Very interesting Joe! and it is clear you have a lot of experience in measurements and RF. You also can teach as I followed a lot of your discussion. When you mentioned that the mismatch on one input would reflect back through the splitter and affect the signal going into the other channel it really rang a bell with me in how there are just so many pitfalls in RF measurements that only someone with a lot of experience, and likely having made a few mistakes along the way, would know. Thanks for sharing your knowledge!
Tek made a probe (before my time) P6048. That's a probe for 1MOhm scopes that behaves akin to a resistive probe for 50ohm scopes. It is described in their "Oscilloscope Probe Circuits" publication. They get the bandwidth by terminating into a T-coil, with an adjustable cap to augment scope's input capacitance. If one starts adding inductors to their terminations, why not go full monty? 😁 [EDIT:] Crunched some numbers: T-coil is conceptually clean and can be matched to the cable. Soo... If, for example, cable is .21uH/m and 83pF/m then 16pF cap corresponds to 40nH, i.e. two 20nH inductors (and a 50 ohm resistor to the ground, of course).
They claim under a pF tip capacitance but only 100MHz BW and only a 2.5ns Tr. I wonder how well it actually performs. In that link to Lasmux's active probe project in the description, I had compared several of my probes. You can see how the PP005 probe I showed in this video compared with a resistive probe. The cross over point was around 60-70MHz. www.eevblog.com/forum/testgear/12-ghz-active-probe-project/msg4988716/#msg4988716
0:40 Your through terminators with the SMD resistors look like a very nice electrically helpful geometry. However I would worry that one side being attached to the pin would subject the resistors to mechanical stress during plugging/unplugging, possibly cracking them, disconnecting them, likely intermittently. That sort of problem might go unnoticed -- quite annoying. It we be so nice to somehow create a more flexible connection to the resistors -- though I don't have any ideas how to do that at the moment.
There were a few clues in this video that I left for the savey viewer to sort. The inductance of the TH resistors may improve the match. Using these leaded parts also mitigates the stress you mention. Consider that when we are talking about using it with a scope, each one would be different and require a different network. Even if you happen to compensate it correctly, the improvements most likely will be minimal. I basically suggest better tools may be the best solution.
19:57 "so just adding this termination resistor has greatly improved the flatness of this" -- I'm not seeing the improved flatness. The knee appears to be in the same place. Which aspect of the curve is flatter? I'm sure I'm missing something here. Hmm... by "improved flatness" do you mean the better ratio between say 1 MHz and 350 MHz? No resistor: 1 : 0.494 = 2.02 versus with resistor: 0.5 : 0.375 = 1.33?
Hello, I really enjoy the channel’s content! Could you share boylestad’s book title? I would like to take a glance at it, looks like a good reference. Thank you!
This was really well explained, and actually easy to follow for a beginner. Thank you very much!
You're welcome.
High-quality content and detailed explanation, as usual. Thank you!
You're welcome.
Very interesting Joe! and it is clear you have a lot of experience in measurements and RF. You also can teach as I followed a lot of your discussion. When you mentioned that the mismatch on one input would reflect back through the splitter and affect the signal going into the other channel it really rang a bell with me in how there are just so many pitfalls in RF measurements that only someone with a lot of experience, and likely having made a few mistakes along the way, would know. Thanks for sharing your knowledge!
You're welcome.
Thanks Joe.
You're welcome.
Good video, I guess no commercial 50 ohm terminators sold by scope manufacturers do this? I didn't see any from a quick search.
Custom match them to a specific scope? Not that I am aware of.
Tek made a probe (before my time) P6048. That's a probe for 1MOhm scopes that behaves akin to a resistive probe for 50ohm scopes. It is described in their "Oscilloscope Probe Circuits" publication. They get the bandwidth by terminating into a T-coil, with an adjustable cap to augment scope's input capacitance. If one starts adding inductors to their terminations, why not go full monty? 😁 [EDIT:] Crunched some numbers: T-coil is conceptually clean and can be matched to the cable. Soo... If, for example, cable is .21uH/m and 83pF/m then 16pF cap corresponds to 40nH, i.e. two 20nH inductors (and a 50 ohm resistor to the ground, of course).
They claim under a pF tip capacitance but only 100MHz BW and only a 2.5ns Tr. I wonder how well it actually performs. In that link to Lasmux's active probe project in the description, I had compared several of my probes. You can see how the PP005 probe I showed in this video compared with a resistive probe. The cross over point was around 60-70MHz. www.eevblog.com/forum/testgear/12-ghz-active-probe-project/msg4988716/#msg4988716
0:40 Your through terminators with the SMD resistors look like a very nice electrically helpful geometry. However I would worry that one side being attached to the pin would subject the resistors to mechanical stress during plugging/unplugging, possibly cracking them, disconnecting them, likely intermittently. That sort of problem might go unnoticed -- quite annoying. It we be so nice to somehow create a more flexible connection to the resistors -- though I don't have any ideas how to do that at the moment.
There were a few clues in this video that I left for the savey viewer to sort. The inductance of the TH resistors may improve the match. Using these leaded parts also mitigates the stress you mention. Consider that when we are talking about using it with a scope, each one would be different and require a different network. Even if you happen to compensate it correctly, the improvements most likely will be minimal. I basically suggest better tools may be the best solution.
Guess we all just need to own lecroy oscopes...
Or just buy tools that do the job you need them to.
Is the book you use Introductory Circuit Analysis by Robert L. Boylestad and Brian A. Olivari?
Sounds right. It's a pretty old book.
19:57 "so just adding this termination resistor has greatly improved the flatness of this" -- I'm not seeing the improved flatness. The knee appears to be in the same place. Which aspect of the curve is flatter? I'm sure I'm missing something here. Hmm... by "improved flatness" do you mean the better ratio between say 1 MHz and 350 MHz? No resistor: 1 : 0.494 = 2.02 versus with resistor: 0.5 : 0.375 = 1.33?
Hello, I really enjoy the channel’s content! Could you share boylestad’s book title? I would like to take a glance at it, looks like a good reference. Thank you!
Sure, it's "Introductory Circuit Analysis" It only covers the the basics.
12:00 I think you mean an E8357A, not an E8557A - E8557A isn't a valid agilent product
That is correct. I'm surprised that's the only mistake you caught as long as the video is. Normally there are a few.
@@joesmith-je3tq the big picture😁
Thorough and enjoyable. Haven't seen that textbook in over 40 years.Thanks for your time!
I had a friend who was interested in circuit design and I came across a used copy in a book store that I gave to them. It's a good starting point.
@@joesmith-je3tq agreed!
What textbook is it ?
@@patougass watch the video again.