Jump to content
Trash80toHP_Mini

ProtoCache1.1 - IIsi/SE/30 PowerCache Adapter Prototype

Recommended Posts

What about the Applied Engineering board?

On ‎1‎/‎1‎/‎2018 at 4:52 AM, Bolle said:

Probably not going to work with the Carrera.

The adapters for the Carrera (and Sonnet I think as they just rebranded the Carrera?) are slightly different.

 

Also the Carrera was never supposed to be used in a SE/30 and will refuse to boot and say that the machine is not supported.

It would need modifications to either the ROM on the Carrera or the driver to work.

 

 

Anyone try these with other brands of 040 accel?

Share this post


Link to post
Share on other sites

Not sure about that one. Never actually seen it.

Rule of thumb is, if it works in the Daystar adapters, it will work in here.

 

If it is compatible with the SE/30 from a firmware/software point of view is another question. I remember reports from somewhere in the forums that the Mobius as well as the Carrera are supposed to not work in the SE/30 due to firmware and drivers are not willing to run.

 

The pinout differences can be addressed and I might be doing so on an adapter designed for the IIsi, where most of the other 040 accelerators do work. Only a few signals need to be swapped around. Basically the difference between the way Daystar and all the others did 040 accelerators was in the way they shut down the onboard 030 either via /CPUDIS on the cache slot or /BG, /BR and /BGACK signals on PDS adapter implementations.

Daystar is converting the /CPUDIS signal supplied by the accelerator to the matching /BG, /BR, /BGACK signals on the PDS slot to shut the 030 down.

The others are somehow doing this right on the accelerator. Their adapters leave /CPUDIS open and connect the other signals appropriately. 

 

 

Edited by Bolle

Share this post


Link to post
Share on other sites

:ph34r:  Could you test a really nasty WAG for me? Will the SE/30 run without the socketed 68030 on board with PowerCache in our adapters? The Socketed PowerCache does away with it after all.

 

edit: dunno if Bolle's SE/30s are socketed. Can someone else test this CPU removed configuration with a standard DayStar setup for me? If it works, that's one heckuva spot to put a current design PDS Card.

 

Edited by Trash80toHP_Mini
the usual

Share this post


Link to post
Share on other sites

I desoldered the 030 on one of my boards to test that already. It does run without it.

Same for the IIci. I had a board with a broken CPU that I replaced and tested PowerCache and Turbo040 when the broken 030 was removed.

Share this post


Link to post
Share on other sites

Niiiice! Suddenly we have a handy prototyping slot! [:D] Now that you and joe have torn the PowerCache adaptation windmill to its very foundations I need a new impossible dream.

 

Lets call it the PDS2SD for want of a better term for the time being. The main reason for the NuBus in SE/30 project was to get my JackHammer running in there. [}:)] The SE/30 SCSI controller is the last major bottleneck. Time for a new thread!

Share this post


Link to post
Share on other sites

 

Quote

*bonus*

I know what you have been waiting for:

 

IMG_3517.thumb.JPG.496192a8f9cd881256514c3e60f59574.JPG

 

It works and fits into the case :evil:

 

Wow, fantastic work. I believe the answer is no, but this config with a Daystar 040 accelerator will not work, correct? Gamba said no but he didn't have your card...

 

Also, does this config clear the Floppy chassis and HD? 

 

Edited by Von
Forgot a Q...

Share this post


Link to post
Share on other sites

It works AND fits in the case last I heard. Bolle came up the great idea of building only the top half of the DiiMo form factor adapter board. It plugs into the "wrong angle hacked" connector on the NIC at the left on the bottom of the stack and his GS card plugs into the adapter's passthru connector. VERY slick!

 

His second version is the baseline DiiMo form factor for those without a NIC.

 

edit: it clears the full height OEM HDD Cage.

 

Edited by Trash80toHP_Mini
the usual

Share this post


Link to post
Share on other sites

This is all so amazing. I would love to order one of these cards, assembled or even bare, I have several of each of the euro connectors on hand!

 

... I’ve got a IIsi pivot and MacCon... Anyone have a Turbo040 they want to sell? 

Share this post


Link to post
Share on other sites

With a screen shot showing the Asante NiC, DayStar 040 and Micron Xceed card all running together I will buy one as I have all 3.  Is the mod the Asante where its secondary PDS points up required or can a right angle adapter be used?  Also, is the a FW minimum on the 040? Mine says ROM v4.01...

Edited by Von

Share this post


Link to post
Share on other sites

The firmware on the Micron is not compatible with 040 CPUs. This still applies today as it did in the good old Gamba days ;)

 

I still have fully assembled boards of both types for sale. Just PM me which one you want.

 

Another shot for size comparisons:

 

kXMJx5v.jpg&key=9480ba8bbaced8732696b8cb

 

Also take a notice of the compatibilty issues I encountered while testing the cards I have access to with 040 Upgrades in the SE/30. Check my post on page one in this thread right here:

https://68kmla.org/forums/index.php?/topic/53710-protocache11-iisise30-powercache-adapter-prototype/&do=findComment&comment=571472

 

Note that the Turbo040 is especially picky in a two PDS card setup as it seems to be running very aggressive timing on the host bus to achieve part of its speed bost. This does not go too well with more than one PDS card in most of my tests. Probably also because this was never supposed to be used this way.

 

The 030 upgrades are all safe to be used with multiple cards as far as I can tell.

Edited by Bolle

Share this post


Link to post
Share on other sites
20 minutes ago, Von said:

Is the mod the Asante where its secondary PDS points up required or can a right angle adapter be used?

 

It is required. With just plugging in a right angle adapter the stack is most likely not going to hold against the force of the accelerator pulling it down.

Share this post


Link to post
Share on other sites
6 hours ago, Bolle said:

 

It is required. With just plugging in a right angle adapter the stack is most likely not going to hold against the force of the accelerator pulling it down.

I figured this. Is a desoldering gun the recommended tool for the job?

..I just ordered one, I needed it anyways..

Share this post


Link to post
Share on other sites

I do this the old fashioned way using an iron and a solder sucker. Works well enough for me.

A gun should do just as well.

Share this post


Link to post
Share on other sites
23 hours ago, Bolle said:

The firmware on the Micron is not compatible with 040 CPUs. This still applies today as it did in the good old Gamba days ;)

That is what I thought and what I experienced when I tried SE/30 + Asante + Daystar IIsi adapter & 040 Accelerator + Micron Xceed. I tried it once and trusted what Gamba said,

 

SE/40etgs:
This setup would have EtherNet + DayStar Turbo 040 accelerator + Micron Xceed internal grayscale.
But, our testing has confirmed that the Xceed Color 30 video card is not compatible with the DayStar Turbo 040.
This was also verified without the EtherNet card, and using DayStar SE/30 instead of DayStar IIsi PDS adapter.

 

Then I saw this setup with the motherboard running outside the case (I'd love to know how you do that):

 

On 12/9/2017 at 10:24 AM, Bolle said:

As said I decided to make a tall version as well because the IIsi Pivot can not be made to work together nicely with the Turbo040. There seems to be a timing issue that does not occur when the PowerCache is used instead of the turbo.

Problem also does not occur when the two PDS cards are connected directly to each other.

So:

 

MacCon->IIsi Pivot->adapter+Turbo040 = works

IIsi Pivot->MacCon->adapter+Turbo040 = works

adapter+Turbo040->IIsi Pivot->MacCon = works

adapter+Turbo040->MacCon->IIsi Pivot = works

MacCon->adapter+Turbo040->IIsi Pivot = does not work

 

The GAL is buffering the clock signal and introduces a really small delay which seems to be enough to throw things off for the Turbo040 as it seems to be operating on a tighter timing than the PowerCache.

And even though this...

IMG_3524.thumb.JPG.83905ef3956a98d13a1104bbba59550a.JPG

...might look funny it is not actually useful :p

 

The IIsi Pivot and MacCon will still fit nicely on top of that, so no problem if you do not need to run big cards like the XCeed....

 

Did you happen to try:

 

adapter*+Turbo040->MacCon->Xceed (*where adapter is what you created or a IIsi with Accelerator/Cache slot and PDS passthrough)?

This has the the adapter and 040 fist in line on the PDS slot...

 

Gamba said he tried it however I didn't and am away from my machines at the moment.  I do understand the challenge of fitting this rig in the case however I am working on something that could address that.

 

THx

 

 

 

Edited by Von
Gamba link

Share this post


Link to post
Share on other sites

The Micron and 040 do not work together in any possible combination.

It hangs on a gray screen when initializing the Micron no matter if the internal or an external screen is used.

Share this post


Link to post
Share on other sites

Oopsie! I misread/failed to read your question, Von. Booboo's badly in my response.

 

silly notions:

 

Are we positive it's a firmware issue and not merely a timing issue?

 

Has a crystal tweak for the Micron been tested? Could decoupling/massaging its timing inputs and then tweaking its outputs back into phase with the 040 card's bus timing shenanigans allow it function correctly in its own little timing bubble?

 

Tweaking the outputs might not even be necessary? It interfaces directly with the A/B as with its external connector, no?

 

Thinking here is along the lines of a NuBus card operating on a buffered asynchronous 10MHz bus. A synchronous PDS bus card is the equivalent of a bus buffered NuBus card as far as the Slot Manager is concerned. Could we take another page from the NuBus playbook and introduce a near synchronous buffered bus into the PDS world?

 

Sorry if this is techno-noobish babble, gotta ask em' as they come up as a learning experience if nothing else. :mellow:

 

Edited by Trash80toHP_Mini
spellchecking

Share this post


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

The Micron and 040 do not work together in any possible combination.

It hangs on a gray screen when initializing the Micron no matter if the internal or an external screen is used.

I figured that this had been checked, thanks much for confirming!

 

I have a couple of threads on a possible path to a SE/40etgs where the Daystar 040 is replaced by a Carrera 040:

Need help with MicroMac Carrera040 in a SE/30
SE/40 Grayscale: Need help with a Carrera hack

 

User GeekDot engaged on this quest and delivered 2 revs to the Carrera CP, neither of which worked. Our thinking was that the init needed to be patched too.  Problem is that in one of these posts there was much discussion (spring/summer of 2017) and possible direction appears to have been lost in the site upgrade. Both of us have asks into the Admins to see if the missing posts can be recovered...

Share this post


Link to post
Share on other sites
5 hours ago, Bolle said:

The accelerators CPU clock is not present on the PDS connector. Everything on the PDS is synced to the 16MHz clock.

We're not talking about the same thing, I should have said "CPUCLK"

 

If the Micron card is running off the SE/30's native clock signal on PDS pin A38 CPUCLK signal and the 040 card's aggressive timing on that line is what's upsetting the Micron, I've got one of my knuckledraggin' Neanderthaler tests in mind:

 

On your adapter, cut the Passthru's trace to A38 (CPUCLK) and jumper it to the C38 (C16M) 16MHZ reference signal that's used to get SE/30 cards to work on the IIsi's 20MHz PDS.

 

It's a long shot, but might be worth a try? WTH! :ph34r:

Share this post


Link to post
Share on other sites

The short of it:

 

Got linkage to the PowerCache drivers for IIci/IIsi/SE/30 handy?

 

The long of it:

 

After wigging out on the new USPS tracking disservice, my bargain basement, untested IIci board for finally taking my long untested 50MHz PowerCache hit the bench today!

 

Seems to work despite a Maxell fizzle that ate away the entire positive contact in the battery holder! [:D] Can't get a startup chime on it without PRAM bat with the speaker from the IIcx, but I'm not getting one from that machine either. Gotta test with another speaker a/o bat.

 

Forgot about the lack of cutout for the VampireVid connector in the IIcx case. That'll be desoldered if it's the only physical incompatibility. I've got at leas three times as many VidCards as I have Macs to put them in and IIci video shouldn't happen to anybody but a writer with a Portrait Display, a DTP jockey who needs spot color on a miniscule CRT to go with a "real display" or a gamer. [;)]

 

Looks like there are no ROMs on the IIci board? Boots GREAT from BMOW's SIMM/ROM image! Lovin' that acquisition. [:)]

 

The one time I have a chance to test the IIci Cache Card mcd sent my way I can't even find the dang thing! LOL! I've got him to thank for the PowerCache he sent along with a Mac II adapter, great guy, much missed here. You can blame him for my years of babbling about this adapter cloning project. Idling curiosity about that funky Mac II adapter with discrete CPU and MMU connections was the genesis, actually starting the string of threads came later.

Share this post


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

 

On your adapter, cut the Passthru's trace to A38 (CPUCLK) and jumper it to the C38 (C16M) 16MHZ reference signal that's used to get SE/30 cards to work on the IIsi's 20MHz PDS.

CPUCLK and C16M are the same and connected to each other already on the SE/30 logicboard.

Share this post


Link to post
Share on other sites

iCrap! Cut it and add an "in phase" oscillator for a clean signal? My indecently uneducated WAG here is that the 040 is latching on the very edges of the 030 clock to the point (it jitters???) the Micron card finds it unpalatable?

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

×