Topics

Tek 2465B Calibrator issues

Marcial Gomez Varas
 

Hi everyone,

I bought some time ago (2017) a TEK 2465B that I didn't have too much time to play with (having 2 kids and a travelling job doesn't help!)
I have been toying with it lately, and I've discovered that the calibrator is showing some really awful square signal.

It's clearly NOT square ! It has quite a bit of slope in the rise and fall lines, and bouncing on the high level.
I have checked and both the output levels and frequency seem to be OK, but somehow the signal is distorted.

Could it be some capacitor? Looking at the service manual it shows to be in the A1 board, section 5 (That should be in the middle left side, am I right?)

Any help would be appreciated !

Cheers,
Marcial

Roy Thistle
 

Hi:
It's obvious, so you probably are... but, are you using a properly calibrated probe?
Regards.

Tom Gardner
 

I don't like the "calibrator" output on the 24x5 series: the period changes according to the timebase so that there is always 5 cycles on the display. At high sweep speeds (i.e. much faster than1ms/div) the poor risetime of the calibrator signal is evident.

The only practical use for a calibrator output is to calibrate the low frequency frequency response of a *10 "high" impedance probe. For that, all you need is a couple of cycles of a 1kHz waveform with a flat top - and you just tweak the scope probe so that its response looks flat. There are many descriptions of how to do that.

Summary: if you have fast sweep speeds (say <1us/div) and the waveform is distorted, don't worry because that is normal. If, OTOH, the waveform is distorted at 200us/div, then tweak the probe. If a problem remains, post a picture.

On 26/03/19 13:29, Marcial Gomez Varas wrote:
Hi everyone,

I bought some time ago (2017) a TEK 2465B that I didn't have too much time to play with (having 2 kids and a travelling job doesn't help!)
I have been toying with it lately, and I've discovered that the calibrator is showing some really awful square signal.

It's clearly NOT square ! It has quite a bit of slope in the rise and fall lines, and bouncing on the high level.
I have checked and both the output levels and frequency seem to be OK, but somehow the signal is distorted.

Could it be some capacitor? Looking at the service manual it shows to be in the A1 board, section 5 (That should be in the middle left side, am I right?)

Any help would be appreciated !

Siggi
 

The point of this on the 2465s is that it allows the user to validate the
time base accuracy on the spot. For this purpose, the poor rise and fall of
the calibrator are unimportant, as you just line some part of the rise or
fall up to the graticule for a check.

It sure looks ugly at fast sweeps though :).

On Wed, Mar 27, 2019 at 20:43 Tom Gardner <tggzzz@...> wrote:

I don't like the "calibrator" output on the 24x5 series: the period
changes
according to the timebase so that there is always 5 cycles on the display.
At
high sweep speeds (i.e. much faster than1ms/div) the poor risetime of the
calibrator signal is evident.

The only practical use for a calibrator output is to calibrate the low
frequency
frequency response of a *10 "high" impedance probe. For that, all you need
is a
couple of cycles of a 1kHz waveform with a flat top - and you just tweak
the
scope probe so that its response looks flat. There are many descriptions
of how
to do that.

Summary: if you have fast sweep speeds (say <1us/div) and the waveform is
distorted, don't worry because that is normal. If, OTOH, the waveform is
distorted at 200us/div, then tweak the probe. If a problem remains, post a
picture.


On 26/03/19 13:29, Marcial Gomez Varas wrote:
Hi everyone,

I bought some time ago (2017) a TEK 2465B that I didn't have too much
time to play with (having 2 kids and a travelling job doesn't help!)
I have been toying with it lately, and I've discovered that the
calibrator is showing some really awful square signal.

It's clearly NOT square ! It has quite a bit of slope in the rise and
fall lines, and bouncing on the high level.
I have checked and both the output levels and frequency seem to be OK,
but somehow the signal is distorted.

Could it be some capacitor? Looking at the service manual it shows to be
in the A1 board, section 5 (That should be in the middle left side, am I
right?)

Any help would be appreciated !



Brian Cockburn
 

Hi Marcial,

Have a look at W2AEW's YouTube channel here: https://www.youtube.com/channel/UCiqd3GLTluk2s_IBt7p_LjA search there for 'compensation' and watch the probe compensation videos. This might help.

Also check out EEVblog's https://www.youtube.com/watch?v=OiAmER1OJh4

Cheers, Brian.

Tom Gardner
 

I'm unconvinced it could check the measurement accuracy in one sense, since I assume the cursors and calibration signal are derived from the same source.

It could be used to check the X linearity and the X magnification, but personally I'd just use other independent equipment for that.

On 28/03/19 01:18, Siggi wrote:
The point of this on the 2465s is that it allows the user to validate the
time base accuracy on the spot. For this purpose, the poor rise and fall of
the calibrator are unimportant, as you just line some part of the rise or
fall up to the graticule for a check.

It sure looks ugly at fast sweeps though :).

On Wed, Mar 27, 2019 at 20:43 Tom Gardner <tggzzz@...> wrote:

I don't like the "calibrator" output on the 24x5 series: the period
changes
according to the timebase so that there is always 5 cycles on the display.
At
high sweep speeds (i.e. much faster than1ms/div) the poor risetime of the
calibrator signal is evident.

The only practical use for a calibrator output is to calibrate the low
frequency
frequency response of a *10 "high" impedance probe. For that, all you need
is a
couple of cycles of a 1kHz waveform with a flat top - and you just tweak
the
scope probe so that its response looks flat. There are many descriptions
of how
to do that.

Summary: if you have fast sweep speeds (say <1us/div) and the waveform is
distorted, don't worry because that is normal. If, OTOH, the waveform is
distorted at 200us/div, then tweak the probe. If a problem remains, post a
picture.


On 26/03/19 13:29, Marcial Gomez Varas wrote:
Hi everyone,

I bought some time ago (2017) a TEK 2465B that I didn't have too much
time to play with (having 2 kids and a travelling job doesn't help!)
I have been toying with it lately, and I've discovered that the
calibrator is showing some really awful square signal.
It's clearly NOT square ! It has quite a bit of slope in the rise and
fall lines, and bouncing on the high level.
I have checked and both the output levels and frequency seem to be OK,
but somehow the signal is distorted.
Could it be some capacitor? Looking at the service manual it shows to be
in the A1 board, section 5 (That should be in the middle left side, am I
right?)
Any help would be appreciated !

Chuck Harris
 

That is not what the OP is talking about!

The 2465 family has a variable frequency calibrator, which
always shows a couple cycles of the calibrator signal,
regardless of the sweep speed.

The purpose of this fancy calibrator is threefold:

1) scope probe calibration
2) sweep speed and cursor/graticule verification
3) it was nifty, and made the scope look special compared
to all scopes that came before.

[The calibrator's frequency is derived from the crystal
oscillator that is the microprocessor's reference on the
A5 controller card. The calibrator's amplitude is derived
by the DAC, and its reference, and a digital constant that
is determined and stored during scope calibration.

The sweep's frequency (rate) is derived by the Miller
integrator's precision capacitors, and some digital calibration
constants stored during scope calibration.

The cursor's relationship to the graticule is determined
by the DAC and the trimmer pots that adjust horizontal and
vertical size and centering... all adjusted during calibration.

It is good to be able to relate all three to each other, and
gives a nice indication of the calibration state of the
scope.]

When you make a calibrator whose frequency rises in lock
step with the sweep speed, you will eventually reach a point
where the scope's bandwidth is too low to show the calibrator's
wave form as a square wave.

Remember, Fourier Transforms: a square wave is a sum of an
infinite series of sine waves all harmonically related,
something like:

f + 1/3 f3 + 1/5 f5 + 1/7 f7 + ... + 1/inf finf

Where f is the frequency, f3 is the third harmonic, f5 is
the fifth harmonic...

Well, to see something that resembles a square waveform,
the vertical amplifier has to be able to pass up to about
the 7th harmonic without distorting its amplitude. Even
then the wave you see will have ripple on its top and bottom.

The operator's guide doesn't adequately inform the user about
this fact, so many a new 2465 user eventually runs into this
on his own, and thinks his scope is broken. For me, that
happened in 1985 after I brought my brand new 2565 home and
ran it through its paces.

I'm not sure which is the worst marketing: showing your flagship
scope does something ugly, or letting your user think his new
scope's broken because it is doing something ugly.

It's all water over the bridge at this point.

-Chuck Harris

Brian Cockburn wrote:

Hi Marcial,

Have a look at W2AEW's YouTube channel here: https://www.youtube.com/channel/UCiqd3GLTluk2s_IBt7p_LjA search there for 'compensation' and watch the probe compensation videos. This might help.

Also check out EEVblog's https://www.youtube.com/watch?v=OiAmER1OJh4

Cheers, Brian.