|
After a bit more playing about, I think some of the fast and slow ripple is due to the limitations of the Vrms measurement that I have done with the scope. The accuracy of this is limited by the scope itself and also by the 5GSa/s sample rate. The scope can compute Vrms on a cyclic basis or on the display data and both methods have their accuracy limitations in this case. The impact of this is fairly minor though. The rest of the ripple will be a combination of ripple in the probe and some minor mismatch uncertainty in the test setup.
I think the reason I'm getting good results is because I'm using a scope with >1GHz bandwidth.
The input of a typical scope isn't 1Meg Rp in parallel with (say) 15pF Cp at RF frequencies. The 1Meg Rp only applies at very low frequencies. At RF frequencies the Rp will fall as the frequency increases. Cp will also change with frequency. It generally falls up at higher frequencies. Different scopes will show a different Cp and Rp vs frequency characteristic. This will make it harder to achieve consistent probe compensation up at UHF especially if you use an old and relatively slow scope like a Tek 2465A to test a 500MHz probe. I've not measured Rp and Cp for a 2465A but it isn't going to look the same as the Cp and Rp of a much faster scope that has lower Cp to begin with.
That's partly why I think it going to be better to look at a pulsed waveform using a decent pulse generator as I demonstrated earlier. My fastest homebrew pulser probably has a rise-time of about 400ps and it is adequate for testing a 500MHz scope probe. I've got another one that has a rise-time of about 2ns. I've got a faster BJT based pulser but it doesn't give a flat square wave output. The rise-time of the 2465A is going to be about 1ns and I think this scope would still be able to demonstrate how much better a genuine 10073C probe performs compared to (say) a Tek P6139A when using the pulse test method. By better, I mean how much flatter the pulse waveform looks after the initial rising edge has finished. The 10073C has the advantage that it is a 2.2Meg probe and this makes it easier to compensate across LF-UHF when compared to the 10Meg P6139A probe.?
|
The next thing I tried was to put a tiny 10R SMD resistor across the 50R termination. This brings the effective source impedance down to about 7 ohms. This should really reduce the droop at 500MHz. When I do this with the simulation, it predicts 1dB droop by 500MHz when the system is loaded by the 10073C probe. When I did it for real, with the real test system, it gave the result below with the 10073C probe. This also shows about 1dB droop. This is a really good result I think! It would be nice to do this with a better scope. My Tek scope doesn't have a great front end design. It's noisy and uses cheap SMD technology on a regular PCB for the front end. It does offer 1GHz BW but only above 5mV/div. Some of the ripple and noise in the plot below will be in my noisy old Tek scope. Some of the ripple will be in the 10073C probe, but at the moment it's hard to know (for certain) what causes the subtle ripple in the plot below. The simulation predicts a fairly smooth 1dB decline by 500MHz with none of the ripple seen in the plot below. 
|
I managed to write some code to control the sig gen and the 1GHz Tek scope using the classic levelling system described a few posts ago. I tested the 50R input of the scope first and it was flat to within about 0.2dB up to 500MHz. This doesn't prove very much because this scope has a different path for the 50R input, but it was still interesting to do it. I then swapped to the 1Meg input and compensated the 10073C probe using the usual procedure and then tested it for the frequency response when fed with a 50R source and a 50R termination. This is the Tektronix method. See below for the result. It looks quite similar to the simulation I did earlier. The response is 3.5dB down at 500MHz. I was expecting it to show a bit more droop than this, but it appears to show the correct response when the system is loaded by the probe. 
|
To muddy things a bit, the probe will have internal compensation that can aim to flatten the response and I think that this will make it harder to predict how much droop will actually be seen using this test method.
For lots of reasons I much prefer to look at how the probe performs in the time domain using a pulsed source. However, it might be interesting to see how the probes compare in the frequency domain as well.
I guess the thing to do is try it... I'll put some code together and try testing using this 50R | 50R method as used by Tektronix. I've got a 1GHz scope and a decent (dual 50R) splitter rated to many GHz and various sig gens that are suitable for this. I've also got a decent RF power meter to use as the detector in the feedback arm of the system.
Note that the scope probe loading can't pull down the source voltage at the input port of the splitter. This is because of the feedback in the system. This node will appear as an ultra low impedance voltage source. So the probe can't affect this voltage if the system can remain in closed loop. However, the probe loading can pull down the voltage across the termination resistor. My simulation showed it can pull it down by about 3.5dB by 500MHz.
|
All depends what is important for your application and the accuracy of measurements. Sopes are actually used for signal integrity tests, this is more complex than scalar voltage response. If you want to know if the voltage readings on scope represents the actual value on probe you need to solve the complete problem. If the probe will interfeer with the DUT performance , its important to know how much.. in this case ALSO the input impedance over frequency ?are needed.?
Ing. Patricio A. Greco
Taller Aeronáutico de Reparación 1B-349
Organización de Mantenimiento?Aeronáutico de la Defensa OMAD-001
Laboratorio de Calibración ISO 17025?AREA: RF/MW?
Gral. Martín Rodríguez 2159
San Miguel (1663)
Buenos Aires
T: +5411-4455-2557
F: +5411-4032-0072
toggle quoted message
Show quoted text
On 17 Apr 2024, at 11:35, jmr via < jmrhzu@...> wrote:
In this case the aim would be to fit a 50R termination as the DUT and then place the probe across the 50R DUT. This would effectively drive the probe with a well defined 25R source impedance.
Up at VHF and UHF the loading of the probe will therefore pull down the source voltage as you say. It's not a perfect system but I think some scope probes are formally measured (by Tektronix?) using a 50R source and a 50R termination like I proposed. I could reduce the series resistor in the DUT arm of the splitter or lower the termination resistance to effectively lower the source impedance and this would show a wider bandwidth.
The link below has already been posted by someone on this thread but it touches on this issue on page 7.
In essence the new Agilent bandwidth specification method assumes a zero ohm source impedance, a source unaffected by probe loading. A probe characterized by Agilent’s method would have an effective bandwidth that is lower than its calculated bandwidth because the probe is used in real world environment with real impedances that are not zero. The Tektronix bandwidth measurement method by comparison is optimized for an effective 25? source impedance (50? source in parallel with a 50? termination), which is much closer to real world conditions for measuring high speed signals in a terminated 50? environment
|
Using the 1 port model measured by my VNA I can try and predict the loading effect of the 10073C probe using this 50R || 50R method. See the plot below. If the probe (and the scope) was perfectly flat, it would show -3.5dB loading by 500MHz because of the probe. It does show a fairly smooth response in terms of the loading effect, but the real test would be to measure the output of the probe using a fast scope. I'd expect the response to show some more droop at 500MHz, partly because of any additional droop in the scope and also some extra droop in the response of the probe itself. It therefore might droop >5dB by 500MHz using my test method? 
|
Based on the info in the video below, Tektronix appear to aim for some excess margin in the claimed probe BW. Therefore, when a Tek 500MHz probe is connected to a Tek 500MHz scope, the probe should still show something approaching 500MHz bandwidth. See 3:10 onwards in the video to see what I mean.
|
In this case the aim would be to fit a 50R termination as the DUT and then place the probe across the 50R DUT. This would effectively drive the probe with a well defined 25R source impedance.
Up at VHF and UHF the loading of the probe will therefore pull down the source voltage as you say. It's not a perfect system but I think some scope probes are formally measured (by Tektronix?) using a 50R source and a 50R termination like I proposed. I could reduce the series resistor in the DUT arm of the splitter or lower the termination resistance to effectively lower the source impedance and this would show a wider bandwidth.
The link below has already been posted by someone on this thread but it touches on this issue on page 7.
In essence the new Agilent bandwidth specification method assumes a zero ohm source impedance, a source unaffected by probe loading. A probe characterized by Agilent’s method would have an effective bandwidth that is lower than its calculated bandwidth because the probe is used in real world environment with real impedances that are not zero. The Tektronix bandwidth measurement method by comparison is optimized for an effective 25? source impedance (50? source in parallel with a 50? termination), which is much closer to real world conditions for measuring high speed signals in a terminated 50? environment
|
In case of high impedance input you still have the problem of voltage divider between the 50Ω resistor of splitter and the probe input impedance. Its required to know the probe impedance to calculate the probe input voltage. More stable results would be get if you insert a pass-thru load. But again you will need to know the input impedance of probe and it includes the scope too.?
Ing. Patricio A. Greco
Taller Aeronáutico de Reparación 1B-349
Organización de Mantenimiento?Aeronáutico de la Defensa OMAD-001
Laboratorio de Calibración ISO 17025?AREA: RF/MW?
Gral. Martín Rodríguez 2159
San Miguel (1663)
Buenos Aires
T: +5411-4455-2557
F: +5411-4032-0072
toggle quoted message
Show quoted text
On 17 Apr 2024, at 10:39, jmr via < jmrhzu@...> wrote:
I tried looking at the 500MHz P6139A probe with the pulse test. As expected, it couldn't match the 10073A and 10073C probes in terms of the overall pulse response, but it did show a slightly faster rise-time.
I think that it's expecting too much for a 10M probe like the P6139A to have similar compensation performance as a 1M or 2.2M probe. So the result wasn't that surprising.
I tried to think of an alternative way to test the probe using a sig gen and I think it could be done using a self levelling system as pictured below. I normally use this system to test the frequency response of a 50R spectrum analyser (up to many GHz) or it can be used to test the response of a 50R scope input.
By replacing the device under test (DUT) with a 50R termination and then probing the termination with a fast scope, it should be possible to measure the frequency response of the probe.
The advantage of the system pictured below is that (when operating in closed loop) the source match is solely defined by the quality of the series 50R resistor inside the 11667 splitter that feeds to the DUT. So this means the source match feeding to the DUT will typically be < 1.03:1.
The level accuracy is defined by the accuracy of the thermocouple power meter. This is typically < +/- 0.1dB over 10-500MHz for example. I'll try and set something up to demo this system. It will need me to write some code to remotely operate the scope and the sig gen as this will let me store and plot the results for each probe type. This test method won't be as critical (of the probe) as the pulse test, but it might still be interesting to compare the results.?
<RF_Levelling.gif>
|
I tried looking at the 500MHz P6139A probe with the pulse test. As expected, it couldn't match the 10073A and 10073C probes in terms of the overall pulse response, but it did show a slightly faster rise-time. I think that it's expecting too much for a 10M probe like the P6139A to have similar compensation performance as a 1M or 2.2M probe. So the result wasn't that surprising. I tried to think of an alternative way to test the probe using a sig gen and I think it could be done using a self levelling system as pictured below. I normally use this system to test the frequency response of a 50R spectrum analyser (up to many GHz) or it can be used to test the response of a 50R scope input. By replacing the device under test (DUT) with a 50R termination and then probing the termination with a fast scope, it should be possible to measure the frequency response of the probe. The advantage of the system pictured below is that (when operating in closed loop) the source match is solely defined by the quality of the series 50R resistor inside the 11667 splitter that feeds to the DUT. So this means the source match feeding to the DUT will typically be < 1.03:1. The level accuracy is defined by the accuracy of the thermocouple power meter. This is typically < +/- 0.1dB over 10-500MHz for example. I'll try and set something up to demo this system. It will need me to write some code to remotely operate the scope and the sig gen as this will let me store and plot the results for each probe type. This test method won't be as critical (of the probe) as the pulse test, but it might still be interesting to compare the results.? 
|
Barry and Dave D, ?you are right. They may have the same beginnings but definitely they are different.
I apologise for the misleading information.
Pete
I was thinking the same thing but wanted to compare them first.
Also, I think the Petola sockets are gold-plated while the 131-4244-00 sockets are not.
DaveD ? That looks similar to a Peltola but I think it has slightly different dimensions.? Also, I the Peltola has only three "stilts" whereas these have four.
Barry - N4BUQ
This is the Peltola connector, named after its inventor.
Pete That's it! I've seen them on PCBs, but never as NOS.
DaveD Also 131-2766-03 Some on Ebay but $118.31 and $47.91 shipping ($Cdn). Others $9.63 but still $19.40 shipping. I don't buy from Ebay because of the shipping. If they were from China they would be $1 and the shipping would be free.
<IMG_1329.JPG> Tektronix made a small two-piece probe socket that mounted on a PCB. The two parts were a barrel into which the cylindrical probe ground barrel was inserted and a small closed barrel into which the probe tip is inserted, the two pieces to be soldered concentrically onto the PCB. The probe was inserted into the socket vertically.
I looked it up this morning; the part number is 131-4244-00. I was able to find some instances of it on tge web (e.g., at Mouser and Octopart). One could probably make a version of it.
I did not find it on the TekWiki site.
DaveD Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. <P6139A.jpg>
|
I was thinking the same thing but wanted to compare them first.
Also, I think the Petola sockets are gold-plated while the 131-4244-00 sockets are not.
DaveD
toggle quoted message
Show quoted text
On Apr 15, 2024, at 12:18, n4buq via groups.io <n4buq@...> wrote:
? That looks similar to a Peltola but I think it has slightly different dimensions.? Also, I the Peltola has only three "stilts" whereas these have four.
Barry - N4BUQ
This is the Peltola connector, named after its inventor.
Pete That's it! I've seen them on PCBs, but never as NOS.
DaveD Also 131-2766-03 Some on Ebay but $118.31 and $47.91 shipping ($Cdn). Others $9.63 but still $19.40 shipping. I don't buy from Ebay because of the shipping. If they were from China they would be $1 and the shipping would be free.
<IMG_1329.JPG> Tektronix made a small two-piece probe socket that mounted on a PCB. The two parts were a barrel into which the cylindrical probe ground barrel was inserted and a small closed barrel into which the probe tip is inserted, the two pieces to be soldered concentrically onto the PCB. The probe was inserted into the socket vertically.
I looked it up this morning; the part number is 131-4244-00. I was able to find some instances of it on tge web (e.g., at Mouser and Octopart). One could probably make a version of it.
I did not find it on the TekWiki site.
DaveD Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. <P6139A.jpg>
|
That looks similar to a Peltola but I think it has slightly different dimensions.? Also, I the Peltola has only three "stilts" whereas these have four.
Barry - N4BUQ
This is the Peltola connector, named after its inventor.
Pete That's it! I've seen them on PCBs, but never as NOS.
DaveD Also 131-2766-03 Some on Ebay but $118.31 and $47.91 shipping ($Cdn). Others $9.63 but still $19.40 shipping. I don't buy from Ebay because of the shipping. If they were from China they would be $1 and the shipping would be free.
<IMG_1329.JPG> Tektronix made a small two-piece probe socket that mounted on a PCB. The two parts were a barrel into which the cylindrical probe ground barrel was inserted and a small closed barrel into which the probe tip is inserted, the two pieces to be soldered concentrically onto the PCB. The probe was inserted into the socket vertically.
I looked it up this morning; the part number is 131-4244-00. I was able to find some instances of it on tge web (e.g., at Mouser and Octopart). One could probably make a version of it.
I did not find it on the TekWiki site.
DaveD Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. <P6139A.jpg>
|
This is the Peltola connector, named after its inventor.
Pete
That's it! I've seen them on PCBs, but never as NOS.
DaveD ? Also 131-2766-03 Some on Ebay but $118.31 and $47.91 shipping ($Cdn). Others $9.63 but still $19.40 shipping. I don't buy from Ebay because of the shipping. If they were from China they would be $1 and the shipping would be free.
<IMG_1329.JPG> Tektronix made a small two-piece probe socket that mounted on a PCB. The two parts were a barrel into which the cylindrical probe ground barrel was inserted and a small closed barrel into which the probe tip is inserted, the two pieces to be soldered concentrically onto the PCB. The probe was inserted into the socket vertically.
I looked it up this morning; the part number is 131-4244-00. I was able to find some instances of it on tge web (e.g., at Mouser and Octopart). One could probably make a version of it.
I did not find it on the TekWiki site.
DaveD ?Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. <P6139A.jpg>
|
That's it! I've seen them on PCBs, but never as NOS.
DaveD
toggle quoted message
Show quoted text
On Apr 15, 2024, at 11:44, peter bunge via groups.io <bunge.pjp@...> wrote:
? Also 131-2766-03 Some on Ebay but $118.31 and $47.91 shipping ($Cdn). Others $9.63 but still $19.40 shipping. I don't buy from Ebay because of the shipping. If they were from China they would be $1 and the shipping would be free.
<IMG_1329.JPG> Tektronix made a small two-piece probe socket that mounted on a PCB. The two parts were a barrel into which the cylindrical probe ground barrel was inserted and a small closed barrel into which the probe tip is inserted, the two pieces to be soldered concentrically onto the PCB. The probe was inserted into the socket vertically.
I looked it up this morning; the part number is 131-4244-00. I was able to find some instances of it on tge web (e.g., at Mouser and Octopart). One could probably make a version of it.
I did not find it on the TekWiki site.
DaveD ?Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. <P6139A.jpg>
|
Also 131-2766-03 Some on Ebay but $118.31 and $47.91 shipping ($Cdn). Others $9.63 but still $19.40 shipping. I don't buy from Ebay because of the shipping. If they were from China they would be $1 and the shipping would be free.
Tektronix made a small two-piece probe socket that mounted on a PCB. The two parts were a barrel into which the cylindrical probe ground barrel was inserted and a small closed barrel into which the probe tip is inserted, the two pieces to be soldered concentrically onto the PCB. The probe was inserted into the socket vertically.
I looked it up this morning; the part number is 131-4244-00. I was able to find some instances of it on tge web (e.g., at Mouser and Octopart). One could probably make a version of it.
I did not find it on the TekWiki site.
DaveD ?Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. <P6139A.jpg>
|
Tektronix made a small two-piece probe socket that mounted on a PCB. The two parts were a barrel into which the cylindrical probe ground barrel was inserted and a small closed barrel into which the probe tip is inserted, the two pieces to be soldered concentrically onto the PCB. The probe was inserted into the socket vertically.
I looked it up this morning; the part number is 131-4244-00. I was able to find some instances of it on tge web (e.g., at Mouser and Octopart). One could probably make a version of it.
I did not find it on the TekWiki site.
DaveD
toggle quoted message
Show quoted text
On Apr 15, 2024, at 07:45, jmr via groups.io <jmrhzu@...> wrote:
?Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. <P6139A.jpg>
|
Yes, that's the same grounding technique I used when I did the scope rise-time experiments. The coil of wire needs to have a degree of 'spring' to it to allow a reliable ground connection. Some scope probes come with one of these spring coils for this very purpose. It really does give good results. I managed to find the accessory pouch that came with my Tek scope and the P6139A probe was still in it. I've never used it. I've got a lot of scope probes here already. You can see how tiny the tip section is on this probe when compared to the 10073C tip that I've shown alongside it. They are both 500MHz probes but the 10073C is a 2.2Meg probe and the P6139A is a 10Meg probe. 
|
Great idea, Mikek. Especially since I can't seem to find any probe sleeves at all on Ebay.
|
This idea of letting it on the board is such a simple but great one! I'll remember that Jim!
Renaud
|
Patricio A. Greco
Pete_G4GJL
Jinxie