Daystar Universal PowerCache P33 in SE/30

[JDW]

I wish I could get my Turbo 040 to run again.  xx(

Here’s the thread which discusses the problem in exhaustive detail:





I would be willing to mail my 040 card off (temporarily) to any of you veteran forum members who have the parts and/or ability to carefully and constructively tinker, with aim of restoring it to functionality again.   And if you need it, I could also include my ARTMIX TS Adapter as well. My 040 card has the newest ROM, but I also have a second chip with an older version of the ROM on it, although that has ROM nothing to do with the problem this card has. 

 

[jessenator]

I know that the one in the picture is the older "IIci only" non universal version, but thanks for the heads up.

It still seems to work fine in the SE/30 though as far as I can tell.

Sorry for the remote necro post.

Bolle does this comment mean that your adapter board will allow pre-P33 PowerCache to be used in the SE/30?

I've been digging and this was what I found here. Apologies if this covered elsewhere.

 

[Bolle]

It did run just fine at least for me with both my adapter and the original Daystar SE/30 adapter.

I know that some people were having problems with the old PowerCache in the SE/30 but I couldn’t reproduce those in my tests.

 

[james_w]

It does. It also applies to the adapters I built so far. You will have to jumper the /CENABLE pin (C13 on the cache slot connector) to ground, otherwise the cache won’t get activated.

I'm going to carry out this adaptation to the ProtoCache 1.1 adapter I got from @Bolle. I'll report back, but assume it should all go fine (as long as I have better luck than with my recent Epic Blunder with Turbo 040 card)

 

[james_w]

So I just added the jumper wire onto my accelerator adapter - it's the ProtoCache 1.1 by @Bolle



Took just a few minutes and I'm very pleased with the HUGE performance difference it has made now that the cache is being used properly.



Thank you everyone for figuring this out and sharing how to fix this!

 

[ktkm]

Super! Nothing better to match your profoundly gorgeous clear case!

 

[james_w]

Super! Nothing better to match your profoundly gorgeous clear case!

Thanks to @maceffects 

 

[hfrazier]

The T040 even works with the MacCon stuck under it. I could never get this to work reliably up to now. If I knew it would have been as simple as inverting that one clock signal.

It is running rock solid for an hour now. Downside is that the P33 won't work with the clock inverted but I might be able to solve that and go completely jumperless on the adapter.

Another mystery solved.

@Bolle I have two GAL T040's that I just can't seem to get to stay stable, and aim to do the MacCon -> Short Adapter + T040 -> iisi Pivot combo.. I have a few 74*68 chips laying around, and your ProtoCache1.1 short adapter I can modify. What are the jumper settings I need and besides the 5V and GND, how should I connect the logic chip to invert the clock? A38 through a gate and then out to pin 18 where the GAL was?

This is an awesome discovery!

 

[JDW]

Hi all, I tried adding the jumper to my Twinspark to see what effect it had on my Daystar Turbo 040 - I was blown away by the improvement!

The MacBench scores speak for themselves. Below I tested my SE/30 with a Daystar Turbo - one with the Jumper and one without:

View attachment 28513

Did you also run before/after Speedometer 3.06 and/or Speedometer 4.x scores too?  I get the same before/after-jumper scores in both of those.

NOTE: I have the older model GREEN colored TS Adapter, but that shouldn't matter.  I added the jumper across the correct 2 pins.  I tried two different SCSI hard drives, but the results are the same.

When I run MacBench 3.0 under System 7.1, I get a 4.79 for the Processor score and 1.0 for Floating Point and 20.52 for Disk Mix, which are quite similar to your scores "w/jumper."  I didn't have older MacBench 3.0 scores using with the Turbo040 saved though, and I didn't want to desolder the jumper to run the tests again.

I also ran Norton System Info under System 7.1 and my scores today were pretty much on par with past scores.

I always load QuadControl, so I don't know if that has anything to do with my scores remaining consistent. My 40MHz Turbo040 has the newest 4.11 ROM. (This is the same card @Bolle kindly repaired for me by swapping out the bad Daystar chip.)

Now here's the interesting part...

I booted into 7.6.1 so I could run MacBench 4.0 (I clicked "Continue" when it complained about the lack of a 640x480 display), and I found differences with past results:

Processor w/jumper: 36

Processor without: 28

Floating Point w/jumper: 14

Floating Point without: 12

Disk w/jumper: 113

Disk without: 93

Those differences seem large enough that it doesn't seem like a normal variation that would appear when you run benchmarks multiple times.