likely even more problems if the adapter is connected with a cable. I am genuinely surprised about the high number of failing CF cards that work just fine with this adapter - from the results we have found so far, I'd expect a huge number of "compatibility issue"-threads in other forums. However, these might have been avoided by resellers pairing the right card with an unbuffered adapter. You only run into problems if you buy CF card and adapter separately.

  • I nedd to resolve PCMCIA not being recognised but it seems to be software issue because PrepCard does see it.

    On a side note, the next iteration of this buffered interface aims to decouple PCMCIA and IDE as well, as there is a possibility for PCMCIA accesses to confuse non-compliant CF cards. It's the same "total separation" that the ACA500plus uses.