Jump to content
Trash80toHP_Mini

Farallon ETHERMAC LC NSC w/NuBus drivers in the SE/30 PDS?

Recommended Posts

D15-D11 wrapped - #2 booboo found in the map.

 

To just get started already, I went ahead and cemented Rev.7 of the wrap map to the perf board. Decided documenting the booboos on the wrap side as I went would be easy enough. I'll cement a change order list to the left of the LC Slot and a final, corrected map over the error ridden map on the connector side at some point. What a PITA to do checking and wrapping at once, but errors are easily fixed with the unwrap tool on the other end of the wrap tool. Not as easy as moving wires around on a breadboard, but far more compact, much better connection integrity and structurally robust. I need to be able to send this assembly off to a NIC wizard if I get the connections debugged and still can't get it online. I'll be happy if the SE/30 will boot with this monstrosity in place. I'll be ecstatic if the Declaration ROM on the NIC shows up in the NuBus/PDS report of TattleTech.

 

Before storing away tools I already had strewn about the place, I fabricated the clear plexi cover plate armor for the rat's nest of wires/chassis to hold the three cards together in handling. NIC and RA adapter are bolted up to it, ready to rock. I'd said I'd celebrate tomorrow if I finished enough downsizing by tonight with a wire wrap party, but that probably won't happen. So I'm rewarding myself with a bit of wrapping as a carrot when the timer goes off and I can stop beating myself up with the stick for a bit. Just spent way too much time getting everything squared away from the plexi fab and getting the hardware together, but I'm in a good place all in all.

 

 

That said: if someone could please proof/error check this version of the wrap mappings it would be a huge help. It's not as up to date as Rev.7, but it's the printout I'm using for markup as I wrap.

 

RosettaCache-01-002.jpg

 

Change Order:

LC B11 Maps to IIsi C19 - DATA24 was mapped to B19 GROUND on the IIsi

LC C14 Maps to IIsi B23 - DATA14 was mapped to B23 the same pin as DATA13

Share this post


Link to post
Share on other sites

Nope, not at all. It's an early version that should be easier to debug as the columns aren't flip flopped to match pins of the connector orientations of the board in the Rev.7 wrap map. It's a massive PITA to work from it, but the visual gymnastics of doing so seem to be helping me to find the real booboos as I make translation process booboos in the wrap.

 

A bunch more wrapped and I'm up to six booboos identified. I even made a booboo while highlighting one of the booboos on the card's map already! ::)

 

P1010039.thumb.JPG.097037617911e5515dab62a6847a9987.JPG

 

Orange data bus complete, just six more connections to wrap and the yellow address bus will be complete. At that point it's down to the last (first) eight rows of control, IRQ and clock lines to go. But my mid-stacation wire wrap party is over and I'm back to downsizing so it's all going back into the project tray on the shelf. Maybe I can pick at it now and then, but even when I'm done with wrapping I have to flip it over and buzz every single line for proper connection, shorts power/ground disasters in the making and what not. ETA undetermined, the only thing I'd said was that I hoped to start a wrap party today in the start of the process sense. No ETA on a wrap party in the sense of a an entertainment production end celebration.

 

I'm very shot on hacking stuff here in the virtual world ATM and demands IRL are piling up, but we'll get there! :blink:

Edited by Trash80toHP_Mini

Share this post


Link to post
Share on other sites

I'm keen to start messing with this idea, but I have a bunch of SE's, and no SE/30. On top of that, the SE seems like a good place to start as they seem to be more common than the SE/30. In other words, blowing up an SE would be less heartbreaking than an SE/30. Is there a card out there I should try to find that I'd have better luck trying on an SE?

Share this post


Link to post
Share on other sites

No, but designing an SE<->LC PDS conversion PCB would be much more economical than this SE/30 connector strapped proposition. SE and LC share the 96-pin "NuBus" connector form factor which is less than 10cm long an fits on inexpensive two sided PCBs, so that's the way I'd give that a try.

 

- 8MHz SE clock might be problematic in terms of timings for Ethernet from an unmodified LC NIC? Dunno.

- LC PDS is 68020 bus, so it should be 68000 bus PDS compatible,IIRC? Dunno maybe at least done a bit more so than the LC 020/LCIII 030 to full blown SE/30 bus 030 PDS? Again, dunno.

 

Back to my fascination with color coded wiring. The Worm Ourbouros patterns are necessitated by orientatingf the two cards so they install as compactly as possible on the conversion board. That I find the color-coded patterns aesthetically pleasing is a side benefit. Can't wait to see the two blue contro line whorls emerge with red, green, purple and black highlights intertwined. [:)]

 

Looks like loopy breadboard wires because I've kept "trace length" to approximately equal the longest connection necessary, which of course would be the address jumper block detour. ::)

Edited by Trash80toHP_Mini

Share this post


Link to post
Share on other sites

Looks AMAZING!  Don't worry guys, if this works out well, I don't mind funding the SE version of the adapter.  It should be slightly cheaper anyway.  I'm really hoping this project can work, it would really be amazing for everyone to have Ethernet cards that aren't $150+

Edited by maceffects

Share this post


Link to post
Share on other sites

HRMMM???? Wondering where/how the card might be shoehorned into an SE? Later models with the SE/30 compatible chassis would be no problem.

 

One thing to think about is that a single adapter in the form factor of the MicroMac Performer could be SE/Plus/Classic compatible. I wonder about the cards? Clear plastic Killy Klips beckon! Know anybody with experience applicable to developing the tooling for a run of little clear plastic doodads? Stamping out the metal contacts should be a piece of cake by comparison. How many of these suckers did you say you had? [:P]

Share this post


Link to post
Share on other sites

I thought the SE and SE/30 chassis was the same as well. Seems like there’s plenty of room in there. Card position doesn’t really matter... crimp and RJ-45 onto a short CAT5E cable, route to a 3D printed bracket that holds a keystone punch down jack on the back of the SE or SE/30, and you’re good to go. Seems to me like the big hurdle is just seeing if these things are electrically compatible or not.

 

If there’s anything I can to do help... I’m game. Including 3D printing plastic doodads.

Share this post


Link to post
Share on other sites

SE PDS expansion card position is horizontal, stretched across the logic board. The PDS connector has solid metal overhead. So NIC and adapter can't be located in the upper case. The assembly has to fit below decks like cards designed for the 128k-Plus that were Killy Klipped to the 68000. Not a lotta cubic available under there for matching up cards on EuroDIN connectors on two separate planes needing that much vertical clearance. Of the pics I've posted this one probably shows it best:

 

ClearBucketAdapter.jpg

 

Any expansion assembly must fit between logic board components and the slanted black line defining the hard deck of the SE Chassis. Whereas the SE/30 and late production SE have that convenient access hatch punched out for the upright  PDS card spec. of the SE/30.

 

If it can't be fit below decks, maybe some kind of ribbon cable slot extension might work to get it above deck? Dunno, tired and off to bed.

Share this post


Link to post
Share on other sites
18 hours ago, PotatoFi said:

I'm keen to start messing with this idea, but I have a bunch of SE's, and no SE/30. On top of that, the SE seems like a good place to start as they seem to be more common than the SE/30. In other words, blowing up an SE would be less heartbreaking than an SE/30. Is there a card out there I should try to find that I'd have better luck trying on an SE?

I think I misinterpreted your question. A driver for the SE would be the first thing you need to nail down. Did Farallon make an SE version of this card? If not, you'll be looking for an SE NIC developed in conjunction with a NuBus version for the 68020 Macintoch II sharing the same chipset making drivers available for both versions.

 

Availability of NuBus and LC PDS versions of this card/driver combo would be the heart of the matter. Don't recall offhand, but lack of a suitable driver was likely the stumbling point the many times suggestions for the LC NIC to SE/30 conversion came up. The SE/30 is a Macintosh IIcx stuffed into the SE form factor. That made its PDS an outgrowth of the NuBus/Slot Manager architecture. The SE, its PDS and drivers of its cards are not compatible with that architecture and require a very different type of driver.

 

TLDR: SE PDS is is the natural progression of 68000 architecture expansion. Apple II model slot based expansion had been in the minds of the Macintosh development engineers from day one. SJ and his project managers quashed the notion of expansion for the Macintosh at every turn: folklore.org: Diagnostic Port The workaround for that intentional limitation was likely coming to a boil long before Burrell & co. left Apple to form a startup they named Radius to develop the FPD.  I wonder if they chose that word for its connotations of lengthening/expanding a circle of Macintosh capability far outside the box in every direction?

Share this post


Link to post
Share on other sites
21 minutes ago, Trash80toHP_Mini said:

Did Farallon make an SE version of this card?

Great question... I have no idea. The only PDS ethernet card that I know of for the SE is the Asante MacCon (sometimes I see it called the Asante MacCon+). They are typically advertised for $150 on eBay.

 

Thanks for the background information... I'll have to look and see if a Farallon card exits. But for the SE... I'm starting to think that a custom SCSI ethernet adapter is the way to go. But engineering that is way beyond my skill level (I can might LED's blink in fancy ways on an Arduino, that's about it).

Share this post


Link to post
Share on other sites
11 minutes ago, PotatoFi said:

The only PDS ethernet card that I know of for the SE is the Asante MacCon (sometimes I see it called the Asante MacCon+). They are typically advertised for $150 on eBay.

There's your missing link, Asante MacCon IIsi SE/30 is the gold standard, also available in NuBus versions.

 

edit: dang, now I've got 3D models of the compacts banging around in my head playing bi-directional Jenga again. :blink: Good news it that if an SE in your batck has a late production date, you may find the SE/30 compatible chassis inside and your test bed setup is ready for sourcing connectors and doing a board design in the 10cm x 10cm. format. If the requirement extends to a third layer I've got a quantity 250 box of ten position jumpers. Time for a new topic.

Edited by Trash80toHP_Mini

Share this post


Link to post
Share on other sites

I opened some FDHD and SuperDrive models they both have the same chassis as the SE/30.  But I bet there is a unquie way to make it work that we haven’t thought of yet. I’ll spend some time on that if the wire wrap turns out successful.  

 

Share this post


Link to post
Share on other sites

I have looked into the issue more, and I think there is a reasonable solution we can do.  As a result, we need not be dissuaded from pursuing that goal as well.  Let's just wait and see what the final wire wrap reveals for us.  I bet when complete TattleTech will see it, but the drivers won't work. 

Share this post


Link to post
Share on other sites

That would be my prediction as well, but still holding some hope out for the NuBus driver working. I've yet to mention the fourth header in the address line block on the board. That's for the /NuBus line of the 68030/IIsi PDS. SE/30 is BusMaster hostile where the IIsi is not. LC NIC/all LC slot cards sound to be NonMaster implementations?

 

RosettaCache-01-002.jpg

 

Explanation of the LC NIC's /SLOTIRQ address line above (hardwired $E implementation) as a NonMaster NuBus IRQ address interrupt request presents a possible a detour around SE/30 PDS driver addressing issues when used in conjunction with the address location of the IIsi NuBus Adapter Slot. It was an interesting blurb in the LC II DevNote, may be applicable, may not, but implementation on the board is trivial being but one unused pin on the 030 PDS.

 

Wire wrap may not be the most current form of prototyping, but rework for connection corrections/modifications to wire wrap circuits on a prototype board is trivial.

Share this post


Link to post
Share on other sites

The original SE has a different chassis without the cutout and mounting tabs to allow vertical expansion cards.

Later SEs and the SE/30 share the same chassis with the cutout and the tabs to properly mount vertical cards.

 

I have modified a MacCon SE to mount vertically because I have an accelerator mounted on top of the 68000 which blocks the use of a standard horizontal SE upgrade card.

Share this post


Link to post
Share on other sites
8 minutes ago, Bolle said:

The original SE has a different chassis without the cutout and mounting tabs to allow vertical expansion cards.

Thanks for that crystal clear explanation. That hack was my reason for suggesting a vertical slot adapter/card LC NIC installation should such become a feasible proposition. What are your thoughts on /NuBus and chances for its use with the NuBus drivers in a workaround should the need arise?

 

 

edit: also wondering about the 8MHz clock issue in the SE. C16M is present on its PDS, but will the NIC clock asynchronous?

 

Edited by Trash80toHP_Mini

Share this post


Link to post
Share on other sites

Back to the SE/30 PDS adaptation again for an update!

 

P1010038.thumb.JPG.7190dad031be3dc54728f4fa8aa891b0.JPG

 

Something like twenty Blue Control lines and the three Purple IRQ lines left to be hooked up on the 030 PDS side. Everything is wired up on the LC side so I can sort out the wires from those connections and knock them down by the numbers, row by row. That's the easy part, flipping it over for testing and debugging is next up.

Edited by Trash80toHP_Mini

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×