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.
  • I'm sorry, but the only upgrade path we can offer is from the ACA1211.


    With the ACA1234 offering quite a few things that are either not usable (IDE speeder) or don't really make a lot of sense (yet another CF card slot) for owners of an ACA500plus, you'll have a huge market for offering your existing 68030 card.

  • The ACA1233n-55 is faster in terms of CPU performance, and possibly for the IDE speeder, which is fairly aggressive on the 55MHz card. The ACA1234-50 is more conservative on the IDE speeder, because you can get even more performance from the local CF slot, and you only have one speed to choose from for the IDE speeder, where the ACA1233n-55 lets you choose one out of two speeds.


    Further, the ACA1233n-55 offers more flexibility when "downgrading": It can do on-the-fly clock changes (where the ACA1234 needs a reboot for every clock rate change), and it lets you use memory for the 68ec020 CPU if the 68030 processor is switched off. I had to cut away "something" in order to achieve affordable prices in this chip-shortage environment :-).


    So if you have the ACA1233n-55, I wouldn't change anything on your setup. You can compare the performance with the AIBB modules that I have published for the ACA1234 cards (see Wiki).

  • Looking at the AIBB data it does indeed look like the 1233n is the faster card (glad I own one) however it's great that we still have the option to buy new 030 cards nowadays with part shortages and all that so the 1234 is a very welcome addition and I'm sure will be a hit.

  • I am happy with my 1233n 55 except for one thing: if I try to use IDE speeder my CF card corrupts and it takes a lot of effort to fix bad blocks.


    If you have any advice I will happily stick to my 1233n 55.

  • Can't give any advice without more information about the internals of your system: The IDE speeder must go through the Address/Data bus of the A1200, so anything installed there may have an influence - even Kickstart ROMs. I test the IDE speeder with a buffered IDE adapter only, as CF cards usually don't have a lot of drive strength on the data bus.

  • In my A1200 there is Only ACA 1233n 55 and Indivision AGA MK3. CF is connected directly via external CF adapter (no cable just angled PCB). ROM is 3.X not stock 3.1. I am upgrading to 3.2 now.. CF is 32GB with 3 partitions


    If I increase PIO mode it starts to getbad sectors. Is there a partition limitation or fike system recomendation? Maybe advaced setup in HDD Tool? I will backuo all and repartitiin it if needed.

  • CF is connected directly via external CF adapter (no cable just angled PCB).

    That's the most likely explanation: A CF card has "little to no" drive strength, so it takes a lot of time for data to become valid on the bus. I have been preaching not to connect mass storage to the A1200 IDE bus since 1996. Use a buffered IDE interface, and the probability of the IDE speeder working as expected are a lot higher.

  • Thank you Jens. I will try real 2.5" HDD ir SSD converter before I try buffered CF. Honestly, for general purpose of WHDLoad CF is more than enough (DOM maybe even more reliable) but I am confident that SD cards are not reliable enough.

  • Thank you Jens. I will try real 2.5" HDD ir SSD converter before I try buffered CF.

    OK, so you intentionally go against my recommendation.


    Honestly, for general purpose of WHDLoad CF is more than enough

    I don't see how the type of usage will affect drive strength of data bus drivers. It's either "things work by design", or they fail. A data transfer mechanism usually does not judge the type of data that it's transferring, nor does it rate the "degree of use".


    Use a buffered IDE interface. You can identify them on the number of chips: Three chips are required to really buffer the IDE interface.

  • Quote

    OK, so you intentionally go against my recommendation.

    Quite the opposite. I want to go away from CF cards. I have both 2.5" HDD and interface for M type SSD and SSD laying around to try.

    Quote

    I don't see how the type of usage will affect drive strength of data bus drivers. It's either "things work by design", or they fail.

    I didn't state otherwise but people looking for an occasional game in their Amiga are OK with those cheap solutions.

  • Use a buffered IDE interface. You can identify them on the number of chips: Three chips are required to really buffer the IDE interface.

    Would you recommend some concrete buffered IDE solution which is known to be good? Of course, to be also available for purchase. I guess you don't have something in your store? At least I don't see it.

  • They don't. However, I don't have buffered Amiga IDE interface. I am trying to find one to buy. However, I have managed to find only on AmigaKit but it will be a month before it arrives. I have to play with it in a meantime.


    To be clear: you say that CF can be accelerated on a buffered interface?

  • We currently don't have such a product, that's correct. However, I was the first to make such adapters - see the "MLC buffered IDE interface", released in 1996, and the IDE-fix/IDE-fix express adapters. There are over 20.000 of these adapters in the field that are my design (including non-branded ones sold through Eyetech in large quantities). Another 40.000 of the DCE-made buffered interfaces are in the market - sold through Vesalia back in the days. While the DCE adapters were not working properly with more devices (after all, it was supposed to support up to 4 drives), it will certainly do the trick in terms of buffering the bus for a single CF card.


    With such huge quantities in the market, it should be easy to find them used.


    What AmigaKit is currently offering is a direct knock-off of the "express" version of the IDE-fix adapters (not even giving credit, but that's OK - we have other professional pirates in the market).


    To be clear: you say that CF can be accelerated on a buffered interface?

    To be even more clear: The probabilities are much higher. I can't guarantee it, as I have no control over the CF card's design.

  • Is there anywhere else we might be able to get a buffered interface, or an open source one we can build? There doesn't seem to be a single one for sale anywhere on Ebay at the moment, and AmigaKit are, well... I would personally really prefer not to have to deal with them based on my previous personal experiences, and also their current seemingly hostile and overreaching behaviour towards others in the community.

  • I haven't heard anything bad about AmigaKit - any particular event that I should be aware of? Please link to a news page, don't describe the case here, as that may not be neutral enough.


    I've just dug out the old GAL sources - it should be possible to implement that with pure TTL logic, eliminating the need for a GAL programmer. Back in the days, GAL chips were dirt cheap, but they're not made any more. More stuff on the todo list...

  • Thanks - please understand that I have deleted the post, as it's not really news pages, but forum posts, where flamewars are cross-posted over different forums. I'd like to use this forum primarily for customer support, and as an archive that people go to when they are looking for a solution.


    The buffered IDE adapter can be made, but I fear that SMD soldering is not something that everyone can do. Note that CF card slots are only available as SMD parts. Further, there's a few tricks for a truly working buffered IDE adapter that I wouldn't like to open-source. I haven't looked at AmigaKit's knock-off to check if it's a straight copy, or if they "optimized away" the tweaks (which is likely, as it may look like an oversight to the untrained eye).

  • The last reply was more than 365 days ago, this thread is most likely obsolete. It is recommended to create a new thread instead.