View Full Version : Dazzler and dynamic memory

July 24th, 2014, 02:26 AM
Has anyone gotten their Cromemco Dazzler to work with dynamic memory. The manual states static only.


July 24th, 2014, 04:12 AM
Hi All;

There's a reason why the Manual states Static only..
If I remember correctly it has to do with how the refresh rate is done on the early machines.. It hogs too much processor time..


August 30th, 2014, 11:10 AM
I remember the Cromemco Dazzler... kept trying to justify buying one but with my tuition costs I opted out. While the colors were cool the text mode was too coarse... I went with a Miniterm Associates MERLIN board instead.

IF you get your dazzler running, it would be cool to upload some videos of that for the vintage community. :)

August 30th, 2014, 09:06 PM
Will not work with DRAM, in fact will not run with a number of other S100 boards.

August 31st, 2014, 03:53 AM
Will not work with DRAM, in fact will not run with a number of other S100 boards.

This is why I'm thinking my Dazzler actually works now just not in my sol-20. I have to try it in an Altair. My Altair motherboards are full and I keep delaying adding my spare. I'd pull two cards to try it but the Rev-0 motherboards are too thin to risk keep pluging and unplugging. I may borrow another Altair with an aftermarket thicker 12+ slot motherboard in the future to try it which would then remove my procrastination on adding the extra board to my Altair.


August 31st, 2014, 06:49 AM
...I'm thinking my Dazzler actually works now just not in my sol-20. I have to try it in an Altair....
You can probably resolve why, (other than dynamic ram timing issues) by pulling the schematics of your S-100 boards and tabulating the S-100 signals and polarity for each, looking for conflicts. It really wasn't standard until IEEE-696.

I built a Cromemco Z2 back in 1978 and bought boards from various companies, which probably increased the odds of incompatibility. I had to reverse the polarity of one board's control signal and I had other special boards making incompatible use of the extra pins with poorly defined signals.

I just modified the boards so they'd all work together. I found an unused inverter gate on the board with the bad polarity signal and ran solder tacked wire-wrap wires to fix it (hated having to cut traces but it had to be done). The three signal conflict was tricky but with the 100 pins tabulated for each board I could see the available S-100 pins to make it work.

Generally speaking memory boards seldom cause problems unless they play with the PHANTOM signal. Usually the worst offenders are Display cards and Floppy Disk Controllers (in my case it was a Mini-Term Assoc. MERLIN display and a NorthStar MDS).

I haven't found my S-100 notes, just all my documentation and hardware, so I'll probably regenerate that tabulation some day. As this is the third time I've mentioned in in VCF, its probably time I dig for more illuminating details.

I found 3 Dazzler PDFs on the maben archives... I'll look through them later.

...My Altair motherboards are full...
You may be able to confirm the Dazzler works by thinning the odds of card conflict merely by removing any unnecessary S-100 boards from the cage, leaving only what you need to get the Dazzler tested. If you prove the Dazzler is fine then you can cycle power with the remaining set of boards added in a few at a time until you discover an incompatible pair... then you've got less work tabulating the S-100 signals between those two boards.

Just be careful not to lose track of whether power is ON or OFF when doing repetitions like this. Its a good way to learn a double-check or triple-check methodology for working in situations where a mistake might destroy a card. What you may lose tends to determine how much you need to ramp up safety checking like this.

September 1st, 2014, 03:28 AM
My Altair is a rev-0 with two 4 slot motherboards with CPU, four 4k MITS static ram cards, MITS 1702 EPROM card, SIO-A and MITS PIO parallel card. All my cards are rev-0 so I'm guessing they don't do anything too weird since the dazzler is the only "approved" card from a 3rd party that MITS would accept at the time. My concern is the actual motherboards. One of the reason people replaced the MITS ones was they are thin and flex. So I don't want to mess with mine since the Altair is running perfectly. That's why it makes more sense to borrow another one with a PT motherboard that is thicker and has lots of open slots. If it all works then I have a reason to add 4 more slots to my system with another motherboard.

I was trying to use one of my Sol-20 to test the dazzler so I would know if I should put the extra motherboard in the Alrair, but I'm now guessing between the dynamic ram cards, built in video and IO and all the other stuff the Dazzler won't work in a sol-20 which was I guess the original reason for this thread. the Sol-20 is usually my "card" testing machine with the built in riser and a built in crowbar circuit. It's actually one of the reason I have two. I keep one at my "work area" as a generic S100 machine, but apparently it's not generic enough. :-)


September 1st, 2014, 06:26 AM
As the Cromemco Dazzler is such an odd card anyhow, it would make some sense keeping it out of the Altair system so the Dazzler doesn't upsetting its operation not its original set of MITS cards.

The Sol-20 is probably a better home for it, even if you remove stuff there for testing.

What were you wanting the Dazzler to do in the system? Were you intending to write code on the Altair to experiment with it?

September 2nd, 2014, 02:21 AM
I agree the Sol-20 would be a good home for it, but I just think that they conflict. The Altair will be a good home with another 4 slot motherboard to give me more slots, I have the board, it's just I don't have the 100 wires precut, that is the reason I think I'm procrastinating. The main reason for the Dazzler for me is the few color games as an example of early hobby computing. So thinking about 1975/1976 as a setup, it belongs in the Altair...