Re: SUCCESS! The "sick" 475A is now the "fixed" 475A!


On Tue, Dec 8, 2020 at 02:06 AM, Dave Peterson wrote:

Hi Dave,
Please find my comments embedded, after your text.

I bought a cheap function generator that I wouldn't take home to meet my
parents. Much less brag about here.

However, it is a modern piece of technology that I think can outperform my old
465 well enough to tell it how to dance. Modern run of the mill digital
electronics are running at several orders of magnitude higher frequencies than
this old scope.
Not too fast:
A cheap function generator
- generating signals up to and beyond 100 MHz (465 BW) and
- with a predictable/calibrated amplitude?

I'm currently using it to generate square waves to use in place of a time mark
generator. From what I've seen so far from the calibration procedure, and I've
done a little tweaking on my existing scope with an even more embarrassing
source of square waves (a free function generator app on my cell phone), the
primary purpose is to provide a time standard to adjust sweep timing and also
geometry edges for rotation and "Geometry, R1442" vertical alignment with the
graticule. So giving you clean square edges for aligning the X and Y axis of
your trace. The 465 can just make out the vertical edges of around 10-20ns.
Plenty good for 465 edges.
Not for (rising and falling) edges: the 465's spec is 3.5 ns.
Adjusting geometry affects vertical and horizontal amplitude calibration, mostly at the edges.
Trace rotation is gravity- and therefore also orientation-dependent. That's why it is easily adjusted from the front of the 'scope.

Other entries in the calibration procedure use the time mark generator to
calibrate (check and adjust) the B-sweep. I'm not clear on the process yet.
I'm sure there are others. What I've seen of time mark generator output look
like pulses more than square waves. But the effect is the same - regular
vertical edges. The cheap FG will do pulses too. Anything actually.
As long as the distance between rising or falling edges of the FG (IOW the period) are accurate to within at least 1%, that's ok.

The point being, I think a decently competent square wave generator can
suffice as a stand-in for the time mark generator.
Very limited at best for horizontal calibration of a 465's (and 475's) high-speed horizontal timing (1 - 50 ns).

Standards for level and
time are subject to suspicion. But again, my sneaking suspicion is that the
cheap function generator is within the tolerances of what the 465 can measure.
I disagree on several points, see above.

I continue to contemplate how to validate it's time and levels.
It's perfectly possible to use DC voltages, checked with your DMM, for adjustment of vertical deflection (sensitivity), IOW "levels".

But this is far better than nothing, for a lot less $$.
If you intend to expand your collection and perform calibration and adjustment (can't help myself), consider hunting for a PG506 and TG501 at least. Prices have gone down - and so has availability...

Calibrated? No. Tuned up? Hell yes!

Join to automatically receive all group messages.