开云体育

Jon,
Do you mean that you couldn't see the following?
Sometimes in the web interface you have to hit the button "Show Quoted Text'.

Raymond

Raymond ACK on all,

I forgot about " show quoted text" to see interleaved response.

I referenced Jim W avalanche and Tek TD pulsers only as possible alternatives to Leo Bodnar, the gold standard

The PG506 is reasonable on fast but nowhere near as good the Leo B. I never used the avalanche JW.

I tried old Tek TD pulser adapter, but tweaky, high level input required.

Am lucky to have two of the old versions of the Leo Bodnar, handy to compare two plugins or channels.


Jon

So I started to go through the cal steps - the power supplies, DAC ref and CRT (cal08) steps seem reasonably straightforward. Then I tried cal01, the horizontal cal step.

The initial step of probing TP800 and twiddling the delta ref knob seemed straightforward, but I'd love to know what this is for, the manual does not say.

Oh well, onto the next step, connect up a time mark generator. I used the Siglent function generator, it seems to be adequate when using it in Pulse waveform mode, specifying the period e.g. 0.1mS where it says set the time markers. You need to set a suitable pulse width, not too large but not so small that the pulses become difficult to see. But it seems adequate.

I believe I was able to get up to step 16 correct, but the next steps (starting from 17) baffled me. The manual says connect the time marker to both the 2465B and a second scope's CH1, and the Gate Out of the 2465B to the second scope's CH2 and trigger off that channel. But then the next paragraph (jj in the manual) completely confused me. What is it all about? What's the purpose of the second scope, what is a 'usable time mark' ? It's as clear as mud.

- Keith

On Tue, Jan 21, 2025 at 12:15?PM Keith via groups.io <keith=
[email protected]> wrote:

The initial step of probing TP800 and twiddling the delta ref knob seemed
straightforward, but I'd love to know what this is for, the manual does not
say.

...

I believe I was able to get up to step 16 correct, but the next steps
(starting from 17) baffled me. The manual says connect the time marker to
both the 2465B and a second scope's CH1, and the Gate Out of the 2465B to
the second scope's CH2 and trigger off that channel. But then the next
paragraph (jj in the manual) completely confused me. What is it all about?
What's the purpose of the second scope, what is a 'usable time mark' ? It's
as clear as mud.

Hey Keith,

I remember wondering the same things when I calibrated my 2467, though at
the time I didn't understand enough of how these scopes work to do more
than follow the instructions.
There is some mention of calibration and calibration constants in the
Theory of Operation section, maybe that'll help?

The principle involved is that you first calibrate the CRT and the cursors,
and then everything else is adjusting gains & offsets to a given time
reference to align sweep times, and delay offsets (I believe) to the
calibrated CRT (and cursors), as well as calibrating the the two sweeps (A
and B) with respect to each other. Note that sweeps on these scopes are
calibrated even with the horizontal VAR control out of detent, which is not
the case for earlier scopes (I believe).
The sweeps need their delays calibrated as you can e.g. measure time delays
with dual delay display and e.g. overlaying edges or waveforms. The A sweep
delays can be calibrated by displaying the B-trace after a delay (or dual
delays), but the B-sweep delays are calibrated by looking at the B sweep
gate, I believe.
If you look at the sweep schematics, you'll see that the delay gate of the
B-sweep hybrid is routed to the sweep gate output, which I believe is only
useful for calibration, and hence I believe the steps using the B-delay
gate are calibrating the B-sweep delay.
It's been a while since I dug into this. I'm fairly sure I never sussed out
the purpose of each of the steps, and my reasoning might have been off to
begin with, but maybe you'll figure this out?

... time passes ...

Now that I think about it, I can see how the A sweep delay(s) are used to
highlight the A sweep, as well as to implement B delay sweeps and dual
delay sweeps. I'm struggling to figure out what the B sweep delay reference
is actually used for. After reading the ToO section, I believe the B-sweep
may be used for options to measure stuff, maybe e.g. pulse timings, meaning
it'd need calibrated delays?

Siggi

Hi Siggi,

Thanks, I've been able to run through the CAL01 (X) and CAL02 (Y), however I'm still lost at step 'jj'. I connect up a second scope as it says, with CH1 from the marker generator (in my case the Siglent) and CH2 coming from the Gate output of the 2465B, and set triggering to CH2.

Then for cal step 17 set the Siglent pulse gen to 500mV p/p (is that supposed to be 0 to +500mV, or -250mV to +250mV? I used the latter), 1uS period, 100nS pulse width. Then I see that waveform as expected on the second scope.

But next, the manual says:

"adjust the Delta control for roughly the listed number of markers over the center 8 divisions, then superimpose markers on bench scope screen"

- I've got the markers i.e. the pulse gen output visible on the bench scope, and the gate waveform, what does 'superimpose' mean? Have the two traces on top of each other?

It then says "Manually set SEC/DIV setting of bench scope to keep a usable time mark as listed in Table 5-3"

- I presum if just means keeping the pulse train on CH1 and the gate waveform on CH2 visible?

Then it says "Use IUT DELAY POS to bring markers on screen".

- Err, which screen? Is that the 'Delta REF / DELAY POS' control? Where are they if they are not 'on screen'?

Next, "When markers are superimposed"

- Again, on what screen?

I just skipped thru steps 17-34 as it was not clear how to do these steps. I do wish Tek had added some images of what you're supposed to be doing on the Table 5-3 steps. The waveforms after cal seem to have the correct amplitude and time with the A timebase, but probably the delayed timebase is not right as you mention.

- Keith

Siggi: We had similar questions years ago when we did our frst full CAL.

Very important to do every section and be sure thhat CAL suceeds to write (rather than the deraed "out of range")

Most citical is the CTT option CAL.

1/ The gen spec for certain CAL sections is very specific, and needs the orig TEK cal gens or a very good replication.
IWe use the TEK and avoid the copies like siglent Chinese.

I suspect Chinese not produce the exact levels and duty cycle, freq required by TEK.

2/ The 246X/A/B CAL has been detailed extensivel especiall for the unlear steps and screens.

Search both this group and Tekscpes2 @ groups.io and EEVBLOG 2465 teardown for CAL errors and proceedure notes.

3/ Especiall Chuck Harris is the expert but banned from Tekscopes for unknown reasons years ago.

After your search and further work suggest to Ask Chhuck on Tekscops2.

Bon chance,

Jon

On Wed, Jan 22, 2025 at 6:54?AM Keith via groups.io <keith=
[email protected]> wrote:

Then for cal step 17 set the Siglent pulse gen to 500mV p/p (is that
supposed to be 0 to +500mV, or -250mV to +250mV? I used the latter),

I believe TG501 has positive-going pulses, but I don't think it matters for
horizontal calibration. It is perhaps safer to use positive going pulses,
as the scope under calibration might use a fixed DC level trigger.
It does matter for vertical calibration that you use GND to the indicated
level, though.

- Err, which screen? Is that the 'Delta REF / DELAY POS' control? Where
are they if they are not 'on screen'?

Next, "When markers are superimposed"

- Again, on what screen?

I only did this once, years ago, for my 2467, and I just don't remember the
details. Keep it in mind that the instructions are written assuming you're
using another analog scope as a bench scope, presumably that's why the
instructions on how to set up delays and such. With a digital storage scope
you don't need to worry about any of that, as you can just position the
trigger wherever you want on the screen, and the timing calibration on the
bench scope is not going to be in question.
As to what you're supposed to align, I just don't remember, I hope someone
who's done this more recently than I steps in here.

Assuming this is calibrating something to do with the B sweep delay,
presumably you ought to be aligning two copies of the timing marks.
If you look at A1U850A, you'll see that DLY_REF_0 and _1 can both make
their way to the B-sweep DR (delay ref) pin, so presumably this is
calibrating out something to do with the alternate delay paths?
If you have the case off, you can verify this by looking at the control
signal on U850A while sitting at step (jj). If it's alternating, then the
two delay refs are being alternately routed to the B-sweep delay ref.
Note that since the control signal is under microcontroller-control, the
switching will be quite slow, so maybe you need to bump the persistence on
your bench scope to see two instances of the time markers?

On Wed, Jan 22, 2025 at 10:43?AM Sigur?ur ?sgeirsson via groups.io <siggi=
[email protected]> wrote:

Assuming this is calibrating something to do with the B sweep delay,
presumably you ought to be aligning two copies of the timing marks.

... on the bench scope ...

If you look at A1U850A, you'll see that DLY_REF_0 and _1 can both make
their way to the B-sweep DR (delay ref) pin, so presumably this is
calibrating out something to do with the alternate delay paths?
If you have the case off, you can verify this by looking at the control
signal on U850A while sitting at step (jj). If it's alternating, then the
two delay refs are being alternately routed to the B-sweep delay ref.
Note that since the control signal is under microcontroller-control, the
switching will be quite slow, so maybe you need to bump the persistence on
your bench scope to see two instances of the time markers?

It occurs to me that alternatively you might be aligning the start time of
the B-sweep through two alternate paths. There is a mention in the ToO that
the B sweep can run concurrently with the A sweep, so perhaps the delay
reference is used to make sure the two sweeps are in time, whichever way
gates them.

... time passes ...

So I looked at the service manual for my 2467, which has essentially the
same instructions for step (jj), but it's only possible to route -1.25V or
the A-sweep delay reference to the B-sweep delay reference.
So, it's still possible that this is calibrating out something to do with
the two paths the B-sweep can be initiated. See page 3a-22, where it
discusses the BDCA signal. I could imagine alternating between the BDCA and
regular B delay sweep modes, adjusting the A and B delay reference voltages
in order to effectively measure the initiation delay of the B sweep through
the BDCA path.
Honestly, I just don't know. but again, if you probe the BDCA signal while
hanging out on the (jj) step, you'd be able to find out :).

On Wed, Jan 22, 2025 at 11:14?AM Sigur?ur ?sgeirsson via groups.io <siggi=
[email protected]> wrote:

On Wed, Jan 22, 2025 at 10:43?AM Sigur?ur ?sgeirsson via groups.io <siggi=
[email protected]> wrote:

Assuming this is calibrating something to do with the B sweep delay,
presumably you ought to be aligning two copies of the timing marks.

... on the bench scope ...

If you look at A1U850A, you'll see that DLY_REF_0 and _1 can both make
their way to the B-sweep DR (delay ref) pin, so presumably this is
calibrating out something to do with the alternate delay paths?
If you have the case off, you can verify this by looking at the control
signal on U850A while sitting at step (jj). If it's alternating, then the
two delay refs are being alternately routed to the B-sweep delay ref.
Note that since the control signal is under microcontroller-control, the
switching will be quite slow, so maybe you need to bump the persistence

on

your bench scope to see two instances of the time markers?

It occurs to me that alternatively you might be aligning the start time of
the B-sweep through two alternate paths. There is a mention in the ToO that
the B sweep can run concurrently with the A sweep, so perhaps the delay
reference is used to make sure the two sweeps are in time, whichever way
gates them.

... time passes ...

So I looked at the service manual for my 2467, which has essentially the
same instructions for step (jj), but it's only possible to route -1.25V or
the A-sweep delay reference to the B-sweep delay reference.
So, it's still possible that this is calibrating out something to do with
the two paths the B-sweep can be initiated. See page 3a-22, where it
discusses the BDCA signal. I could imagine alternating between the BDCA and
regular B delay sweep modes, adjusting the A and B delay reference voltages
in order to effectively measure the initiation delay of the B sweep through
the BDCA path.
Honestly, I just don't know. but again, if you probe the BDCA signal while
hanging out on the (jj) step, you'd be able to find out :).

And perhaps I'm just totally wrong, and this is simply calibrating the B
time base and/or delay:
/g/TekScopes/topic/2465b_vert_cal_02_cal_01/29871399.
Chuck will know this better than I do, and I don't pretend to offhand
understand how the +-B error thang works.

Hmm. It does rather look like you need to use an analog scope with a delayed B timebase as the 'bench' scope, from reading Chucks's comment #154417.

Time to read all of that thread...

- Keith

Hey Keith,

I'm scrubbing through the calibration instructions, trying to remember what
I'd sussed out when I was looking into this.
The principle is that the CRT is set up against a reference sweep, whose
deflection factor D V/DIV has been measured using DLY_REF_0 & _1, using
you, the puny human, as a comparator. This sweep is then mated to the CRT
by adjusting the horizontal gains.
Subsequent sweep speeds are then set up by initiating a strawman sweep
using a strawman current Is that is close to the correct slope, then
measuring the strawman slope Ss V/SEC using the DLY_REF_0 and _1, again
with the human as a comparator. Given Is and Ss, it's then possible to
interpolate the correct I and S for the sweep.

Here are the important steps, and what I believe they do:
o) Verify that DLY_REF_0 and _1 are in concordance by lining up both
B-sweeps on a reference. I assume this will produce that dreaded LIMITS
error if the internal control voltages (aka DAC codes) don't line up fairly
precisely.
r) Find the slope of the reference sweep timing by aligning both B sweeps
on different timing signal reference markers.
aa) Adjust horizontal gain to fit the reference sweep precisely on CRT.
The CRT is now calibrated in horizontal D V/DIV deflection factor, and the
scope knows the deflection factor D.

ff et al) Measure the slope S of various timings the same way as in step r
(using the human as comparator). Given that the scope knows the reference
current producing S, it can interpolate the correct reference current to
calibrate S to the deflection factor D.

In the 2465, as Chuck mentions, the B sweep timings were set simply by
aligning time markers to the graticule.
In the 2465A and later, the calibration procedure for the B sweep works the
same way as the A sweep, except the B gate and a bench scope are used to
display the waveforms for the human comparator.
The way this has to work is by alternating B sweep delays, which should
still produce a trace on the DUT CRT.
The difference is that twiddling the delta and delta ref knobs here will
change the timing of the B-gate with respect to the timing signal, rather
than to show where on the A trace the B-trace sits.

jj et al) Measure the slope of B sweeps by overlaying two delayed versions
of the timing reference on a bench scope, then interpolate the correct
reference current as in ff).

And that's all I have to say about that (though let the truth prevail if I
haven't reasoned this out right :).

Siggi

On Wed, Jan 22, 2025 at 11:41?AM Keith via groups.io <keith=
[email protected]> wrote:

Hmm. It does rather look like you need to use an analog scope with a
delayed B timebase as the 'bench' scope, from reading Chucks's comment
#154417.

Time to read all of that thread...

- Keith


On Wed, Jan 22, 2025 at 04:27 PM, Sigur?ur ?sgeirsson wrote:

And perhaps I'm just totally wrong, and this is simply calibrating the B
time base and/or delay:
/g/TekScopes/topic/2465b_vert_cal_02_cal_01/29871399.
Chuck will know this better than I do, and I don't pretend to offhand
understand how the +-B error thang works.