Finally last night I was able to sit down and figure out the probe
adjustment.
I have 5 probes. One 10x Tek probe, one 20x Tek probe, one
Probemaster 1x & 10x probe, and two other 1x & 10x probes from China.
The 10x Tek and the Probemaster probes are 100 MHz rated and made in
the USA. The Tek 20x probe is 200Mhz rated and USA made. The Chinese
probes are 100MHz rated.
I realized the probes need to be adjusted only in the 10x mode. I
didn't know that. That for sure was part of the problem.
Looking at signals from my function generator it seemed that the
voltage was off. Then I realized that my meter displays the RMS value
of the AC signal and the 'scope displays the peak values.
I got all the probes to display a nice square wave in the 10x
setting. I don't know why I got really weird results before but it may
be because I was in my machine shop with CNC machines running. At home
with everything quiet all the probes behaved like they were supposed
to. At the 10x setting.
But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.
It didn't seem to matter if the 'scope was set to AC or DC
coupling. It also didn't matter if the signal being measured came from
the 'scope or from the function generator.
Is this normal, for the square wave displayed to be perfect in the
10X setting and slightly distorted in the 1X setting?

The 'scope is a Tek465B.
The 20X probe is a TEK model P5120
The 10X probe is a Tek model P6053B
The Probmaster is model 3901-2
The Chinese probes only have a model number-P6100

So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?
Thanks,
Eric

On 2/14/19 1:25 PM, wrote:
The Chinese probes only have a model number-P6100

That alleges they are clones of the Tek P6100

So, as near as I can tell, the probes are all right,
at least they all act the same. But what causes the
distorted square wave at the 1x setting?

All scopes have a basic input impedance. For a 1:1
probe to "look right" the probe has to match that.
Mostly it's a matter of the scopes input stage being
properly compensated for the probe's cable.


--
"I am a river to my people."
Jeff-1.0
WA6FWi
http:foxsmercantile.com

On Thu, 14 Feb 2019 11:25:53 -0800, wrote:


So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?

It's caused by the lack of high frequency response (i.e. rise time) in
X1 mode. For a typical X10/X1 switchable problem, the frequency
response is MUCH higher in X10 than in X1. For example:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-250mhz-x1-x10
Notice that the bandwidth is 250MHz in X10 but only 6MHz in X1.
This one is closer to what I think you bought:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-60mhz-x1-x10
60MHz bandwidth in X10 and 6MHz bandwidth in X1.




--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

wrote:



So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?


** What frequency is your test wave ?

About 1kHz is usual.


.... Phil

On 2/14/2019 11:25 AM, wrote:

But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.

If you emailed your doctor and said, "I feel slightly bad,"
what diagnostic would you expect?

At least describe it in such a way that one could draw an
accurate picture from your description. And yes, the time
dimension is relevant.


What's the output impedance of your function generator?
What do you see when you plug the function generator directly
into the scope, via what kind of cable?

On Thursday, February 14, 2019 at 2:17:31 PM UTC-5, wrote:
Finally last night I was able to sit down and figure out the probe
adjustment.
I have 5 probes. One 10x Tek probe, one 20x Tek probe, one
Probemaster 1x & 10x probe, and two other 1x & 10x probes from China.
The 10x Tek and the Probemaster probes are 100 MHz rated and made in
the USA. The Tek 20x probe is 200Mhz rated and USA made. The Chinese
probes are 100MHz rated.
I realized the probes need to be adjusted only in the 10x mode. I
didn't know that. That for sure was part of the problem.
Looking at signals from my function generator it seemed that the
voltage was off. Then I realized that my meter displays the RMS value
of the AC signal and the 'scope displays the peak values.
I got all the probes to display a nice square wave in the 10x
setting. I don't know why I got really weird results before but it may
be because I was in my machine shop with CNC machines running. At home
with everything quiet all the probes behaved like they were supposed
to. At the 10x setting.
But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.
It didn't seem to matter if the 'scope was set to AC or DC
coupling. It also didn't matter if the signal being measured came from
the 'scope or from the function generator.
Is this normal, for the square wave displayed to be perfect in the
10X setting and slightly distorted in the 1X setting?

The 'scope is a Tek465B.
The 20X probe is a TEK model P5120
The 10X probe is a Tek model P6053B
The Probmaster is model 3901-2
The Chinese probes only have a model number-P6100

So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?
Thanks,
Eric

Cool. Hey Eric, I just want to say that one reason I enjoy
sharing whatever small amount of knowledge I have with you,
is that you'll always follow up and tell me how it turned out.

I wish I had more customers like that. They send me
a problem... I send them off a few potential solutions,
and then ~1/2 the time I don't hear back. I assume something
worked... but I don't which of the three solutions it was,
or they made some bone head mistake and are embarrassed.

Oh, here's a machinist question for you. When tapping
holes in an easy metal, Al 6061 say, with a through hole
I've got these 'gun'* taps that shoot the chip out the
bottom of the hole. If I'm doing a blind hole is there
a tap that shoots the chip back up the tap.
(chip may not be the right word.)
I tap mostly small stuff
4-40, 6-32, 2-56... in approximate order of usage

George H.
*gun taps is what I call them, there is a GN on the
stem.

On 15/2/19 12:50 pm, wrote:
On Thursday, February 14, 2019 at 2:17:31 PM UTC-5, wrote:
Finally last night I was able to sit down and figure out the probe
adjustment.
I have 5 probes. One 10x Tek probe, one 20x Tek probe, one
Probemaster 1x & 10x probe, and two other 1x & 10x probes from China.
The 10x Tek and the Probemaster probes are 100 MHz rated and made in
the USA. The Tek 20x probe is 200Mhz rated and USA made. The Chinese
probes are 100MHz rated.
I realized the probes need to be adjusted only in the 10x mode. I
didn't know that. That for sure was part of the problem.
Looking at signals from my function generator it seemed that the
voltage was off. Then I realized that my meter displays the RMS value
of the AC signal and the 'scope displays the peak values.
I got all the probes to display a nice square wave in the 10x
setting. I don't know why I got really weird results before but it may
be because I was in my machine shop with CNC machines running. At home
with everything quiet all the probes behaved like they were supposed
to. At the 10x setting.
But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.
It didn't seem to matter if the 'scope was set to AC or DC
coupling. It also didn't matter if the signal being measured came from
the 'scope or from the function generator.
Is this normal, for the square wave displayed to be perfect in the
10X setting and slightly distorted in the 1X setting?

The 'scope is a Tek465B.
The 20X probe is a TEK model P5120
The 10X probe is a Tek model P6053B
The Probmaster is model 3901-2
The Chinese probes only have a model number-P6100

So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?
Thanks,
Eric

Cool. Hey Eric, I just want to say that one reason I enjoy
sharing whatever small amount of knowledge I have with you,
is that you'll always follow up and tell me how it turned out.

I wish I had more customers like that. They send me
a problem... I send them off a few potential solutions,
and then ~1/2 the time I don't hear back. I assume something
worked... but I don't which of the three solutions it was,
or they made some bone head mistake and are embarrassed.

Oh, here's a machinist question for you. When tapping
holes in an easy metal, Al 6061 say, with a through hole
I've got these 'gun'* taps that shoot the chip out the
bottom of the hole. If I'm doing a blind hole is there
a tap that shoots the chip back up the tap.
(chip may not be the right word.)
I tap mostly small stuff
4-40, 6-32, 2-56... in approximate order of usage

George H.
*gun taps is what I call them, there is a GN on the
stem.

GN for Gun Nose.

You want a spiral flute tap, see pictures he
https://www.cnccookbook.com/different-types-thread-taps-complete-guide-choosing/

Clifford Heath.

On Thu, 14 Feb 2019 17:50:38 -0800 (PST), wrote:

On Thursday, February 14, 2019 at 2:17:31 PM UTC-5, wrote:
Finally last night I was able to sit down and figure out the probe
adjustment.
I have 5 probes. One 10x Tek probe, one 20x Tek probe, one
Probemaster 1x & 10x probe, and two other 1x & 10x probes from China.
The 10x Tek and the Probemaster probes are 100 MHz rated and made in
the USA. The Tek 20x probe is 200Mhz rated and USA made. The Chinese
probes are 100MHz rated.
I realized the probes need to be adjusted only in the 10x mode. I
didn't know that. That for sure was part of the problem.
Looking at signals from my function generator it seemed that the
voltage was off. Then I realized that my meter displays the RMS value
of the AC signal and the 'scope displays the peak values.
I got all the probes to display a nice square wave in the 10x
setting. I don't know why I got really weird results before but it may
be because I was in my machine shop with CNC machines running. At home
with everything quiet all the probes behaved like they were supposed
to. At the 10x setting.
But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.
It didn't seem to matter if the 'scope was set to AC or DC
coupling. It also didn't matter if the signal being measured came from
the 'scope or from the function generator.
Is this normal, for the square wave displayed to be perfect in the
10X setting and slightly distorted in the 1X setting?

The 'scope is a Tek465B.
The 20X probe is a TEK model P5120
The 10X probe is a Tek model P6053B
The Probmaster is model 3901-2
The Chinese probes only have a model number-P6100

So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?
Thanks,
Eric

Cool. Hey Eric, I just want to say that one reason I enjoy
sharing whatever small amount of knowledge I have with you,
is that you'll always follow up and tell me how it turned out.

I wish I had more customers like that. They send me
a problem... I send them off a few potential solutions,
and then ~1/2 the time I don't hear back. I assume something
worked... but I don't which of the three solutions it was,
or they made some bone head mistake and are embarrassed.

Oh, here's a machinist question for you. When tapping
holes in an easy metal, Al 6061 say, with a through hole
I've got these 'gun'* taps that shoot the chip out the
bottom of the hole. If I'm doing a blind hole is there
a tap that shoots the chip back up the tap.
(chip may not be the right word.)
I tap mostly small stuff
4-40, 6-32, 2-56... in approximate order of usage

George H.
*gun taps is what I call them, there is a GN on the
stem.
The taps that direct the chips back out of the hole are called high
spiral taps. Spiral point taps, AKA gun taps, eject the chips ahead
of the tap. I use high spiral taps a lot but I don't like them. They
are weaker than spiral point taps. I know, the terminology is
confusing, especially since the high spiral taps are actually high
helix taps.
For blind holes in ductile materials, such as wrought aluminum and
mild steel, I like to use roll form taps, AKA form taps, AKA roll
taps. In fact, I use roll taps whenever I can because of the stronger
thread and because no chips are generated. These taps require a larger
hole and hole size must be controlled well because the tap just
displaces the metal. With a roll tap you can tap right down to the
bottom of a hole because there are no chips generated.
Because form taps displace the material the thread is stronger than
a cut thread. Virtually all the fasteners you will find in your local
hardware store will have formed threads.
Materials that can be roll or form tapped include all the wrought
aluminum alloys, such as 6061, 6063, 5052, etc. Mild steel, 1144
steel (stressproof), 12L14 steel (leadloy) and 303 stainless steel
also tap well with form taps. 304 and 316 can be form tapped but great
care must be taken to avoid broken taps. Tap life will also be short
in 304 and 316. But if you need threads all the way to the bottom and
want to do it in one pass roll taps can work.
Materials that should not be formed tapped are the cast aluminum
alloys, half hard brass, cast iron, and the like. This is because the
material is not ductile and the resulting thread will have tiny
fractures, resulting in a weak thread.
I have tapped thousands of holes with 2-56 form taps.
Always use cutting oil when drilling and tapping holes, never motor
oil. Using motor oil leads to broken taps and bad finishes and poor
tool life.
Hope that helped.
Eric

On Thu, 14 Feb 2019 13:47:40 -0600, Fox's Mercantile
wrote:

On 2/14/19 1:25 PM, wrote:
The Chinese probes only have a model number-P6100

That alleges they are clones of the Tek P6100

So, as near as I can tell, the probes are all right,
at least they all act the same. But what causes the
distorted square wave at the 1x setting?

All scopes have a basic input impedance. For a 1:1
probe to "look right" the probe has to match that.
Mostly it's a matter of the scopes input stage being
properly compensated for the probe's cable.
Can this basic input impedence be adjusted? Should it be?
Eric

On Thu, 14 Feb 2019 13:09:32 -0800, Jeff Liebermann
wrote:

On Thu, 14 Feb 2019 11:25:53 -0800, wrote:


So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?

It's caused by the lack of high frequency response (i.e. rise time) in
X1 mode. For a typical X10/X1 switchable problem, the frequency
response is MUCH higher in X10 than in X1. For example:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-250mhz-x1-x10
Notice that the bandwidth is 250MHz in X10 but only 6MHz in X1.
This one is closer to what I think you bought:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-60mhz-x1-x10
60MHz bandwidth in X10 and 6MHz bandwidth in X1.
The frequency I was using was 1KHz. That's what the 'scope puts out
and what the function generator was set at. With a 100MHz probe
shouldn't a 1 KHz signal be low enough to avoid the rise time issue? I
didn't even know about the difference in the frequency response. Is
this one of the reasons that people recommend using the 10X setting
unless the 1x setting is needed? That the only time the 1X setting is
used is when the signal needs to be measured un-attenuated?
Thanks,
Eric

On Thu, 14 Feb 2019 16:30:11 -0800, Mike wrote:

On 2/14/2019 11:25 AM, wrote:

But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.

If you emailed your doctor and said, "I feel slightly bad,"
what diagnostic would you expect?

At least describe it in such a way that one could draw an
accurate picture from your description. And yes, the time
dimension is relevant.


What's the output impedance of your function generator?
What do you see when you plug the function generator directly
into the scope, via what kind of cable?
I forgot to write the frequency. The frequency being used was 1KHz
at 300 MV, from both the 'scope and the function generator.
I only have 1 cable that connects to the function generator, it has
a BNC connector at one end and alligator clips at the other end. So at
this time I cannot directly connect it to the 'scope, I must use the
probes. But the probes are what I was testing and I got the same
results no matter where the signal was coming from, as I stated in my
original post.
The output impedence of the function generator is 50 ohms.
If I was able to connect the function generator directly to the
'scope with the proper cable what would this tell me? I guess if the
square wave shape was still distorted the same way and the same amount
then it would be the 'scope that has the problem? And if the square
wave was perfect then it would be the probes causing the distortion.
Look, I try to post a question with adequate information. It seems
like I'm always leaving something out though. It's not because I'm
being lazy or trolling. It's because I miss ****, just like eveyone
else. And I don't mind being asked for more info, especially if the
reasons for the requested info are listed. This helps me to be more
accurate in my replies. And helps me learn, which is why I post
questions in the first place.
Thanks,
Eric

In article ,
wrote:

Can this basic input impedence be adjusted? Should it be?

Sometimes, but that's usually a "service adjustment and repair" sort
of thing, not something that one is usually expected to do during
normal use. In a lot of scopes there's no adjustment that doesn't
require e.g. unsoldering parts.

There's enough variation between types of scope, that it's probably
most common to have a set of probes for each scope. Pick (and if
necessary adjust) the probes to match the input characteristics of the
scope, and then don't fiddle with 'em.

On Friday, February 15, 2019 at 12:09:31 PM UTC-5, wrote:
On Thu, 14 Feb 2019 17:50:38 -0800 (PST), wrote:

On Thursday, February 14, 2019 at 2:17:31 PM UTC-5, wrote:
Finally last night I was able to sit down and figure out the probe
adjustment.
I have 5 probes. One 10x Tek probe, one 20x Tek probe, one
Probemaster 1x & 10x probe, and two other 1x & 10x probes from China.
The 10x Tek and the Probemaster probes are 100 MHz rated and made in
the USA. The Tek 20x probe is 200Mhz rated and USA made. The Chinese
probes are 100MHz rated.
I realized the probes need to be adjusted only in the 10x mode. I
didn't know that. That for sure was part of the problem.
Looking at signals from my function generator it seemed that the
voltage was off. Then I realized that my meter displays the RMS value
of the AC signal and the 'scope displays the peak values.
I got all the probes to display a nice square wave in the 10x
setting. I don't know why I got really weird results before but it may
be because I was in my machine shop with CNC machines running. At home
with everything quiet all the probes behaved like they were supposed
to. At the 10x setting.
But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.
It didn't seem to matter if the 'scope was set to AC or DC
coupling. It also didn't matter if the signal being measured came from
the 'scope or from the function generator.
Is this normal, for the square wave displayed to be perfect in the
10X setting and slightly distorted in the 1X setting?

The 'scope is a Tek465B.
The 20X probe is a TEK model P5120
The 10X probe is a Tek model P6053B
The Probmaster is model 3901-2
The Chinese probes only have a model number-P6100

So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?
Thanks,
Eric

Cool. Hey Eric, I just want to say that one reason I enjoy
sharing whatever small amount of knowledge I have with you,
is that you'll always follow up and tell me how it turned out.

I wish I had more customers like that. They send me
a problem... I send them off a few potential solutions,
and then ~1/2 the time I don't hear back. I assume something
worked... but I don't which of the three solutions it was,
or they made some bone head mistake and are embarrassed.

Oh, here's a machinist question for you. When tapping
holes in an easy metal, Al 6061 say, with a through hole
I've got these 'gun'* taps that shoot the chip out the
bottom of the hole. If I'm doing a blind hole is there
a tap that shoots the chip back up the tap.
(chip may not be the right word.)
I tap mostly small stuff
4-40, 6-32, 2-56... in approximate order of usage

George H.
*gun taps is what I call them, there is a GN on the
stem.
The taps that direct the chips back out of the hole are called high
spiral taps. Spiral point taps, AKA gun taps, eject the chips ahead
of the tap. I use high spiral taps a lot but I don't like them. They
are weaker than spiral point taps. I know, the terminology is
confusing, especially since the high spiral taps are actually high
helix taps.
For blind holes in ductile materials, such as wrought aluminum and
mild steel, I like to use roll form taps, AKA form taps, AKA roll
taps. In fact, I use roll taps whenever I can because of the stronger
thread and because no chips are generated. These taps require a larger
hole and hole size must be controlled well because the tap just
displaces the metal. With a roll tap you can tap right down to the
bottom of a hole because there are no chips generated.
Because form taps displace the material the thread is stronger than
a cut thread. Virtually all the fasteners you will find in your local
hardware store will have formed threads.
Materials that can be roll or form tapped include all the wrought
aluminum alloys, such as 6061, 6063, 5052, etc. Mild steel, 1144
steel (stressproof), 12L14 steel (leadloy) and 303 stainless steel
also tap well with form taps. 304 and 316 can be form tapped but great
care must be taken to avoid broken taps. Tap life will also be short
in 304 and 316. But if you need threads all the way to the bottom and
want to do it in one pass roll taps can work.
Materials that should not be formed tapped are the cast aluminum
alloys, half hard brass, cast iron, and the like. This is because the
material is not ductile and the resulting thread will have tiny
fractures, resulting in a weak thread.
I have tapped thousands of holes with 2-56 form taps.
Always use cutting oil when drilling and tapping holes, never motor
oil. Using motor oil leads to broken taps and bad finishes and poor
tool life.
Hope that helped.
Eric

Helped yes.. thanks. I don't do much tapping.. and all of that by hand
on the mill. (with starter drill in the chuck, centering the hand tap.)
I looked at those spiral taps and thought.. 'dang I'm just going to bust
on of those.' I never heard of form taps, but I'll look into them.

Thanks again,
George

On 2/15/2019 11:34 AM, wrote:
On Thu, 14 Feb 2019 16:30:11 -0800, Mike wrote:

On 2/14/2019 11:25 AM, wrote:

But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.

If you emailed your doctor and said, "I feel slightly bad,"
what diagnostic would you expect?

At least describe it in such a way that one could draw an
accurate picture from your description. And yes, the time
dimension is relevant.


What's the output impedance of your function generator?
What do you see when you plug the function generator directly
into the scope, via what kind of cable?
I forgot to write the frequency. The frequency being used was 1KHz
at 300 MV, from both the 'scope and the function generator.
I only have 1 cable that connects to the function generator, it has
a BNC connector at one end and alligator clips at the other end. So at
this time I cannot directly connect it to the 'scope, I must use the
probes. But the probes are what I was testing and I got the same
results no matter where the signal was coming from, as I stated in my
original post.
The output impedence of the function generator is 50 ohms.
If I was able to connect the function generator directly to the
'scope with the proper cable what would this tell me? I guess if the
square wave shape was still distorted the same way and the same amount
then it would be the 'scope that has the problem? And if the square
wave was perfect then it would be the probes causing the distortion.
Look, I try to post a question with adequate information. It seems
like I'm always leaving something out though. It's not because I'm
being lazy or trolling. It's because I miss ****, just like eveyone
else. And I don't mind being asked for more info, especially if the
reasons for the requested info are listed. This helps me to be more
accurate in my replies. And helps me learn, which is why I post
questions in the first place.
Thanks,
Eric

And yet, the best description you can give for the observed waveform
is "distorted". How can you expect a solution to a problem you won't
disclose?
At the risk of repeating myself...
At least describe it in such a way that one could draw an
accurate picture from your description. And yes, the time
dimension is relevant.

Mike wrote:



At the risk of repeating myself...
At least describe it in such a way that one could draw an
accurate picture from your description. And yes, the time
dimension is relevant.


** Ask "et" to describe a match box.

Bet he will struggle and make no sense.




..... Phil

SNIP

And yet, the best description you can give for the observed waveform
is "distorted". How can you expect a solution to a problem you won't
disclose?
At the risk of repeating myself...
At least describe it in such a way that one could draw an
accurate picture from your description. And yes, the time
dimension is relevant.
Whoops! It looks like the wave is over compensated. So it looks like
at the beginning of the flat waveform it has a little peak that then
drops down to form a flat line for the rest of the wave. This of
course repeats but in the negative direction for the negative part of
the wave. I thought you meant frequency for the time dimension but now
I think what you mean is how long does it take for the waveform to
become flat. It looks like it takes .1 mS. And the overshoot is about
30mV
Eric

wrote:



But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.

It didn't seem to matter if the 'scope was set to AC or DC
coupling. It also didn't matter if the signal being measured came from
the 'scope or from the function generator.



** Now that you have finally DESCRIBED the "distortion" ( a 10% overshoot) the answer is obvious.

The 465 scope is the culprit, the vertical attenuator needs frequency adjustment.

With the various probes in 10:1 setting, the trim cap allows you to compensate for the error in the scope.



..... Phil

On Fri, 15 Feb 2019 11:07:18 -0800, wrote:

On Thu, 14 Feb 2019 13:09:32 -0800, Jeff Liebermann
wrote:

On Thu, 14 Feb 2019 11:25:53 -0800, wrote:


So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?

It's caused by the lack of high frequency response (i.e. rise time) in
X1 mode. For a typical X10/X1 switchable problem, the frequency
response is MUCH higher in X10 than in X1. For example:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-250mhz-x1-x10
Notice that the bandwidth is 250MHz in X10 but only 6MHz in X1.
This one is closer to what I think you bought:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-60mhz-x1-x10
60MHz bandwidth in X10 and 6MHz bandwidth in X1.

The frequency I was using was 1KHz. That's what the 'scope puts out
and what the function generator was set at. With a 100MHz probe
shouldn't a 1 KHz signal be low enough to avoid the rise time issue?

No. The rise time and bandwidth are not related to the frequency of
the input waveform. Basically:
Bandwidth * RiseTime = 0.35
https://www.tek.com/support/faqs/how-bandwidth-related-rise-time-oscilloscopes
If the rise time of the 1KHz waveform is very small, you will need a
probe and scope with a high bandwidth in order to see it. If these
are lacking, the scope will display a much slower rise time than what
the 1KHz generator is actually producing.

For the two scope probes I mentioned above, both probes have 6MHz
bandwidth.
RiseTime = 0.35 / Bandwidth = 0.35 / 6*10^6 = 0.058 * 10^-6
= 58 * 10^-9 = 58 nsec rise time
with an X1 probe.

You can also mangle the waveform by setting the vertical amplifier set
to AC instead of DC. That will cause the top and bottom of the square
wave to sag like a capacitor charge/discharge curve.

I can't tell which of these is the problem because your description of
the waveform as "really weird results" and "slightly distorted" isn't
really sufficient.

I didn't even know about the difference in the frequency response.

Now, you know. There should be something on the web that covers all
this. Search for:
https://www.google.com/search?q=analog+oscilloscope+basics

EEVblog #453 - Mysteries of x1 Oscilloscope Probes Revealed
https://www.youtube.com/watch?v=OiAmER1OJh4 (29:48)

ABCs of Probes
https://download.tek.com/document/02_ABCs%20of%20Probes%20Primer.pdf
See Pg 8, 35-37.

Is
this one of the reasons that people recommend using the 10X setting
unless the 1x setting is needed?

Yep.

That the only time the 1X setting is
used is when the signal needs to be measured un-attenuated?

Nope. There are plenty of uses for an X1 probe. If the signal is
very low amplitude, it's easier to see using either a direct
connection to the scope input or an X1 problem. If the frequency is
low enough that you don't care about bandwidth, then X1 is just fine.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On Fri, 15 Feb 2019 14:56:25 -0800, Jeff Liebermann
wrote:

On Fri, 15 Feb 2019 11:07:18 -0800, wrote:

On Thu, 14 Feb 2019 13:09:32 -0800, Jeff Liebermann
wrote:

On Thu, 14 Feb 2019 11:25:53 -0800, wrote:


So, as near as I can tell, the probes are all right, at least they
all act the same. But what causes the distorted square wave at the 1x
setting?

It's caused by the lack of high frequency response (i.e. rise time) in
X1 mode. For a typical X10/X1 switchable problem, the frequency
response is MUCH higher in X10 than in X1. For example:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-250mhz-x1-x10
Notice that the bandwidth is 250MHz in X10 but only 6MHz in X1.
This one is closer to what I think you bought:
https://www.tiepie.com/en/usb-oscilloscope/accessories/oscilloscope-probe-60mhz-x1-x10
60MHz bandwidth in X10 and 6MHz bandwidth in X1.

The frequency I was using was 1KHz. That's what the 'scope puts out
and what the function generator was set at. With a 100MHz probe
shouldn't a 1 KHz signal be low enough to avoid the rise time issue?

No. The rise time and bandwidth are not related to the frequency of
the input waveform. Basically:
Bandwidth * RiseTime = 0.35
https://www.tek.com/support/faqs/how-bandwidth-related-rise-time-oscilloscopes
If the rise time of the 1KHz waveform is very small, you will need a
probe and scope with a high bandwidth in order to see it. If these
are lacking, the scope will display a much slower rise time than what
the 1KHz generator is actually producing.

For the two scope probes I mentioned above, both probes have 6MHz
bandwidth.
RiseTime = 0.35 / Bandwidth = 0.35 / 6*10^6 = 0.058 * 10^-6
= 58 * 10^-9 = 58 nsec rise time
with an X1 probe.

You can also mangle the waveform by setting the vertical amplifier set
to AC instead of DC. That will cause the top and bottom of the square
wave to sag like a capacitor charge/discharge curve.

I can't tell which of these is the problem because your description of
the waveform as "really weird results" and "slightly distorted" isn't
really sufficient.

I didn't even know about the difference in the frequency response.

Now, you know. There should be something on the web that covers all
this. Search for:
https://www.google.com/search?q=analog+oscilloscope+basics

EEVblog #453 - Mysteries of x1 Oscilloscope Probes Revealed
https://www.youtube.com/watch?v=OiAmER1OJh4 (29:48)

ABCs of Probes
https://download.tek.com/document/02_ABCs%20of%20Probes%20Primer.pdf
See Pg 8, 35-37.

Is
this one of the reasons that people recommend using the 10X setting
unless the 1x setting is needed?

Yep.

That the only time the 1X setting is
used is when the signal needs to be measured un-attenuated?

Nope. There are plenty of uses for an X1 probe. If the signal is
very low amplitude, it's easier to see using either a direct
connection to the scope input or an X1 problem. If the frequency is
low enough that you don't care about bandwidth, then X1 is just fine.
Thanks Jeff,
Below is an answer I gave to Mike about the wave shape and other
features. The description should give you a better idea of what I am
seeing. At least I hope it does.

"Whoops! It looks like the wave is over compensated. So it looks
like
at the beginning of the flat waveform it has a little peak that then
drops down to form a flat line for the rest of the wave. This of
course repeats but in the negative direction for the negative part of
the wave. I thought you meant frequency for the time dimension but now
I think what you mean is how long does it take for the waveform to
become flat. It looks like it takes .1 mS. And the overshoot is about
30mV
Eric"

Thanks Again,
Eric

Jeff Liebermann wrote:


You can also mangle the waveform by setting the vertical amplifier set
to AC instead of DC. That will cause the top and bottom of the square
wave to sag like a capacitor charge/discharge curve.



** No sign of that will be seen with a 1kHz square wave.

I can't tell which of these is the problem because your description of
the waveform as "really weird results" and "slightly distorted" isn't
really sufficient.


** That info was posted by the OP 3 or 4 hours ago.

You must read all the posts in a thread before posting replies with dumb error like that.

..... Phil

On Fri, 15 Feb 2019 15:33:01 -0800, wrote:

Below is an answer I gave to Mike about the wave shape and other
features. The description should give you a better idea of what I am
seeing. At least I hope it does.

"Whoops! It looks like the wave is over compensated. So it looks like
at the beginning of the flat waveform it has a little peak that then
drops down to form a flat line for the rest of the wave.

Under compensated or too little capacitance. The overshoot of the
rising part of the square wave means you have too much high frequency
response.

This of
course repeats but in the negative direction for the negative part of
the wave.

Yep. If the negative going part of the waveform were not identical,
then your input square is not symmetrical.

I thought you meant frequency for the time dimension but now
I think what you mean is how long does it take for the waveform to
become flat.

No. The flat part of the waveform is the low frequency component,
which is independent of the rise time. If the flat part sags, then
you are capacitor coupling (AC coupling) somewhere, probably the AC/DC
switch on the vertical input.

It looks like it takes .1 mS. And the overshoot is about
30mV

I can't tell if that means that it sags 0.1msec or the overshoot lasts
0.1msec. Perhaps a posting photo of the screen would be helpful.

More reading:
"The Secret World of Oscilloscope Probes"
http://www.dfad.com.au/links/THE%20SECRET%20WORLD%20OF%20PROBES%20OCt09.pdf

The link in the video to EEVblog doesn't work. This does:
"EEVblog #453 - Mysteries of x1 Oscilloscope Probes Revealed"
http://www.eevblog.com/forum/blog/eevblog-453-mysteries-of-x1-oscilloscope-probes-revealed/?all

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On 16/2/19 7:05 am, wrote:
I never heard of form taps, but I'll look into them.

They make roll-forming screws also. Google "tri-lobe screw". They're
much better and more secure than self-tappers, and you can remove and
replace them multiple times without weakening the thread.

Clifford Heath.

On 2/15/2019 4:15 PM, Phil Allison wrote:
wrote:



But with the probes set at 1x the square wave isn't quite square.
The 3 probes that have the 1x setting all act the same, with the
square wave being slightly distorted.

It didn't seem to matter if the 'scope was set to AC or DC
coupling. It also didn't matter if the signal being measured came from
the 'scope or from the function generator.



** Now that you have finally DESCRIBED the "distortion" ( a 10% overshoot) the answer is obvious.

The 465 scope is the culprit, the vertical attenuator needs frequency adjustment.

With the various probes in 10:1 setting, the trim cap allows you to compensate for the error in the scope.



.... Phil

Anyone wanting to store this list of TEKtronix probe characteristics
should do it now. The link is no longer functional, I went to the
Wayback Machine to locate it.
https://web.archive.org/web/20180104...age_probes.asp
It has info on about 100 different TEK probes.

Mikek

On Fri, 15 Feb 2019 19:02:44 -0800, Jeff Liebermann
wrote:

On Fri, 15 Feb 2019 15:33:01 -0800, wrote:

Below is an answer I gave to Mike about the wave shape and other
features. The description should give you a better idea of what I am
seeing. At least I hope it does.

"Whoops! It looks like the wave is over compensated. So it looks like
at the beginning of the flat waveform it has a little peak that then
drops down to form a flat line for the rest of the wave.

Under compensated or too little capacitance. The overshoot of the
rising part of the square wave means you have too much high frequency
response.

This of
course repeats but in the negative direction for the negative part of
the wave.

Yep. If the negative going part of the waveform were not identical,
then your input square is not symmetrical.

I thought you meant frequency for the time dimension but now
I think what you mean is how long does it take for the waveform to
become flat.

No. The flat part of the waveform is the low frequency component,
which is independent of the rise time. If the flat part sags, then
you are capacitor coupling (AC coupling) somewhere, probably the AC/DC
switch on the vertical input.

It looks like it takes .1 mS. And the overshoot is about
30mV

I can't tell if that means that it sags 0.1msec or the overshoot lasts
0.1msec. Perhaps a posting photo of the screen would be helpful.

More reading:
"The Secret World of Oscilloscope Probes"
http://www.dfad.com.au/links/THE%20SECRET%20WORLD%20OF%20PROBES%20OCt09.pdf

The link in the video to EEVblog doesn't work. This does:
"EEVblog #453 - Mysteries of x1 Oscilloscope Probes Revealed"
http://www.eevblog.com/forum/blog/eevblog-453-mysteries-of-x1-oscilloscope-probes-revealed/?all
Thanks again Jeff. I have watched EEVblog several times and haven't
had time yet to check out completely the episode mentioned. I looked
for it and just got it started when the dryer belt broke. Anyway, when
I spoke above about .1mS I meant that the overshoot lasts that long.
So there is a peak at the beginning of the square wave and then the
trace curves down to become flat. And the time it takes to become flat
is about .1 mS. The computer I am posting from is running XP and so I
can't use Dropbox. I should look for another easy to post to picture
site.
Eric

On Fri, 15 Feb 2019 19:02:44 -0800, Jeff Liebermann
wrote:

On Fri, 15 Feb 2019 15:33:01 -0800, wrote:

Below is an answer I gave to Mike about the wave shape and other
features. The description should give you a better idea of what I am
seeing. At least I hope it does.

"Whoops! It looks like the wave is over compensated. So it looks like
at the beginning of the flat waveform it has a little peak that then
drops down to form a flat line for the rest of the wave.

Under compensated or too little capacitance. The overshoot of the
rising part of the square wave means you have too much high frequency
response.

This of
course repeats but in the negative direction for the negative part of
the wave.

Yep. If the negative going part of the waveform were not identical,
then your input square is not symmetrical.

I thought you meant frequency for the time dimension but now
I think what you mean is how long does it take for the waveform to
become flat.

No. The flat part of the waveform is the low frequency component,
which is independent of the rise time. If the flat part sags, then
you are capacitor coupling (AC coupling) somewhere, probably the AC/DC
switch on the vertical input.

It looks like it takes .1 mS. And the overshoot is about
30mV

I can't tell if that means that it sags 0.1msec or the overshoot lasts
0.1msec. Perhaps a posting photo of the screen would be helpful.

More reading:
"The Secret World of Oscilloscope Probes"
http://www.dfad.com.au/links/THE%20SECRET%20WORLD%20OF%20PROBES%20OCt09.pdf

The link in the video to EEVblog doesn't work. This does:
"EEVblog #453 - Mysteries of x1 Oscilloscope Probes Revealed"
http://www.eevblog.com/forum/blog/eevblog-453-mysteries-of-x1-oscilloscope-probes-revealed/?all
Greetings Jeff,
Well, I watched that video you mentioned plus another two, one put out
by Tektronix. I now understand WAY better what's going on. Thanks for
the link and suggestions. I know it shouldn't matter but the EEVblog
guy's voice bugs me. But his explanation was clear. And the pdf you
posted a link to was also very educational.
Thanks,
Eric

On Tue, 26 Feb 2019 12:16:27 -0800, wrote:

On Fri, 15 Feb 2019 19:02:44 -0800, Jeff Liebermann
wrote:
More reading:
"The Secret World of Oscilloscope Probes"
http://www.dfad.com.au/links/THE%20SECRET%20WORLD%20OF%20PROBES%20OCt09.pdf

The link in the video to EEVblog doesn't work. This does:
"EEVblog #453 - Mysteries of x1 Oscilloscope Probes Revealed"
http://www.eevblog.com/forum/blog/eevblog-453-mysteries-of-x1-oscilloscope-probes-revealed/?all

Well, I watched that video you mentioned plus another two, one put out
by Tektronix. I now understand WAY better what's going on. Thanks for
the link and suggestions. I know it shouldn't matter but the EEVblog
guy's voice bugs me. But his explanation was clear. And the pdf you
posted a link to was also very educational.
Thanks,
Eric

Y'er welcome. There's also a mess of magic (and confusion) with high
voltage probes, where the frequency response varies with the applied
voltage. See graph on Pg 1-12:
https://download.tek.com/manual/070822305.pdf

Also active probes that work at GHz frequencies with input loading
around 1M and 1pf. You can really get spoiled with those working
around RF.
"Active probes: why they are worth buying"
https://www.eetimes.com/document.asp?doc_id=1279659#

I have the same problem with Dave Jones of EEVblog. Recently, he's
made an attempt to shorten his videos and clean up his presentation,
but it's still difficult viewing. Here's a clue:
https://www.youtube.com/user/EEVblog
"No Script, No Fear, All Opinions"
Methinks a script and an editor might help.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On 2/26/2019 1:29 PM, Jeff Liebermann wrote:
On Tue, 26 Feb 2019 12:16:27 -0800, wrote:

On Fri, 15 Feb 2019 19:02:44 -0800, Jeff Liebermann
wrote:
More reading:
"The Secret World of Oscilloscope Probes"
http://www.dfad.com.au/links/THE%20SECRET%20WORLD%20OF%20PROBES%20OCt09.pdf

The link in the video to EEVblog doesn't work. This does:
"EEVblog #453 - Mysteries of x1 Oscilloscope Probes Revealed"
http://www.eevblog.com/forum/blog/eevblog-453-mysteries-of-x1-oscilloscope-probes-revealed/?all

Well, I watched that video you mentioned plus another two, one put out
by Tektronix. I now understand WAY better what's going on. Thanks for
the link and suggestions. I know it shouldn't matter but the EEVblog
guy's voice bugs me. But his explanation was clear. And the pdf you
posted a link to was also very educational.
Thanks,
Eric

Y'er welcome. There's also a mess of magic (and confusion) with high
voltage probes, where the frequency response varies with the applied
voltage. See graph on Pg 1-12:
https://download.tek.com/manual/070822305.pdf

That has nothing to do with the frequency response.
Probes are VOLTAGE derated with frequency.
Think about it...
At infinite frequency, all the caps look like shorts and
it lets the smoke out.


Also active probes that work at GHz frequencies with input loading
around 1M and 1pf. You can really get spoiled with those working
around RF.
"Active probes: why they are worth buying"
https://www.eetimes.com/document.asp?doc_id=1279659#

I have the same problem with Dave Jones of EEVblog. Recently, he's
made an attempt to shorten his videos and clean up his presentation,
but it's still difficult viewing. Here's a clue:
https://www.youtube.com/user/EEVblog
"No Script, No Fear, All Opinions"
Methinks a script and an editor might help.

wrote:
I got all the probes to display a nice square wave in the 10x
setting.
But with the probes set at 1x the square wave isn't quite square.

You got lots of answers in this 1.5 year old thread, but the group
seems to have missed one explanation that might make sense to you.

Someone posted the bandwidth of two different probes in 1x mode, and
that bandwidth was 6Mhz.

A square wave can be represented as an infinite sum of sine waves, one
sine wave at the base frequency, and every odd harmonic of that base
sine wave up to infinity.

See: https://en.wikipedia.org/wiki/Square...urier_analysis

Your probes, when in 1x mode, are acting as low pass filters, with a
-3db point at the 6Mhz bandwidth of the probe. So for the input square
wave, consisting of an infinite sum of odd harmonics, the 1x probe is
filtering away most of the harmonics above 6Mhz.

When you low pass filter away most of the higher frequency harmonics
that sum to create a square wave, what you have left is a *distorted*
square wave.