Recently, Somebody Somewhere wrote these words
> Hi Declan,
>
> > What you need to know is that the AD767s latch an input with a short
> > negative
> pulse (as short as 40nS) on the /CS line. I'm not getting this action.
>
> Maybe you are but you don't see it?
Exactly my concern. The AD767s don't see it either, but we are down to
component level here on a fast ttl isa card
>
> > The pulse appears to come from a Pal. I sent the Pals to the
> > manufacturer who tried them in another unit and they worked all day.
> > I feel I am looking for a voltage level or timing problem.
>
> Longer wires? Different pcb? Different voltages? PAL's are often very
> sensitive for the timing of signals on the inputs, where race-
> conditions can occure, leading to wrong states that will not give your
> pulse. All depends on the logic of the programming.
Same conditions, same board, approximately the same spot in the ISA bus.
>
> I would check the circuits and timing that connect to the PAL's
> inputs.
I'm straight onto the ISA bus. These pals have the device adress (0 x
0300) coded into them.
>
> > I have a 20Mhz scope here and can't see the pulse, or trigger on it.
> > My thoughts turned to building a latch trap - some latch circuit
> > that would come up low and latch high if the pulse appeared. I
> > could then tack this onto the chip, and look for the pulse.
>
> A 20MHz scope will not catch this. I have a 400 MHz tektronix that
> could do this. But I would still use my logic analyser (175 MHz,
> Philips). Can't you borrow a logic analyser somewhere? They really
> are very handy for catching one-time pulses, they are made for that.
> And you can see all pins at the same time.
>
/smiles wistfully
No, I a can't borrow. I can hire bhut they will only have a 2ghz 182
channel one available, and that will cost half a house. Frankly, if I
can catch it with a circuit, fine. I'll build that.
> > Such a latch could be made from a 74F<something>. Any thoughts on
> > how to check the level (e.g. does it go below 0.8V) or dwell (does
> > it last 40+nS) with this or something similar?
>
> 40 ns is very short, you could use a set-reset flipflop/latch
> 74<whatever>, preferrably something fast (ACT?), to see if the pulse
> is seen. You need short wires and good decoupling!
Can do.
>
> For the scope: A 40 ns pulse has a base frequency of 1/40e-9 = 25 MHz.
> To "see" a reasonable pulse on a scope you need as rule 10* this, so
> 250 MHz bandwidth. Even double this if you assume the pulse to be the
> half of a sinewave.
>
--
With best Regards,
Declan Moriarty.
--
Author: Declan Moriarty
INET: [EMAIL PROTECTED]
Fat City Hosting, San Diego, California -- http://www.fatcity.com
---------------------------------------------------------------------
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB CHIPDIR-L
(or the name of mailing list you want to be removed from). You may
also send the HELP command for other information (like subscribing).
