[time-nuts] Using a frequency synthesizer replacement for motherboard oscillator

David davidwhess at gmail.com
Sun Dec 2 04:37:55 UTC 2012


On Sat, 01 Dec 2012 19:10:54 -0800, Hal Murray
<hmurray at megapathdsl.net> wrote:

>davidwhess at gmail.com said:
>> One of my favorite tricks back when the ISA bus was still available was to
>> use a custom expansion board I built and an oscilloscope to measure the
>> interrupt latency. 
>
>You can do the same trick without special hardware.  Use the printer port.  
>Of course, that assumes your PC/Laptop still has a printer port.

I almost always end up installing a PCI/PCIe serial and parallel
expansion card into my systems even now.

>I have a laptop with a printer port on the thing Dell calls a MediaBase.  
>It's an extender/fattener that includes a CD drive with connectors for 
>printer port and serial port.  It plugs into a big connector on the bottom of 
>the laptop.
>
>Many laptops plug into docking stations for use at a desk.  Sometimes/often 
>they have printer ports and/or serial ports.
>
>You could probably do it with just a serial port by flapping one of the 
>output modem control signals.

I could not always depend on having access to built in serial or
parallel ports.  These days of course both are pretty much deprecated
in favor of USB which is useless for this type of work.

>If you don't have a printer port or serial port, how are you getting the 
>interrupt into your box?

Since this was back in the ISA days, I had access to the whole ISA bus
including the interrupts.  I never built a version for PCI but I did
consider it.

The card I built had a pair of 8254 timers, hexadecimal display,
keypad, a bunch of auxiliary I/O, and all of the decoding to use it. I
usually end up building something similar for microcontroller projects
that operates via SPI although serial to a PC running a terminal
program is often better.

>Many years ago, I helped a friend with this sort of thing.  We were working 
>for DEC back in the days before Intel had captured everything and 
>workstations still needed lots of chips and chips had big pins so you could 
>get a scope probe on them.  My part was to connect scope probes to the 
>interrupt line from the ethernet chip and the chip select for the MAC address 
>ROM.  He patched the driver's interrupt routine to read the ROM.

I have also done that on occasion and sometimes still do.  Usually I
solder a little grab point for the probe into place.  Sometimes I will
just add a resistor or transistor buffer depending on impedance issues
and an RG-316 pigtail.

The funny part is that back then, I was using one of the early
Tektronix series TDS oscilloscopes.  Now I do the same thing with an
older Tektronix 2230, 2232, or 2440 series oscilloscope and I have a
word recognizer for my 2440 which works surprisingly well.  At some
point, I need to pick up a DSO that supports variable and infinite
persistence.



More information about the time-nuts mailing list