Macintosh SE SuperMac SpeedCard accelerator

[jasa1063]

I just picked up a Macintosh SE SuperMac SpeedCard accelerator. This accelerator has a MC68000 clocked at 15.67MHz along with a MC68881 math coprocessor at the same clock speed. Finally it has 16K SRAM cache. I have been interested in getting one of these for quite some time to see just what kind of improvement over a stock SE I would get. I bought this from an eBay seller in Japan. Shipping was only $14 via DHL and was delivered in just 2 days! There were a couple of things that I like about this accelerator. The 16K SRAM Cache, a 68881 math coprocessor and the ability to boot in nomral 68000 mode for compatibility. The accelerator was in excellent condition and installed with no isue. I ran SpeedOMeter 3.23 on System 6.0.8, System 7.0.1 and System 7.1. I used the SuperCard Control Panel v1.7. Overall a 230x increase in CPU, 70% increase in graphics and 440% in SANE math performance is not too shabby. Here is a front and back picture of the accelerator.

 

[jasa1063]

Here are the results from SpeedOMeter 3.23 on System 6.0.8.

 

[jasa1063]

Here are the results from SpeedOMeter 3.23 on System 7.0.1.

 

[jasa1063]

Here are the results from SpeedOMeter 3.23 on System 7.1.

 

[lilliputian]

What is the P2 connector for? Interesting that the main PDS connector is just the pins!

 

[jasa1063]

What is the P2 connector for? Interesting that the main PDS connector is just the pins!

It's for plugging in an additional expansion card like a video card. I think you could get one specifically for this accelerator from SuperMac. The manual does not go into any detail on it.

 

[JDW]

Here's a few photos of my SpeedCard installed on my Mac SE motherboard...


(And yes, I did later cover the windowed chips with tape to prevent erasure over time by UV.)

And here's a photo of the SpeedCard owned by forum member @BFEXTU , who was an engineer who once worked for SuperMac:



@BFEXTU was the gentleman who kindly provided me with the v1.7 Control Panel I put on Macintosh Garden, which was never released back in the day.

 

[jasa1063]

Here's a few photos of my SpeedCard installed on my Mac SE motherboard...

View attachment 20945 View attachment 20947 View attachment 20948
(And yes, I did later cover the windowed chips with tape to prevent erasure over time by UV.)

And here's a photo of the SpeedCard owned by forum member @BFEXTU , who was an engineer who once worked for SuperMac:

View attachment 20946

@BFEXTU was the gentleman who kindly provided me with the v1.7 Control Panel I put on Macintosh Garden, which was never released back in the day.

Thanks for posting the v1.7 Control Panel. That is really what this vintage computing community is all about, helping one another and preserving computing history!

 

[jasa1063]

I looked through my parts inventory and found a 40MHz MC68882 math coprocessor. I swapped it out for the 16MHz MC68881 on the accelerator and ran through SpeedOMeter 3.23 under System 7.0.1 (The daily driver OS I run on this Macintosh SE). Here are the results, nothing earth shaking, but the MC68882 is a bit faster. I am also including a picture of the accelerator as installed on Macintosh SE motherboard. I took the crumbling paper off of the two EPROMs and covered them both with electrical tape.

 

[phipli]

I do have a PGA 68882 in couple of things. I may give it a try someday, but based on past experience, the increase would not be that great. I have tried both a 68881 and 68882 in the FPU card on Macintosh LC. The speed difference was 5-10% at most, depending on the benchmark. When I post some more detailed scores this weekend, i will test System 6.0.8, 7.01 and 7.1 for comparison.

I looked through my parts inventory and found a 40MHz MC68882 math coprocessor. I swapped it out for the 16MHz MC68881 on the accelerator and ran through SpeedOMeter 3.23 under System 7.0.1 (The daily driver OS I run on this Macintosh SE). Here are the results, nothing earth shaking, but the MC68882 is a bit faster. I am also including a picture of the accelerator as installed on Macintosh SE motherboard. I took the crumbling paper off of the two EPROMs and covered them both with electrical tape.

That's interesting that it was such a small difference. I don't have a 68881 to compare, but old publicity materials claim double the speed or there about if I remember correctly. 12% is... a lot less!

I guess it makes more sense.

Silly question, but since you have System 7.x installed, could you possibly run the System Info benchmarks? I'm much more familiar with them (and trust them more). They published details of what each sub-part of the benchmark does and there is an option in the menu to give a detailed breakdown. It would show if the faster FPU was universally only 12% different, or if it actually was strong in some areas and weak in others.

You can also export the results by copying them and pasting them into a spreadsheet which is handy if you want to do custom analysis, but probably less applicable here.

 

[jasa1063]

I tried running Norton Utilities System Info 3.5 on my Mac SE. Unfortunately, it get a Bad F-Line Instruction error when it tries to display the results.

 

[phipli]

I tried running Norton Utilities System Info 3.5 on my Mac SE. Unfortunately, it get a Bad F-Line Instruction error when it tries to display the results.

3.5 might be a bit new, 3.2 is probably more period. But don't worry (I realise it isn't generally trivial to throw software on old machines).

I'll leave you in peace to play Prince of Persia on your SE

 

[jasa1063]

3.5 might be a bit new, 3.2 is probably more period. But don't worry (I realise it isn't generally trivial to throw software on old machines).

I'll leave you in peace to play Prince of Persia on your SE

Thank you sir!

 

[jasa1063]

I tried running Norton System Info 3.5.3 on both System 7.0.1 and System 7.1 and skipping the Video tests. When trying display the results, I get an Address Bus Error. I then ran Snooper 2.0 the CPU test came out to 229% of an stock SE. The Video test came out at 166% of a stock SE. The FPU tests were skewed and I don't trust the results. Overall the results pretty much mirror what the SpeedOMeter 3.23 results are.

 

[JDW]

Remove all Results files from the folder and try displaying results again. It will likely work. You can then start slowly adding them back and see how many your machine will allow. Maybe 10 or 15?

The problem is the SE has issues when too many results files are in the results folder.

 

[jasa1063]

Remove all Results files from the folder and try displaying results again. It will likely work. You can then start slowly adding them back and see how many your machine will allow. Maybe 10 or 15?

The problem is the SE has issues when too many results files are in the results folder.

I moved the Benchmark Results folder and reran Norton System Info 3.5.3 and it works. I will move back more benchmark results to get a good sample.

Thanks!

 

[JDW]

Great!

Feel free to mention that trick in your public thread, as that info will surely help other people who no doubt will have the same problem. I think it's a bug in System Info pertaining to the lack of sufficient RAM to load all the results files, but that's pure speculation on my part.

 

[jasa1063]

I posted the info about Norton System Info 3.5.3 in my thread on 68KMLA to let others know there as well.

Thanks!

 

[phipli]

Great!

Feel free to mention that trick in your public thread, as that info will surely help other people who no doubt will have the same problem. I think it's a bug in System Info pertaining to the lack of sufficient RAM to load all the results files, but that's pure speculation on my part.

Interesting - I haven't seen that issue before.

Do you have less than 4MB of RAM? I should add a warning to my website as well.

 

[JDW]

Do you have less than 4MB of RAM?

Exactly 4MB on my SE.