1. Home
  2. Embedded Programming
  3. Use fast driver with old slow bus?

Use fast driver with old slow bus?

I want to find proper driver chips for an old bus (Q22Bus). The protocol says: rise/fall times must be no faster than 10ns. The chip I am evaluating is 74lvt125. It meets all other requirements but the rise/fall time is much faster than 10ns. It is not stated in the datasheet, but it would be around

1 ns if we look at the propegating delay time (1ns~4ns). I know there is potentially signal reflection problem. But how bit the risk is? The Q22 bus is a terminated bus, with resistors of 178 to 5v and 383 to gound for one type and 330 to 5v and 680 to gound for another type. The length of the backplane is 10 inches and maximum total length of 16 ft cables can be used to connect several backplanes. I wonder if I use this chip to drive the bus, is the total length of cables the only facter I need to sacrifice to avoid problems? Are there other problems that I am not aware of? Thanks. 
--
http://www.yahoo.com/mscpscsi
Reply to
vax3900
Loading thread data ...

I'm intrigued regarding the who, what, why of a Q-bus build (?) project in the 21st century.

Firstly, a comment on the 16ft of cables, I don't believe that bit. Once upon a time I ran pairs of BA23s buttoned together with DEC cables. Many system crashes later I took to stuffing it all into one BA123, and the 'mileage' improved. Not proof, but a caution. Are you certain 16ft was permitted, it's well outside my experience; i.e. I'm curious.

As my docs are in a boneyard, I can't comment on what the old (qualified) chips were for several days at best. However, I would be inclined if at all possible to continue to use them. Alternatively, even with the terminations, you will see funny waveforms on the Q-bus; 1 ns rise/fall needs careful termination once the signal path exceeds ~7"; cf 70" for 10ns Tr. In other words I doubt you can physically shorten the bus sufficiently. And, the careful termination would start with source resistors dor signals with signiicant edges ... Other issues which might keep one awake with 1ns Tr/f include: edge sensitive (which may include level signalled interupts etc) signals exhibiting multiple transitions at the receiver and ...; signal mangling by the backplane / connector assembly's impedance discontinuities; anything else in Johnson's book of Black Magic (ISBN 0-13-395724-1).

I'd be interested to hear what works.

Regards

DrB

Cross posted to vmsnet.pdp-11 to tap their expertise.

evaluating

much

around

bus

used

Reply to
DrB

I plan to work on a SCSI controller project because SCSI QBUS controllers still carry high prices.

It is stated in the Qbus protocal (page A34, item 8, and ).

They are hard to find and still carry high prices. That's why I am looking for alternatives.

It is really bad news. I borrowed that book months ago but didn't get the time to read. Now it seems a good time.

Thanks. I will post once I find the proper chip.

vax3900

--
http://www.yahoo.com/mscpscsi
Reply to
vax3900

I'd like to learn a little more regarding your project, but

formatting link
appears broken.

Has any work been done on interfacing IDE disks to RQDX3's by bridging the ST-506 bus to IDE ?

DrB

in

Once

Many

(qualified)

and

protocol

is

one

to

Thanks.

Reply to
DrB

sorry it is

formatting link

not that I heard of.

vax, 3900

--
http://www.geocities.com/mscpscsi
Reply to
vax3900

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.