开云体育

On 23/06/18 04:15, lop pol via Groups.Io wrote:

I was able to clean it up a little more. Do you think the ringing may be my
50ohm termination?
/g/TekScopes/photo/59361/5?p=Name,,,20,1,0,0

If that is 5ns/div then the risetime looks like ~3.5ns, which is what I would
expect.

It doesn't look like a termination problem to me; typically those vary
according
to the length of the cable.

I would be satisfied with that trace.

In real circuits, poor probing techniques will introduce more serious errors
than shown on that trace.

I was able to clean it up a little more. Do you think the ringing may be my 50ohm termination?
/g/TekScopes/photo/59361/5?p=Name,,,20,1,0,0

This the the rise time now at 5mV. I could probably get it better but but my
bnc cables and 50ohm terminator are not great quality. Should I keep working
at it?

/g/TekScopes/photo/59361/4?p=Name,,,20,1,0,0

On Thu, Jun 21, 2018 at 12:00 pm, Fabio Trevisan wrote:

Hello lop pol,

I've followed up this topic from the start, without having actually jumped

in,

hoping since the beginning that someone would ask or make remarks about two

or

three important things (to check, or to be aware of).

1. That's a question... What type of Vertical Output board does your 465

have.

the one discrete or the one with the custom Tek IC?
It's important because the High Freq. compensation networks are different

from

one type of board to the other and since your problem manifests in both
channels, there's a great likelihood the problem is located on the Vertical
Output Amplifier board, instead of on the pre-amplifier. Without that
information is difficult to advise further.

2. If your vertical amplifier is of the "Teck IC" kind, look also to the

BIAS

trimpot. If the wiper is open, the amplifier may be operating on the wrong
bias point and that is supposed to impact the high-frequency performance.
Since the procedure for calibrating the bias calls for applying a 100Mhz
signal and adjusting the bias for the highest p-p display, from there we
deduct it has a marked effect on the low nanosecond time constant, which is
exactly where your issue lies.

3. Besides what Raymond already mentioned, that is not uncommon for some of
those trimmer caps fail open (I had 3 of those failed on my 464), I want to
bring to your attention that ALL the trimpots that are used on the H.F.
compensation networks are connected as "rheostats" so, a false contact or

open

wiper will make any of them appear open, or in the best case, the full
trimpot's resistance, which will be just as bad for the step response.
Since many of them are in series with trimmer caps, it's easy to check them

in

circuit, without needing to lift any leg, or remove any component.

Last but not least, despite your problem clearly doesn't seem to be on the
pre-amplifiers, when you find out the culprit and eventually replace one
faulty trimmer cap or faulty trimmer resistor, you will eventually need to

go

through the Vertical High Frequency step response calibration, and for that,
it's important that you follow the procedure, and from your description, I
think you're already doing wrong...
While choosing for the right input level / input attenuator (pass-through)

and

vertical attenuator settings, it's important to do all the High-Frequency
calibration with the oscilloscope set to the most sensitive setting... i.e.:
5mV/div (not 10mV, not 100mV... and not 200mV/div).
In other words, you need to source 25mVpp to the oscilloscope's input (for a

5

division display).
This is what will assure you that none of the high impedance input

attenuators

are upsetting the subtle high frequency adjustments and artifacts.
While - in theory - the high-impedance input attenuators shouldn't degrade

the

step response of the osciloscope, they're not perfect and they do cause some
minor disturbances and aberrations to the step response (and they can also

be

misadjusted, something that you will only correct at a later stage) so, to
make a "clean" H.F. calibration, you need to make sure you got them (the

high

impedance input attenuators) out of the way.
The only compromise to that rule is the input probe itself... If you intend

to

use the oscilloscope with a particular probe (let's say, a 10:1), then you
need to connect that probe to the input, use the scope set to 5mV/div, and
apply, accordingly, a 250mVpp signal to the probe (if it's a 10:1 probe).

This

will assure you will compensate in the scope, for the probe's H.F.
inaccuracies or artifacts.
Yet on this topic, you need to be sure to "source" that signal (regardless

if

directly to the input or through your preferred probe), from a low impedance
source (50Ohm).
The latter is important because you don't want the frequency response of the
incoming signal to be affected by variations caused by the 'scope's input
capacitance, or your probe's input capacitance.
The lower the impedance of the source signal, the less the input capacitance
will matter and, of course, you will also want to assure the interconnect
cable is properly terminated.

Good luck with that,

Krgrds,

Fabio


On Sun, Jun 17, 2018 at 06:47 pm, lop pol wrote:

Hey guys. I have done a lot of work to this 465. I had initially planned

on

selling it to help me buy some plugins for my 7633. I decided I don't want

to

sell it and am keeping it. Today I checked its rise time with a pg506 fast
rise. Both channel 1 and 2 closely resemble each other and they are not

close

to right. Before I start trying to calibrate this thing, does it look like
there is some type of component issue left to fix? I am thinking since

both

channels are so alike there is something wrong besides just out of
calibration. Thanks guys.

here are some pictures

/g/TekScopes/photo/59361/0?p=Name,,,20,1,0,0

/g/TekScopes/photo/59361/1?p=Name,,,20,1,0,0

The Tek 475 is analog, with a CRT screen, so the tools for LCD screens
will probably not work. The intensity knob may be turned down too low
(as Bob mentions), but on my 475, if I turn down the intensity until I
can barely see the trace on just the A sweep, and then go to Mix, I can
still see the A trace along with the B trace. But when I turn it down
further, the B trace is still visible as the A disappears. There may be
a dirty Mix switch or some other switch.


On 6/22/2018 8:27 AM, Jeff Urban wrote:

On Thu, Jun 21, 2018 at 02:29 pm, Bob Albert wrote:

Is the intensity set too low?
Bob

Sounds like you are not seeing the left side of the trace. Couple

possible reasons for this, one is that the brightness or pedestal on your
monitor is set too low. There are monitor test programs for this, as it is
hard to tell just where black is black but none of the grays are black. One
such program is called ntest.exe, which is totally stand alone, would run
off anything, like a CD or stick or whatever. One of the patterns is
exactly for that and says adjust until you see some particular block in the
pattern. That should eliminate any black clipping.

Another possibility is if you have an edge lit LCD monitor. As the

viewing angle changes vertically in those, the brightness changes. This
effect varies, some models are better/worse than others. Some are terrible
and you only get a limited range or it does all kinds of funky crap like
show a negative image or all the color disappears if you are too low or
high, things like that.

The one I got - ntest.exe, is from Nokia. If you can't find a suitable

program I can upload it or you can hit me up with an email and I can send
it as an attachment in a reply.

Bert:
Do you have a CRT that will fit my needs? Sphere does not.

I do not get to see your email address in the "digest view" of the group for which I had opted.
Please post again so I can grab your email address.

Thanks!

Is the intensity set too low?
Bob