An idea: an icomp-branded quality case for ACA500Plus (with add-ons)

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 think that one thing ACA500Plus really lacks is a decent case. It's a super versatile accelerator but when you just pull it out of the box and connect it to your Amiga, it just looks quite ugly, and I think it may really put some people off. Also most cases available online are just crappy, at best, and look out of place.


    But it doesn't have to be like this, does it? Recently, I finally got around getting a 3D-printed case for my ACA500Plus. It's this one from Thingiverse: https://www.thingiverse.com/make:683677. It looks really nice, even though it's not perfect and it's 3D-printed. Check out the attached photos. It also did require some tweaks - there are no spacers to mount the card, no mount holes for additional A1200 card, one of the supports for the top of the case would block the A1200 card, so I had to cut it off, there's no mount point for an RJ45 socket (for X-Surf 500), there's no way to reach the ACA500Plus button, and so on. But if all these issues with the design were addressed, and if it was properly made via injection moulding, I think it would be a hit. As a bonus there could also be an ability to mount LEDs in the case and connect them to the accelerator, but I'm aware that this would require serious messing with the accelerator, so maybe it would be asking too much.




    Nevertheless, I'd buy one for sure, even though I have mine. What do you guys think?

  • That would be two huge toolings, each of them in the 30k EUR area, plus a smaller tooling for the front part, probably 6000-7000 EUR. Make that a round 75000,- EUR for the first sample run, not conting the engineer who makes flow analysis and of course removal of that toxic logo. Another guesstimate for cycle time would be 20 seconds for the big parts, maybe 5-8 seconds for the small one.


    Two injection molding machines would be required, as the big parts need more clamping force. No guesstimate here, as I only have experience with much bigger and much smaller parts. Making 1000-2000 parts would be minimum, which is already ambitious, given that you're coming close to 100% market saturation, requiring each and every customer to be willing to spend about as much on the case as he/she spent on the card. Let's assume we want to pay the toolings with 1000 parts, then it's 75, EUR per part. With some luck, a service company will make these three parts for a combined 15,- EUR (including shipping), then we need to add at least 30% for QC and labour, adding another 27,- EUR.


    Let's assume that this will be shipped to Hungary, then there's 27% VAT on the 117,- EUR figure, resulting in an end-price of 148,59 EUR for this case. Note that all the back-of-the-envelope calculations are made with "zero profit" in mind, only attempting to reach a "black zero". It's not like the bargain of the century, when I have been able to invest less than 100k EUR in the C64c cases for procurement, transport and maintenance/upgrades to them to make them work on modern machines.


    This is completely ignoring storage cost and tooling cost for shipping boxes, which should be designed to be as small as possible in order to make storing "hundreds" of these cases viable. If you buy off-the-shelve boxes, you'll waste lots of volume and overpay on international shipping, as air freight is usually calculated by volume, not weight. So... expect to add another 10,- EUR for cleverly designed packing, and then some for mount material, such as thermoplastic screws.


    So.. mass-production has different rules and requirements. Haven't looked at the wall thickness at all (not a 3D CAD user), but that has a strict requirement of not varying more than a given percentage.


    To summarize: I have my doubts that a case that denies access to the DisMo and networking card will be that popular, given that it will double the price of the expansion.

  • So.. mass-production has different rules and requirements.

    Yea, the costs you mention sound hardcore. I didn't expect it would be THAT expensive.

    Probably not feasible at such costs unless they could somehow be reduced.

    Quote

    To summarize: I have my doubts that a case that denies access to the DisMo and networking card will be that popular

    As I wrote in the OP, the design issues would surely have to be addressed. An RJ45 port mount point could easily be added on the back. There could also be an opening or something similar to access DisMo, or at least the button. DisMo could even be relocated by using a ribbon cable to connect it to the accelerator.

    But yea, looks like given the costs it will remain a 3D-printable thing only, unfortunately.

  • DisMo could even be relocated by using a ribbon cable to connect it to the accelerator.

    No ribbon cable. Neither the address, nor the data lines are buffered, so it should just stay in place.


    But yea, looks like given the costs it will remain a 3D-printable thing only, unfortunately.

    ...which is better, given that there are two DisMos with the button in a slightly different place. That would also have to be covered in a molded version. Something transparent that "klicks in" would be small enough to be made on a double-tooling. The customer would click in the "window" that he/she needs, and shelve the other. A more sophisticated use of "the other" may be to have it "click" into the spot where the RJ45 opening is: I wouldn't care if the cover for the unused RJ45 is transparent, as you don't see it directly anyway.


    As for cost reduction, yes, such plastic molds are crazy-expensive. However, those were guesstimates, given that you want to make a hardened-steel high-precision mold. Since we're talking under 3k units anyway, these molds may easily be cut from Aluminium-7000 alloys, which are good for up to 10.000 cycles. You would only need a hardened steel tooling if you're planning six-digit quantities, which is insane for the Amiga market.

  • If you add a cable, none of the signals has a ground reference any more. Note that the ACA500plus *and* the DisMo are all multi-layer boards with dedicated power- and ground planes. The ground planes serve as ground reference. Only dual-layer foil cables or flat cables where every other wire is GND may be up to the task (such as the UDMA-66 cables that you may remember from PCs).

  • If you add a cable, none of the signals has a ground reference any more. Note that the ACA500plus *and* the DisMo are all multi-layer boards with dedicated power- and ground planes. The ground planes serve as ground reference. Only dual-layer foil cables or flat cables where every other wire is GND may be up to the task (such as the UDMA-66 cables that you may remember from PCs).

    If I understand correctly, signals passed to DisMo are very susceptible to noise, that's why these special considerations would be needed? I thought it was just a 2-digit display bundled with a momentary switch.

    And what if I only relocate the switch? For example by soldering 2 wires to its legs (without desoldering the switch) and connecting them to another switch located elsewhere? Would that be ok and not void my warranty?

  • If I understand correctly, signals passed to DisMo are very susceptible to noise, that's why these special considerations would be needed? I thought it was just a 2-digit display bundled with a momentary switch.

    There's a micro controller on there, and it talks to the main system through address and data lines. The data line buffer is on the DisMo, so one direction could be considered "buffered", but the other direction (on the same lines) isn't.


    And what if I only relocate the switch?

    That will work without problems - the "voiding warranty" part is always tricky. Strictly speaking, any soldering or mechanical modification will void warranty, and the longer the cable, the higher the probability that it picks up a static zap. While today's chips can mostly withstand the 2kV human body model, this model is not always accurate - there's a reason why other (non-commercial) equipment has different test standards. So again, if you want to use a cable, use lots of shielding. A small coax cable will have more protection than two individual wires. Outer conductor wired to GND, that is (yes, one of the button signals is GND).

  • I also ordered the PlexiLaser case for the ACA500plus, and I'm in the USA. The hole in the top for pushbutton access didn't line up, because my version of the ACA500plus was slightly older. It's not a big deal for me, but it's worth confirming which revision you have of the ACA500plus with PlexiLaser when ordering. Overall, it worked out great for me, and the top of the case was a convenient 'platform' on which to mount my Gotek drive. I talked about all this in my A500 restoration video here on YouTube, in case anyone is curious:


    https://youtu.be/xRTThZvt95I


    Enjoy!

    Jeremy

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