Probe into probes. What's up?

[BillyO]

I recently received an inexpensive ($36) no-name 500MHz probe I got of Amazon (called a P-6139A of all things)and I thought I'd compare it to some other probes I have arornd here including my genuine Tek 6139A probes.

First off, the probe seem very well made.  The cable is very flexible (even more so than the Tek probes), the compensation/BNC housing is metal and it has the 10:1 auto-select pin.  The BNC fist my Siglent scopes and Tek very snugly and the accessories all seem to work very well.  Construction is modeled after the Tek probe with the insulating sleeve going all the way to the probe tip and unscrewing in a similar way to the Tek.  The tip does not unscrew from teh cables assembly though.  It's rated a 300V CAT II and 9pf @ 10Mohm.

In any case, using the Bodnar pulser and my own 1.2ns pulser I did some rise time trials with it, the Tek 6139A and a Hantek PP300 (300MHz probe).  The results have me a bit puzzled.

The first two screen shots are of my and the Bodnar pulsers feeding directly into the "improved" SDS2104XP scope.  In both cases we can see a fairly nice and consistent square wave.

Next 2 are the Hantek PP300.  This is about what I'd expect for this $20 probe.  Actually pretty nice for the price.

Nest 2 are the Tek 6139A.  WTF?  while the initial part of the leading edge is sharp and fast, something goes awfully wrong about 65% of the way up.  Does anyone have idea what this is about?  I get the exact same result from the other two Tek 6139A probes I have here.

The next two are no name P-6139A.  The result with the Bdonar pulser is almost identical to the Tek probe!  Again, what's going on?  The result with my pulser is also weird but could be due to soem interaction between the pulser and the probe.  It still starts out okay, but seems to give up steam before the Tek probe does.

I honestly expected a that at least the Tek probe woudl do better than this and I can't explain why the faster probes behave this way.  Any ideas?

[BillyO]

The rest of the screen shots...

[bdunham7]

Any ideas?

Group delay...have a close look at the Tek 6139A datasheet.

Edit:  are your screenshots labelled correctly? 

[jonpaul]

See

https://www.tek.com/en/documents/whitepaper/abcs-probes-primer

Also Tektronix Circuits Concepts book Oscilloscope Probe Circuits
https://w140.com/tekwiki/images/6/62/062-1146-00.pdf
Oscilloscope Probe Measurements
https://w140.com/tekwiki/images/1/19/062-1120-00.pdf

Any scopes avail as second test besides the Chinese?

Any photo of test setup? Block/schema?

Your Probes uses bayonet/BNC adapter or just ordinary ground cable/alligator?

The self inductance of the ground return can affect transient resp, ringing, etc.

We found P6139 rather delicate and use mostly P6137 x10 and P6136 Zo probes on our TEK 2465/P2467B

These probes are fine on Yokogawa digital scopes DL7440, 1740.

Jon

[BillyO]

Edit:  are your screenshots labelled correctly? 

Yes.  Why?

[tautech]

Nest 2 are the Tek 6139A.  WTF?  while the initial part of the leading edge is sharp and fast, something goes awfully wrong about 65% of the way up.  Does anyone have idea what this is about?  I get the exact same result from the other two Tek 6139A probes I have here.

I honestly expected a that at least the Tek probe would do better than this and I can't explain why the faster probes behave this way.  Any ideas?

Do Tek 6139A probes have a hidden (not obvious) HF adjustment ?
The manual doesn't mention such. 
https://download.tek.com/manual/063087005.pdf

All the Asian 500 MHz probes I've seen had LF and HF trimmers.

[BillyO]

Any scopes avail as second test besides the Chinese?

  Unfortunately not, except my old Tek 465.

Any photo of test setup? Block/schema?

A photo can be supplied .. but see next answer.

Your Probes uses bayonet/BNC adapter or just ordinary ground cable/alligator?

Pulser->BNC to probe adapter -> probe.

The self inductance of the ground return can affect transient resp, ringing, etc.

No ground wire used.  Connected through probe adapter.

[BillyO]

Do Tek 6139A probes have a hidden (not obvious) HF adjustment ?

  They do.  Two of them.  But how likely is it that the adjustment it is exactly wrong the same way on 3 Tek probes and one Asian probe?

[tautech]

No ground wire used.  Connected through probe adapter.

Something's not right.
Some waveforms 0V is not on the 0V channel level like they're in AC input mode.

[BillyO]

No ground wire used.  Connected through probe adapter.

Something's not right.
Some waveforms 0V is not on the 0V channel level like they're in AC input mode.

The Bodnar outputs an AC signal.

[bdunham7]

Yes.  Why?

The first screenshot in your second post is labelled "Tek 6139 Bodnar" but the scope is set to DC50 input....

[tautech]

No ground wire used.  Connected through probe adapter.

Something's not right.
Some waveforms 0V is not on the 0V channel level like they're in AC input mode.

The Bodnar outputs an AC signal.

Okay yeah, missed one waveform was 10 MHz and the other 20 MHz.

Pop the lid of the compensation box and have a squiz at what's in there.
You might need to download the old book on Tek probes to follow the schematic of how these were constructed.
https://w140.com/tekwiki/images/6/62/062-1146-00.pdf

[BillyO]

Group delay...have a close look at the Tek 6139A datasheet.

I can't seem to find this.  It no longer seems to be on the Tek site and nothing else I have found makes mentions the group velocity characteristics.

I guess you are suggesting that the different frequency components of teh sharp square waves will have different velocities in the probe system?

[BillyO]

The first screenshot in your second post is labelled "Tek 6139 Bodnar" but the scope is set to DC50 input....

Not that I can see.  It says DC1M.

[bdunham7]

I can't seem to find this.  It no longer seems to be on the Tek site and nothing else I have found makes mentions the group velocity characteristics.

I guess you are suggesting that the different frequency components of teh sharp square waves will have different velocities in the probe system?

Yes, exactly.

https://w140.com/tekwiki/images/a/af/063-0870-05.pdf

[David Hess]

Because of the lumped input capacitance and response of the high impedance input, the high frequency transient response of the probe itself is used to correct the response of the oscilloscope.  Tektronix did not include external adjustments for this, and instead released a different A/B/C version of the probe as newer oscilloscopes were released.

More general purpose probes, like from Texas who also OEMs probes for Rigol and Keysight, include access to the high frequency transient response adjustments.

[BillyO]

It's trivial to get into the HF compensation adjustment on the P6139A.  There is also another device in there.  Not 100% sure what it is.

[bdunham7]

Not that I can see.  It says DC1M.

Is the forum mixnig photos again?

[BillyO]

Must be.  That's not even the Bodnar.  It's my pulser and it's straight into the scope.

[bomp]

Does the  Bodnar pulser need a 50 ohm load when the 10:1 probes are used?

[BillyO]

Does the  Bodnar pulser need a 50 ohm load when the 10:1 probes are used?

Shouldn't, as it's not driving a 50 ohm transmission line.

[BillyO]

Here are soem more images.

The No-name P6139A

Two pictures of the Tek probe disassembled.  In the picture where I have the metal case apart you can see the two HF resistors at the bottom.  The LF compensation is near the center and There is that blue object just above it and to the right.  It's marked C2, but there is not way to get anything into it to adjust it.  It may need a special tool.

The last picture is the tesr setup with the Bodnar.  It's exactly the same with my pulser too as it's just about the same size.

[srb1954]

Does the  Bodnar pulser need a 50 ohm load when the 10:1 probes are used?

Shouldn't, as it's not driving a 50 ohm transmission line.

I don't know what terminating load the Bodnar pulser requires but the conditions for verifying probe performance traditionally calls for a 25 source (ref. Tektronix Circuit Concept series, Oscilloscope Probe Circuits, p27). This source impedance is typically made up of the pulse generator source impedance of 50 paralleled by a 50 through termination. So even if you are not using a interposing cable the output of the pulse generator still requires a terminator.

[tautech]

Does the  Bodnar pulser need a 50 ohm load when the 10:1 probes are used?

Shouldn't, as it's not driving a 50 ohm transmission line.

I don't know what terminating load the Bodnar pulser requires but the conditions for verifying probe performance traditionally calls for a 25 source (ref. Tektronix Circuit Concept series, Oscilloscope Probe Circuits, p27). This source impedance is typically made up of the pulse generator source impedance of 50 paralleled by a 50 through termination. So even if you are not using a interposing cable the output of the pulse generator still requires a terminator.

100%
If it weren’t for a darn power cut I’d put some screenshots up showing exactly this. 

[BillyO]

I could certainly try that.

Edit:  Well!  That just made one heck of a difference!  This is more like what I expected.  With the Tek probe being the best the Hantek the worst and the no-name 6139A being in between.

So that Hantek 300MHz probe is really a 300MHz probe.  The no-name P6129A looks like it would come in around 400MHz and the Tek's rise time is exactly what I calculated it should be @ 500MHz.  Nice.

Shot in the same order as before:  Hantek, Tek, No-name.