[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [N8VEM-S100:6909] Re: Altair Cyclops Info Wanted



Leonard, your comment about shift registers just awakened some very long-term memory.

In the 70’s I was employed just out of high school and during my undergrad years as a bench tech fixing terminals, one particular brand being Wiltek.

These beasts were built to the size of an average office desk.  The electronics existed in a card cage inside – about 7 or 8 PCBs each measuring about 10 by 16 inches, all stuffed with SSI and MSI 74xx.

Some of the boards had the shift register memory you mentioned for temporary storage.

But the mass-storage facility was a rather unusual implementation of magnetic tape.

Two continuous loops, each about 50 feet long, of mag tape with cog holes, making it look very similar to 16mm film, were run back and forth through a transport.

The transport was a set of stepper motors that moved the tape under a read/write head, which handled 8 bits at a time.

One loop stored received data and one stored data to be transmitted.  This would happen in batches when the mainframe would poll the terminal.

The tape was stored in a narrow channel built into the back side of the desk.

The biggest problems with these machines were all centered around the tape storage:

                Driver transistors and damper diodes would go bad, occasionally catastrophically burning holes in the PCB.

                Tape heads would wear and need to be refurbished.

                The tapes themselves would go bad.

I cannot remember ever having to replace any of the shift register ICs, even though they were early NMOS devices.

I would NOT like to return to those days.

 

I do, however, remember the Cyclops article and thinking how cool that was (literally – all cameras of that era were tubes!)

Just a kid in school at the time, it was quite over my head and far in excess of my lawn mowing budget.

 

Bob Bell

 

 

 

From: n8vem...@googlegroups.com [mailto:n8vem...@googlegroups.com] On Behalf Of lcyoung
Sent: Saturday, May 02, 2015 3:47 AM
To: n8vem...@googlegroups.com
Subject: RE: [N8VEM-S100:6909] Re: Altair Cyclops Info Wanted

 

Some terminals I worked on  years ago used shift registers connected in a loop as memory...

 

Leonard

 

From: n8vem...@googlegroups.com [mailto:n8vem...@googlegroups.com] On Behalf Of John Parsons
Sent: Saturday, May 02, 2015 12:15 AM
To: n8vem...@googlegroups.com
Subject: [N8VEM-S100:6909] Re: Altair Cyclops Info Wanted

 

A little detective work that needs a sanity check:

Regarding Gary's idea of using a 4106 or similar: the device shown in the Popular Electronics article makes 30 reads of the memory chip to generate a frame. The output of each bit on each scan is amplified and used to generate a dot, with the dot intensity depending on the charge of the bit in the RAM on that scan. That will give a grey scale dot on the 'scope. The 4106 refreshes the memory on every read, so the bits will always be either a 1 or a zero, depending on the light. The article Gary mentions has no grey scale, only black or white. So no chip that refreshes on every read can be the chip in the Cyclops.

So that leaves another dilemma: if the chip can't refresh on a read, and has no refresh pin, how can a DRAM be the chip? Is there any other refresh mechanism that was in use in 1975? If not, it seems the memory must be static.

Where are all the semiconductor engineers? I guess they have been replaced by computers. We're doomed!!  :-)

JohnP

--
You received this message because you are subscribed to the Google Groups "N8VEM-S100" group.
To unsubscribe from this group and stop receiving emails from it, send an email to n8vem-s100+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "N8VEM-S100" group.
To unsubscribe from this group and stop receiving emails from it, send an email to n8vem-s100+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.