This just crossed my feed today. I was fortunate that my scope did come with probes. As someone who only occasionally uses a scope, thanks for pointing it out. As straightforward as it sounds now, I had no idea. Now I need to check the probes I have when I get home. 😅
Thank you Alan, for this video and for all the work you've been doing over the past years. For many years I've been using my Tek465 purchased it from eBay, repaired it, calibrated it (to the best of my ability) with some cheap oscilloscope probes ($40 cheap). I mostly play with audio - old cassette tape recorders, low noise headphone amplifiers, some HAM. In one of your previous videos you asked for help with your under cabinets lighting, which like you and many others, I have the same issue. My bench lamp with magnifier 5X loupe uses a 22cm diameter LED powered by an switching power supply (planning to replace it with an transformer, good DC filter power supply soon, as others suggested). When I calibrate the microphone inputs for a cassette tape recorder I'm using very low voltage signals 0.77mV. My Tek can barely see those since its lowest input is 5mV/DIV (planning use an OP07 low noise amplifier to see those signals). What I noticed is that my cheap scope probes pickup noise (around the coax cable) and like you I have to turn off lights and use bulbs when I use very low signals like that. So, question to you - what's the best coaxial cable I can use to replace the existing coaxial cable in these cheapo probes -- can I use RG316, its impedance is 50ohm. Or should I use RG-179, although that's a 75ohm (I kind of answered my own question there) or do you know a better coax cable I can use.
What a great informative video. I just purchased an Owon hand held scope and had no idea that I had to somewhat match the probe to the scope. I thought all probes were the same. A 10 x probe came with my scope but I purchased an extra one to go with CH 2. Now I have to go back and see what I purchased? Since my scope is only good up to 40 mhz, I stand a good chance that I may be OK. Thanks for posting this great video. Barry, KU3X
This formula holds true when the probe and the scope both have a single-pole rolloff with frequency. Used to be more common with analog scopes, not as much with today's modern DSOs. Still a good recommendation for the probe BW to exceed the scope BW, provided it can still be properly compensated.
I’m starting to watch lots of your videos and they’re very helpful to me as I’m getting back into electronics a little bit after 40 years of doing other things. One thing I was a little bit confused about is what happens if I have a probe with a compensation range starting at 15 pico farad but my o- scope is rated at 15 pico farad. You touched on it in the video, but it wasn’t clear to me whether or not the probe would still be usable in this situation. I need new probe for my tekTronics 2430a scope and I’m looking at a 1X 10 X probe P 2200 but like the one you show in the video, it starts at 15 pico farad so I’m not sure I should use it. It’s getting harder to find the original P 6133 probes. My scope came with when it was new. I really do want a 1X probe as well.
It is likely that the probe would be able to be adjusted properly for that scope. If the probe's rated compensation range includes 15pF, then it should work fine with a scope that has 15pF input capacitance.
Passives probes don't have a real passband they just have capacitive loading. The higher frequency probes are made with smaller wire and sometime shorter length and as you mention a different compensation range capacity.
Oscilloscope question: Often my measurements will vary depending on how securly the probe is attached, or the relative position of the probe cable. I'm never sure if this means my probe is bad, or if it's due to the shared ground connection, or maybe something else?
This is usually a combination of the probe ground affecting the measurement (ground inductance in the measurement path) as well as the probe/scope ground affecting actual circuit performance.
@@w2aew Since my supply and scope both share the same mains earth, I don't usually use the probe ground lead. But I have noticed the same varying measurements even when using an independent supply and connecting the probe ground. A bit of tension or change of position still affects the reading. I have several probes and I do calibrate them with the scope's built in calibrator (as per your other video).
@@andrewferg8737 The ground is the return path for the sensed/probed signal. By not using the ground, you add a LOT of inductance into the measurement return path. At low frequencies (audio) it may not matter much, but will definitely have an effect as frequency goes up.
Thanks for great video as always. One question that has always been bothering me is that even with P6139 which is one of the best and lowest capacitance probes in the market after all these years (8pF) still the reactance at 500MHz is just 40 ohm. I cannot understand how the system bandwidth could be 500MHz when the reactance drops to such low value at 500MHz? It's even worse for almost all other 500MHz probes from HP or Lecroy because they have 10-12pF capacitance at the tip.
True, the overall measurement BW will depend on the source impedance of the node being probed, and how it reacts to the 8pF load (will that drag it down). The probe/system BW is just a measure of the frequency response with respect to the voltage at the probe tip. BTW - the latest generation of Tek scopes support a new wide-BW passive probe, the TPP1000, which is rated at 1GHz with just 3.9pF loading. The widest BW 10x passive probe in the industry.
@@w2aew that makes sense but I have not seen manufacturers mention what source impedance they consider when they define the system bandwidth for their probes. So, for example for P6139, how exactly the system BW is defined? Is it for example equal to 500MHz if I probe a 50 ohm node? Yes, I know about TPP probes and I have two of them but they only work with newer Tek scopes. P6139 is kind of universal
Probes are tested and specified using a 25 ohm source impedance from a double terminated 50 ohm source. The same applies to the oscilloscope input, however the probe termination may lower the impedance from the probe which allows it to more easily drive the oscilloscope's input capacitance. 25 ohms is unrealistic in most applications, but at least it is objective.
the cheapest and most easiest to find scope probe are p6100 chinese clones. can you do some testing, using three or four different "real" oscilloscopes, and a DSO138 and/or similar "noob dso scope"? :D also, if you have an "real" p6100, HP thing, i think, can you compare them? :)
Don't forget the needs! If all one want to do is to look at audio signals several hundreds of MHz isn't needed as an example. Also getting a probe that can't reach the frequencies needed is wasted money. What voltages? Is 1x, 10x, 100x or... the right choice? Get a cheap to learn with or are verifiable measurements needed?
For very low frequency work, a 1x probe is often suitable provided the circuit can tolerate the ~100pF of capacitive loading. The vast majority of applications will utilize the venerable 10x probes. 100x probes are typically only needed for high voltage work (hundreds of volts). Even if you don't need hundreds of MHz of measurement BW with your probe - keep in mind that very low-BW probes might not be able to be compensated on your higher BW scope. This can lead to amplitude errors on signals at just several kHz frequency range.
Is it fair to say that probe compensation sets a "capacitance divider ratio (series C in the probe w.r.t. shunt C as specified in the probe) to such a value as to get the right amplitude? Is not the overall bandwidth governed by the input C and series resistance of the resistor divider and 50 ohm on the scope input? I think some equivalent circuits would be very illuminating here. Thanks for the video!
Ahh Rin on the scope is typicall 1 MegOhm unless set to 50 ohm.. of course... Also the C compensation is in shunt with the input impedance but that doesn't really explain what it does.
Basically, yes, the compensation adjustment is setting up the appropriate capacitive voltage divider between shunt cap around the 9M series resistor and the scope input capacitance. There is more to it (and more to what determines the BW), but this is the proper high-level description of compensation.
The magic of probe compensation is as follows: the scope has an input impedance of 1Meg//Cin, the probe has a 9 Meg resistor paralled with Ccomp (to be adjusted). When the two time constants 1Meg*Cin and 9 Meg*Ccomp are made equal, then the set "probe & scope input" behaves like a plain voltage dividor 1:10 with a flat bandwith. if your scope input is 50 ohms, then there is no use in using a probe (unless it's an active probe): a 50 ohms coax is enough.
Great video. You have an excellent way of explaining how these things work! There is an issue I'm wondering about. It is my understanding (limited as it is) that the total bandwidth of the scope - probe combination is less than the bandwidth of either the scope or the probe. (More specifically, it's the vector of the rise time of the scope and probe. For the formula, see en.wikipedia.org/wiki/Test_probe). If so, to maximize the benefit of the scope bandwidth, is it fair to say that you want a probe which has a higher bandwidth (say 1.5x to 3.0x the bandwidth) of the scope as long as the probe can be compensated to match the scope?
In most cases, yes. Tektronix is a bit of an exception - they typically underrate the BW of the probe so that if you put a 100MHz probe on a 100MHz scope, you get 100MHz of BW (not 70MHz). Most manufacturers don't do that.
Excellent information as always. I've been using oscilloscopes off and on since the mid 1960s but I still learned a few new things. Thanks and 73 de G3WGV.
Alan thank you for another outstanding video. Just went and checked the capacitance input on my Siglent 200 MHz scope to all the probs (a bunch collected over the years). Found a couple that didn’t match so they were marked as such. Thanks again for the informative video.
I want to do a full alignment on an FT-857D. I have the equipment necessary to do the job but the one piece of information I am missing is what is the best way to probe the radios 7mhz sections. Really probing RF in general. Active probes are prohibitively expensive so I'm looking at probes on ebay with between 1-3pf or just using a piece of coax and a resistor. Would you mind giving me some hints on the best way to go about this? Thanks.
as usual, the circuit is not a (big) problem, but mechanical part will be a nightmare! :) i seen a lot of diy probes, some of them really nice, and those ones was crafted in copper clad board boxes. also, just some random plastic housings, and they was more or less usable! :D
Alan, you have shown us all around your bench, lab etc. How about a video showing us how you shoot the videos ? Camera, camera usage, video equipment used, how you manage to do the videos etc. Just an idea, "73"....
@@w2aew The reason I suggest this, you do an excellent job of making clear precise videos. Even though your moving and changing angles etc. It appears seamless...
Another very informative video. I bought a Tek 2445B 150MHZ scope from a dealer in the US in 2016. It came with 2 probes, P6121 and P6122. I recently came to know after watching this video that the probes that came with the scope was 100 MHZ not 150 MHZ.
So how do we measure, characterize, and compensate for mismatched components, and what happens if you don't? For example when you can't get proper specs for your test gear. I have much research to do. :)
They need to match. If you have the probe set to 10x (to give you the highest BW and lowest capacitive loading), then you should change the scope's vertical scale setting to 10x so that the vertical scale reads correctly.
In my experience probes that are switchable 1X/10X (or 10X/100X) have no way to tell the scope what they are set to. So for that type of probe, you have to set BOTH the probe and scope to the same value. If you can't do that, then you will have to remember to divide the voltages shown on the screen or display by 10 or 100, depending on the ratio of the probe and scope settings.
This just crossed my feed today. I was fortunate that my scope did come with probes. As someone who only occasionally uses a scope, thanks for pointing it out. As straightforward as it sounds now, I had no idea. Now I need to check the probes I have when I get home. 😅
Thank you very much for your videos. I learn a lot from them. Thank you for taking time and making the videos.
Boop... what a pleasure to watch these episodes... Cheers!
Thanks, Alan. There's always something to learn from your videos.
Your videos are pure gems as far of information content is concerned. Many thanks.
Thanks! And it's nice to see the 485 powered up and shiny. :)
Thank you Alan, for this video and for all the work you've been doing over the past years. For many years I've been using my Tek465 purchased it from eBay, repaired it, calibrated it (to the best of my ability) with some cheap oscilloscope probes ($40 cheap). I mostly play with audio - old cassette tape recorders, low noise headphone amplifiers, some HAM. In one of your previous videos you asked for help with your under cabinets lighting, which like you and many others, I have the same issue. My bench lamp with magnifier 5X loupe uses a 22cm diameter LED powered by an switching power supply (planning to replace it with an transformer, good DC filter power supply soon, as others suggested). When I calibrate the microphone inputs for a cassette tape recorder I'm using very low voltage signals 0.77mV. My Tek can barely see those since its lowest input is 5mV/DIV (planning use an OP07 low noise amplifier to see those signals). What I noticed is that my cheap scope probes pickup noise (around the coax cable) and like you I have to turn off lights and use bulbs when I use very low signals like that. So, question to you - what's the best coaxial cable I can use to replace the existing coaxial cable in these cheapo probes -- can I use RG316, its impedance is 50ohm. Or should I use RG-179, although that's a 75ohm (I kind of answered my own question there) or do you know a better coax cable I can use.
Fantastically clear and informative!
Thank you! Succinct, clear, educational, and useful.
Great video. Thanks Alan
Always enjoy your great tips Alan! 73 - Dino KLØS
ALWAYS GOOD, USEFUL INFORMATION FOR ALL.
THANK YOU!
Informative as always Alan. Thanks.
Thanks Alan. Good info, as always!
What a great informative video. I just purchased an Owon hand held scope and had no idea that I had to somewhat match the probe to the scope. I thought all probes were the same. A 10 x probe came with my scope but I purchased an extra one to go with CH 2. Now I have to go back and see what I purchased? Since my scope is only good up to 40 mhz, I stand a good chance that I may be OK.
Thanks for posting this great video.
Barry, KU3X
You learn something new everyday!
Great useful stuff. Thx
Super-great. Thank you.
Thanks Alan! excellent as usual! cheers from Italy!
Final_BW = 1 / SQRT ( SQR ( 1 / Probe_BW) + SQR ( 1 / Scope_BW) ).
They recommend Probe BW to be in between 1.5X and 2.5X the Scopw BW.
This formula holds true when the probe and the scope both have a single-pole rolloff with frequency. Used to be more common with analog scopes, not as much with today's modern DSOs. Still a good recommendation for the probe BW to exceed the scope BW, provided it can still be properly compensated.
I’m starting to watch lots of your videos and they’re very helpful to me as I’m getting back into electronics a little bit after 40 years of doing other things. One thing I was a little bit confused about is what happens if I have a probe with a compensation range starting at 15 pico farad but my o- scope is rated at 15 pico farad. You touched on it in the video, but it wasn’t clear to me whether or not the probe would still be usable in this situation. I need new probe for my tekTronics 2430a scope and I’m looking at a 1X 10 X probe P 2200 but like the one you show in the video, it starts at 15 pico farad so I’m not sure I should use it. It’s getting harder to find the original P 6133 probes. My scope came with when it was new. I really do want a 1X probe as well.
It is likely that the probe would be able to be adjusted properly for that scope. If the probe's rated compensation range includes 15pF, then it should work fine with a scope that has 15pF input capacitance.
@@w2aew Thanks for responding. I’ll give it a go.
Will you make video about mechanical oscillators
uh oh. I already filmed a video on scope probes, so don't think I'm copying you if you see it in a few weeks.
No worries - I'm certain your video will be very good!
Passives probes don't have a real passband they just have capacitive loading. The higher frequency probes are made with smaller wire and sometime shorter length and as you mention a different compensation range capacity.
Oscilloscope question: Often my measurements will vary depending on how securly the probe is attached, or the relative position of the probe cable. I'm never sure if this means my probe is bad, or if it's due to the shared ground connection, or maybe something else?
This is usually a combination of the probe ground affecting the measurement (ground inductance in the measurement path) as well as the probe/scope ground affecting actual circuit performance.
@@w2aew Since my supply and scope both share the same mains earth, I don't usually use the probe ground lead. But I have noticed the same varying measurements even when using an independent supply and connecting the probe ground. A bit of tension or change of position still affects the reading. I have several probes and I do calibrate them with the scope's built in calibrator (as per your other video).
@@andrewferg8737 The ground is the return path for the sensed/probed signal. By not using the ground, you add a LOT of inductance into the measurement return path. At low frequencies (audio) it may not matter much, but will definitely have an effect as frequency goes up.
Great video 👍🏻
Needed this info, coincidentally! Thanks.
Thanks for great video as always. One question that has always been bothering me is that even with P6139 which is one of the best and lowest capacitance probes in the market after all these years (8pF) still the reactance at 500MHz is just 40 ohm. I cannot understand how the system bandwidth could be 500MHz when the reactance drops to such low value at 500MHz? It's even worse for almost all other 500MHz probes from HP or Lecroy because they have 10-12pF capacitance at the tip.
True, the overall measurement BW will depend on the source impedance of the node being probed, and how it reacts to the 8pF load (will that drag it down). The probe/system BW is just a measure of the frequency response with respect to the voltage at the probe tip. BTW - the latest generation of Tek scopes support a new wide-BW passive probe, the TPP1000, which is rated at 1GHz with just 3.9pF loading. The widest BW 10x passive probe in the industry.
@@w2aew that makes sense but I have not seen manufacturers mention what source impedance they consider when they define the system bandwidth for their probes. So, for example for P6139, how exactly the system BW is defined? Is it for example equal to 500MHz if I probe a 50 ohm node?
Yes, I know about TPP probes and I have two of them but they only work with newer Tek scopes. P6139 is kind of universal
Probes are tested and specified using a 25 ohm source impedance from a double terminated 50 ohm source. The same applies to the oscilloscope input, however the probe termination may lower the impedance from the probe which allows it to more easily drive the oscilloscope's input capacitance. 25 ohms is unrealistic in most applications, but at least it is objective.
the cheapest and most easiest to find scope probe are p6100 chinese clones.
can you do some testing, using three or four different "real" oscilloscopes, and a DSO138 and/or similar "noob dso scope"? :D
also, if you have an "real" p6100, HP thing, i think, can you compare them? :)
Don't forget the needs!
If all one want to do is to look at audio signals several hundreds of MHz isn't needed as an example.
Also getting a probe that can't reach the frequencies needed is wasted money.
What voltages? Is 1x, 10x, 100x or... the right choice?
Get a cheap to learn with or are verifiable measurements needed?
For very low frequency work, a 1x probe is often suitable provided the circuit can tolerate the ~100pF of capacitive loading. The vast majority of applications will utilize the venerable 10x probes. 100x probes are typically only needed for high voltage work (hundreds of volts). Even if you don't need hundreds of MHz of measurement BW with your probe - keep in mind that very low-BW probes might not be able to be compensated on your higher BW scope. This can lead to amplitude errors on signals at just several kHz frequency range.
Is it fair to say that probe compensation sets a "capacitance divider ratio (series C in the probe w.r.t. shunt C as specified in the probe) to such a value as to get the right amplitude? Is not the overall bandwidth governed by the input C and series resistance of the resistor divider and 50 ohm on the scope input? I think some equivalent circuits would be very illuminating here. Thanks for the video!
Ahh Rin on the scope is typicall 1 MegOhm unless set to 50 ohm.. of course... Also the C compensation is in shunt with the input impedance but that doesn't really explain what it does.
Basically, yes, the compensation adjustment is setting up the appropriate capacitive voltage divider between shunt cap around the 9M series resistor and the scope input capacitance. There is more to it (and more to what determines the BW), but this is the proper high-level description of compensation.
The magic of probe compensation is as follows: the scope has an input impedance of 1Meg//Cin, the probe has a 9 Meg resistor paralled with Ccomp (to be adjusted). When the two time constants 1Meg*Cin and 9 Meg*Ccomp are made equal, then the set "probe & scope input" behaves like a plain voltage dividor 1:10 with a flat bandwith. if your scope input is 50 ohms, then there is no use in using a probe (unless it's an active probe): a 50 ohms coax is enough.
👍🙏💖
🌟🌷🌟
First
Great video. You have an excellent way of explaining how these things work! There is an issue I'm wondering about. It is my understanding (limited as it is) that the total bandwidth of the scope - probe combination is less than the bandwidth of either the scope or the probe. (More specifically, it's the vector of the rise time of the scope and probe. For the formula, see en.wikipedia.org/wiki/Test_probe). If so, to maximize the benefit of the scope bandwidth, is it fair to say that you want a probe which has a higher bandwidth (say 1.5x to 3.0x the bandwidth) of the scope as long as the probe can be compensated to match the scope?
In most cases, yes. Tektronix is a bit of an exception - they typically underrate the BW of the probe so that if you put a 100MHz probe on a 100MHz scope, you get 100MHz of BW (not 70MHz). Most manufacturers don't do that.
Hello! If we want to communicate business cooperation with you, how can we contact you?
Thank you very much
Excellent information as always. I've been using oscilloscopes off and on since the mid 1960s but I still learned a few new things. Thanks and 73 de G3WGV.
Great information!
Alan thank you for another outstanding video. Just went and checked the capacitance input on my Siglent 200 MHz scope to all the probs (a bunch collected over the years). Found a couple that didn’t match so they were marked as such. Thanks again for the informative video.
I want to do a full alignment on an FT-857D. I have the equipment necessary to do the job but the one piece of information I am missing is what is the best way to probe the radios 7mhz sections. Really probing RF in general. Active probes are prohibitively expensive so I'm looking at probes on ebay with between 1-3pf or just using a piece of coax and a resistor. Would you mind giving me some hints on the best way to go about this? Thanks.
Thanks for the great info!
hi can I use 150 MHz probe on 50MHz oscilloscope?
So simple, but so important 😏
Thank you
I have your videos never enough. Would you like some time to build your own passive probe from scratch? The circuit looks not complicated.
as usual, the circuit is not a (big) problem, but mechanical part will be a nightmare! :)
i seen a lot of diy probes, some of them really nice, and those ones was crafted in copper clad board boxes. also, just some random plastic housings, and they was more or less usable! :D
Time is NOT an illusion. It's the "X" domain on your scope!
Time IS an illusion, lunchtime doubly so 🤣 (with apologies to HHGTTG) 73 de G3WGV
Alan, you have shown us all around your bench, lab etc. How about a video showing us how you shoot the videos ? Camera, camera usage, video equipment used, how you manage to do the videos etc.
Just an idea, "73"....
That's an interesting idea... ...I wonder how much interest there'd be...
@@w2aew The reason I suggest this, you do an excellent job of making clear precise videos. Even though your moving and changing angles etc. It appears seamless...
Another very informative video. I bought a Tek 2445B 150MHZ scope from a dealer in the US in 2016. It came with 2 probes, P6121 and P6122. I recently came to know after watching this video that the probes that came with the scope was 100 MHZ not 150 MHZ.
So how do we measure, characterize, and compensate for mismatched components, and what happens if you don't? For example when you can't get proper specs for your test gear. I have much research to do. :)
Would be nice to have a tutorial on active probe circuitry and uses.
Is it best to set the 10x on the probe or the scope ?
They need to match. If you have the probe set to 10x (to give you the highest BW and lowest capacitive loading), then you should change the scope's vertical scale setting to 10x so that the vertical scale reads correctly.
In my experience probes that are switchable 1X/10X (or 10X/100X) have no way to tell the scope what they are set to. So for that type of probe, you have to set BOTH the probe and scope to the same value. If you can't do that, then you will have to remember to divide the voltages shown on the screen or display by 10 or 100, depending on the ratio of the probe and scope settings.
@@lwilton Thanks I never knew that
@@w2aew Thanks I never knew that always had one set wrong than
Thanks.
Great, "keep us in the loop" video. Thanks for the look.