Frequencies and voltages required to calibrate a 2400-series scope?
A topic that frequently comes up here is how to transfer the NVRAM
contents, or how to "fix" a 2400-series scope that's lost its calibration.
My 2467 was in need of calibration after I restored it to function, both
because it was time, but also because I tweaked all the pots and twiddled
all the settings during repair.
In my ignorance at the time, I believed I needed the equipment specified by
the service manual, and so I now have TG501/SG503/SG504/PG506 gathering
dust in a TM504, as well as e.g. a home-made tunnel diode pulser.
In hindsight, this really wasn't necessary, merely convenient to follow
along with the service manual instructions throughout the calibration
What really was needed was a set of signals with some particular specs. I
believe the specs are fairly easy to meet with misc equipment.
I'm sure I have synthesized RF generators that meet the timing spec with
their dominant hand tied behind their back. The voltages are probably near
enough from my HP 3314A, and probably good enough from any digitized pulse
gen or AWG.
Now, the horizontal timing calibration is all done by using the human eye
for a comparator (half the time with the aid of a second scope). While
TG501 outputs easy to use pulses for the purpose, all you really need is
some kind of synthesized frequency generator. Ideally you want a fast-rise
edge to look at, but at a pinch a sine wave and a hard squint will do.
I think in many cases the question of "how do I copy the NVRAM" would be
best answered with "here's what you need to calibrate your scope".
I volunteer to start that Wiki page, but as I've only done this once, I'd
like ask for help in validating and verifying the page for sanity and
So, as a strawman, the periods involved (that I find from a quick scan of
the 2465 service manual) are:
2ns - this is a sine from TG501.
The vertical and trigger calibration is more automated and it wants a
fairly accurate square wave from ground to:
Other mandatory equipment includes a DMM and a second bench scope -
There are other inputs required for performance validation and transient
response calibration, but IMHO just doing the DAC/CRT, horizontal &
vertical calibration is vastly preferable to copying the calibration
WDYT? Questions, concerns?
I think I could be of some help. Although far from being a metrology professional, I do calibrate all of my scopes (currently 8) using mostly the required equipment stated in the manuals if I have them. But sometimes I don’t. Results are more than acceptable, I think. I get Time Base accuracies up to >= 0.1/02%
I am very interested to follow this thread since I might well need to recalibrate a 2465B and I am not keen on removing the NVRAM without some better de-soldering tools. I have no problem in providing the timing references from some home made synthesised generators and low frequency amplitude calibration from a PG506 but I don't have anything beyond a SG503 for constant amplitude at high frequency.
On Fri, Jul 10, 2020 at 9:40 AM Roger Evans via groups.io <very_fuzzy_logic=
I am very interested to follow this thread since I might well need toHey Roger,
From what I gather (from Chuck's post
https://groups.io/g/TekScopes/message/169366) is that the SG504 or similar
is used for bandwidth performance validation. It sounds like when you play
with the transient response calibration, you'll need to compromise the
response against bandwidth.
I don't have experience with this myself, as when I calibrated my 2467 I
didn't touch the transient response calibration as I didn't have a pulse
generator fast enough at the time.
I have a 2465 I need to re-cap and calibrate one of those days, at which
point I plan to mess with this and make note of the signals required by the