Right angle PDS adapter for SE/30

I explored this last year...

these are hard to find, but I remember seeing a page that explained how to build your own.

My solution? Remove the floppy and HD bracket... remove the floppy and the bracket altogether (who needs it if you are networked?) put the naked HD on its side at an angle next to the PSU, there are holes available in the main chassis such that you can get 2 screws into the HD from underneath the chassis to secure it. (I also put a HD vertical, if facing the back of the exposed SE/30, on the lip on the left... you can drill extra holes in the metal bracket to secure it with at least 2 screws). The last part of the puzzle, how to cover the empty floppy slot, I leave to you. I have found one actual slot cover (from an SE FDHD), it has a nice LED window for the second HD's LED... if you can get into my image gallery, I think I have some pics there. I need one more of these, and I think I'm going to have to construct it from, I think maybe, the lid of an old Performa 400, you know, just rip one strip that fits. If you need pics of the inside HD on its side, let me know, I'll post them.

digikey.com has 'em, $5.32 each. If that link doesn't work, search for AMP part number 650874 (this link is to a Technical Drawing pdf.)

IIRC you have to mod the keys (on the 'adapter') to reverse the orientation of the card.

dan k

Too bad you didn't need a right angle PCI adapter, cuz I have one of those. It's a genuine Apple piece, too.

Now, if I only knew which Macs actually needed such a gizmo.

"You're search has expired", so what terms did you use for the search? Has Mouser got the same thing? I might have a look at my hardcopy catalogs later.

in removing the HD & Floppy bracket, and turning the HD on its side, there are no permanent mods there... The chasis has a couple holes for this already. In covering the floppy slot, as long as you don't use glue or something (actually, scotch tape works fine), its not permanent either. These mods can be entirely reversed and will not reduce the 'market value' of your machine in the slightest. As far as the second HD on its vertical side... I actually didn't drill any holes, but I did only use one screw to secure it (screwed really really tight). If I ever move or something, I will disassemble that... to make sure it doesn't inadvertently smash the CRT... but if dropped hard enough to do that, the CRT is toast anyway.

Being able to return the machines to original condition was an important aspect in the deployment of the mods I did. After all, if you're sticking a "processor accellerator" in there, you're not really accellerating the processor... you're bypassing and replacing the processor (even though it still sits there), the heart and soul of the machine. But I guess if you can live with 16MHz...

I'd love to show you a picture of what it looks like... but my gallery here is MIA... what's up with that? Why can't I see my pictures?

(edit)
I just remembered... to get the adapter card (for the proc accellerator) in there without the 90° adapter... you do have to slice a tiny bit of the chassis... most would tell you its negligable. There are enough trashed SE/30s available, I think, to have a back up "intact" chassis just in case (what, I have no idea).

But my (well, its not really 'mine') hack works for the network adapter, which doesn't need modifying the chassis (but you do have to move the HD & Floppy bracket out of there).
When I have time to pull the both boxes open that each have one of these mods, I'll take a pic and post it here... I think you'll agree slicing the chassis in one spot is worth the extra 34-44 (or 100) or so MHz you gain with accellerator.

other 68k machines are not SE/30s...

I like IIci's, but I have no desire to rip out the inards, and cluster a bunch of mini mobo's inside it.
The SE/30, however, and not the SE, or the Classic, or the Color Classics, or the Plus, 512K, or 128...
I love that box so much... I almost don't care what's in it (almost... I really don't want an x86 in there unless its running OS X). If I had the $, I'd figure out how to get an LCD working, and just how many mini (or 12"pb or iBook) mobos could fit in there... all clustered together... with as many tiny HDs as it would take... and figure out some supurb cooling scheme... I think it would be grand.
...but I digress.

Of course, of course... as you like.

I will tell you then... either make your own right angle adapter (pin outs here--I am remembering a little more about the page I mentioned earlier... the guy used 4 smaller pinned connectors, cut and glued 2 together to get 120 pins... I think you get the idea... not pretty, but who cares, it worked) because finding one is near impossible. Ok, its impossible. Or shell out the $$ for the specialized board:
The Stratos TwinSpark Adapter!

Source

last I heard, they were still available, and still about 180 American clams (too difficult to think in ¥, what is that, like 23K¥?? Numbers too big to deal with)
Its an elegant solution, though. Personally, if I had one, I still rip out the floppy, cover the slot, and put another HD in its place.

Are you kidding? $180 American dollars for this?

Does this thing ship first class on a Lear Jet or something? That's well in excess of $200 Canadian dollars! And for what? There's really nothing to it! I'd say that there's less than $10 in parts there.

In Japan, that's just half the price of a good espresso...

sorry to dig up such an old topic but has anyone found a source or have some really hi res photos of both sides of the twinspark adapter so i can make my own, i refuse to pay 23,000 ¥

High res photos of both sides of a TwinSpark wouldn't do you any good. It's not the traces on the board that are magical, it's the code in the Programmable Array Logic chip. That is what you'd have to reverse-engineer, and no pictures on the internet will give you that.

If you don't want to spend 180 dollars, get a DayStar IIsi adaptor. If you're really serious, get a DayStar IIsi adaptor and reverse engineer the PAL on it... that's what the artmix.com guy did.

If I had done it, I'd have released the code on the internet so those who were willing and able to program their own PALs and etch&stuff their own PCBs could build it themselves. The artmix.com guy decided instead to make money off his effort, which is fine; more power to him. You see that he's done the work and I haven't...

I've moved this thread out of the Wanted forum. The discussion's been interesting, so here it can continue...

So does anyone know what the adapter was exactly? A description will do. Gamba says it's a "120 pin right angle female connector". The drawings look like he just used a male connector and stuck the PCB mount pins into the ethernet card. I don't know how secure this would be.

Perhaps just modding the MacCon would be a more secure connection. Any thoughts? Can anyone share how they made a right angle connection and how secure it is?

Yep, it's just a rightangle connector with its PCB pins stuck into the PDS slot on the board below.

I've never done this so cannot say this with any authority, but I reckon it ought to work OK. I mean you can always strap the RA 'adapter' on there somehow to keep it secure, how hard could that be for a dedicated hacker?

dan k

I bought a couple of adapters off Gamba about three years ago. They are right angle PCB connectors (Eurobus, VMEbus ???) but the fit in a PDS slot is OK. Once your hack is finalised, use hot glue or epoxy to make the connection more secure. Electrically the connection is fine but you need to ensure that it doesn't open up when the Mac is moved etc.

I found the connectors in one of the UK electronics parts catalogues (Farnell or RS, I suppose) but it was cheaper to buy from Gamba.

Phil

see my post, the second reply from the top, with link.

dan k

Yes, it's a simple board with three connectors, a PAL and what looks like a few caps.

However, I'd be surprised if the guy built more than a couple of hundred of them. In those quantities, the circuit board probably cost $15 - $20 each to have produced. The connectors are $3 - $7 each. The PAL and caps are mostly negligble, call it another $1. So there's $30 - $40 worth of materials there.

Then figure the guy probably put in a couple of hundred hours of skull sweat reverse engineering the Daystar board (assuming a friend didn't slip him the HDL code), laying out his board and taking measurements to fit it in the SE/30 properly. If you assign one of those hours to each of the speculative 200 boards, that's another $30 - $50 depending on how you price his time.

Then there was probably a proto-type or two plus test equipment and other sundry. And assembly of the board (soldering the parts down) costs something too, either in his time or to pay an assembly house to do the work.

There's the time-cost of his money for sitting on the un-sold stock, assuming he didn't make his investment back by selling sufficient boards immediately.

Finally there's the overhead of maintaining his website-and it takes a surprising amount of time to take and order, confirm payment, package the item, address and ship it out and then deal with any inquiries and/or problems from the buyer.

Given all that, I'd say that $180 is not so unreasonable. It's more than any of us want to pay, but it's not out of line with his costs.

If he had a market for 10,000 or even 1,000 of the things, it would be a very different story, but for a hundred or two hundred, the costs are high.

I had a couple hundred DIMMs run off once. That's a board about 1.1" X 4.5" and they cost me $6.50 each after careful shopping and negotiation on price. I could easily have paid $8 - $9 each.

Custom DIMMs? For some very special machines? I could see some pioneer doing a short run of some desireable setups, such as 128MB DIMMs for (5|6)500 Takky conversions or 256MB for Beige G3s, etc. If you can get your costs down far enough, and have a guaranteed eBay market or such... Basically bringing old machines up to the most current capacities available with their odd RAM setups.

here's a vote for a run of 128MB 72 pinners

will they even see the full 128 on the chips if they are added up (SIMM slot wise [#'s of slots])? or do these already exist?

RAMSeeker.com lists 64MB 72-pin SIMMs at $15 from 1-800-4-MEMORY, ATM. I'd guess a 128MB SIMM would need to cost about $30-40 to be priced into the main market. The hardcore pre-PCI Mac user might pay more than $40 in special cases. It might be worth a small chunk of change to somebody to get 128MB in a single slot 575/630 board or better capacity on the two-slot models like the 580/631/640/6300 series.

128MB 72 pin SIMMs already exist commercially, so there is no need for a custom run. I have about twenty of them on hand. I have tested them in the Q605/LC/P/475/476 and they work great. I have not tested them in other machines yet. I have Q610, Q650, PM6100, PM7100, PM8100 and Q63x motherboards on hand for when I get around to testing. I don't have a chassis for the Q63x board though (anyone in Austin, TX with one?). I have Power Supplies for all the others. The Q650 should be telling for the Q800 since they're essentially the same board. I guess I have a Q840AV I can test too. But all that's a ways down on the list of projects. Right now it's time to plant potatoes.

BTW, they are SIMMs, not DIMMs. Really just a nitpick. DIMMs have different electrical connections on each side of the connector. On SIMMs, the connector pads opposite each other are electrically identical. So a 72 pin SIMM has 72 pads down each side of the connector. A 168 pin DIMM has 84 pads down each side of the connector, but each pad is electrically unique (or can be, there are some duplicates, such as power and ground).

Not what I built, but one interesting idea would be to build 256MB FPM DIMMs for the 7200 family. The Apple Hardware Developer Notes claim that the 7200 can support a 256MB DIMM giving it a 1GB RAM capacity, but as far as I know, noone ever built an FPM DIMM larger than 128MB.

I doubt that anyone with a 7200 would be all that interested (are there any 7200 users left?) but this should also apply to the Power Computing machines based on teh Catalyst chipset--everything except the PowerWave, PTP and PowerBase.

Is it possible for the mod to split the thread at this point (or a smidge earlier )? I'd like to discuss DIMM possibilities for early PCI Macs but don't want to hijack a great thread about SE/30s.

Phil

I bought a couple of adapters off Gamba about three years ago. They are right angle PCB connectors (Eurobus, VMEbus ???) but the fit in a PDS slot is OK. Once your hack is finalised, use hot glue or epoxy to make the connection more secure. Electrically the connection is fine but you need to ensure that it doesn't open up when the Mac is moved etc.

Finally an actual description of the adapter. I thought maybe there was something more special about it. Thanks for the info.

Kreats,
Did you get any takers? Mouser had all the needed eurodin parts, I've just been waiting until I had a little play money to make a right angle adapter. I'd be happy to try and engineer a cheaper shipment to you.

Maybe I can save up and get Artmix's adapter, in the meantime a cheapie 90 deg adapter without my accelerator will have to suffice. One of these days, that PAL is mine

Maybe I can save up and get Artmix's adapter, in the meantime a cheapie 90 deg adapter without my accelerator will have to suffice. One of these days, that PAL is mine

Are any of you guys game to try to reverse engineer the PAL? I have a IIsi adapter but don't know where to start.

Kreats,
Check your PM, I'm back in the saddle!

Tt,
In reference to the PAL... yes I'd love to reverse engineer one, but I don't have the adapter, and haven't had any luck locating one. I have also not ever tried to work the program out of one, but have had some guidance from some folks that should know. Not that guidance is a replacement for experience, but I'm game to try.

Another Fritterer was kind enough to give me a supposed schematic for the adapter. My plan was to take the datasheet for the PAL/GAL, the schematic for the adapter, and a description of the two sets of PDS ports and try to work out a probing plan where you brute force test different inputs and look at the outputs. I'm sure that someone will weigh in a tell me that won't work. I did notice a gentleman in another Applefritter forum (the GSE-Reactive guy that's marketing new "old" cards for IIe's and IIgs's) has been reverse engineering like a mad man. I figured I might send a polite note to him and see if he'd give me some general pointers.

I looks like you could make an automated prober (I guess depending on the number of I/O lines) from a PC and its printer port. Of course, then the trick would be sifting through all of your data and coming up with a model.

And then maybe I'm outta my head??

I looks like you could make an automated prober (I guess depending on the number of I/O lines) from a PC and its printer port. Of course, then the trick would be sifting through all of your data and coming up with a model.

And then maybe I'm outta my head?? :)

Yeah, someone on LEM was talking about using LEDs and doing it manually, but I was hoping there would be a device I could plug the PAL into and just let the computer do the rest.

The GSE-Reactive guy has mentioned reverse engineering IIgs accelerators. To do this you have to reverse engineer GALs and my guess is that he is not doing this bit him/herself. Ask GSE-Reactive who does it for them. Bill Garber has been very open on comp.sys.apple2 sabout other stuff so just ask.

charlieman,
I've been meaning to write them for the last couple of days and see what the story was... they do seem pretty open about things, as I gathered from the thread in the Apple II section. You are probably right, they have probably found a place that offers such a service, and easily justify the expense from the sales of their end product.

I'll pass along what I find out. I'm beginning to wonder the possibility of starting some group effort on this- once I can cajole a volunteer to supply a IIsi or se/30 adapter. (Good luck) We've got some good talent on Applefritter...

mike