Multibus ST-506/QIC-02 (was: Re: SGI IRIS 1400 ...)

From: Peter C. Wallace <pcw_at_mesanet.com>
Date: Mon Jul 22 11:38:01 2002

On Sun, 21 Jul 2002, Tothwolf wrote:

> On Sun, 21 Jul 2002, Tony Duell wrote:
>
> > > I also just noticed someone cut off the spindle ground strap. I wonder if
> > > it built up a static charge and damaged part of the head amplifier
> > > electronics?
> >
> > Possibly. The preamp is on the flexiprint inside the HDA (:-(). If you
> > can get a 'scope you can look for the servo sigals on the 2 wire jumper
> > between the boards. If you get nothing there, then you can syspect HDA
> > problems, I think. Without a 'scope it's difficult to make any
> > meaningful tests :-(
>
> Would the preamp even be replaceable if it turns out to be faulty? Any
> suggestions for a (temporary) home-brew scope probe?

        For high frequencies (up to 500 MHz for a simple probe), a homemade
resistive divider probe will easily outperform a standard scope probe
(capacitive divider) in pulse fidelity and input impedance. For a 100 to
1 probe, use a non inductive 4.99k 1% resistor connected to the centerlead of
a 50 Ohm coax. For a 10 to 1 probe, use a 453 Ohm input resistor. Terminate
the scope end of the coax with a 50 Ohm terminator (Ethernet (10B2) cables and
terminators will do up to a couple hundred MHz).
        Now you may say that 5K (or 500 Ohm in the 10/1 case) is too low an
input impedance for a good probe: well at a measily 10 MHz your fancy 10M Ohm
/ 20 pF probe has a (reactive) input impedance of around 800 Ohms. So at
anything greater than about 16 MHz, even a 10/1 resistive divider probe with a
500 Ohm input impedance is superior, and typically has much better pulse
fidelity than a standard 10/1 capacitive divider scope probe...


>
> -Toth
>
>

Peter Wallace
Received on Mon Jul 22 2002 - 11:38:01 BST