If you do any work on high-speed signals, you will quickly realize that detection is an art in itself. It is not enough to have a fast oscilloscope. You must have a probe fast enough to process the signal you want to see. In this case, almost all old probes are unusable: the input capacitance of the classic RC probes that you often see on low-bandwidth oscilloscopes begin to seriously load circuits far below 1 GHz. That’s why we are really happy to see [Andrew Zonenberg's]’s new work
.
The design of this new probe seems simple. Called Z
-Probe, transmission line probe or resistance probe, the circuit is used as a voltage divider, composed of 50 ohm input impedance of high-speed oscilloscope input and an external resistor to reduce the load of the circuit under test. In this case, the input resistance has been chosen to be 500 ohms, resulting in a 10-fold probe. In theory, building such a probe is as simple as soldering a resistor to the end of a coaxial cable. You can do this completely, but in reality, optimal design is much more complicated. As you can see in the schematic, just choosing a resistor of the right value will not cut it off at these frequencies. Even small 0402 resistors have parasitic capacitance and inductance, which can affect the response, and choosing a combination of parts that can increase the correct resistance but reduce the overall capacitive load can make a huge difference.
Don't be fooled: the relatively simple schematics obscure the complexity of this design. At these speeds, the PCB layout is as important as the resistor itself, and it is not easy to transmit the transmission line (especially the SMA footprint) correctly. By combining with Sonnet EM simulator for modeling and empirical testing, [Andrew] finally obtained a flat (+/- 1 dB) design in the 1.98 GHz range with a rise time of 10-90% of 161 ps. This is a quick exploration.
The probe has a variety of options, from fully assembled with traceable specifications to DIY versions made by yourself. You may know which options are needed.
We really want to see this kind of knowledge and comprehensiveness into the project, and we would love to see the Kickstarter project reach its goal, but the best part is that the design is open source licensed. In this case, making the circuit board layout open source is the key. The schematic can tell you half of what is actually happening in the circuit, and setting up the PCB correctly can be a long and frustrating job. So please take a look at this project, if you don’t have a probe that fits your fastest oscilloscope, then build a probe or better,
We have seen [Andrew (Andrew)] the working principle of an oscilloscope before, for example
Nice design! Is the housing a light black metal or injection molded plastic?
Related article: The parts price of 1 GHz active probe is 20 dollars, there is a lot of theoretical knowledge, as long as you can use some good characterization and calibration equipment, you can:
1GHz passive probe, no PCB (Therefore, there is no problem in characterizing PCB impedance):
I have built and tested the latter, and it works well. It also needs to be adjusted with a signal analyzer. I want to use a probe that is close to twice the bandwidth of the oscilloscope, so the probe roll-off is not a limiting factor, and it works well.
Every oscilloscope company has such a probe, although the bandwidth is two higher, usually 8-10GHz, the response is flatter.
I recently published an article about this type of probe in the Journal of Signal Integrity.
The disadvantage of several series resistors is that stray capacitance will increase nonlinearity
Yes, there are better detection methods, but none are close to this price range. Of course, products such as the Picoconnect 900 series, LeCroy PP066, Tek 54006A, and R&S RT-ZZ80 will do better, but these products are all priced at more than $1,000.
The probe is designed to compete with Pico TA061, Keysight N2874A and other products, and the price is only a small part. I have benchmarked it with TA061, and I believe it is the same OEM probe design as the N2874A, and has a significantly flatter frequency response and a bandwidth close to 1 GHz. If you know a probe that costs less than $600 and has better performance than my design, then I would love to hear it.
The prototype housing in the photo is 3D printed SLS nylon with 40% glass bead filler.
For various reasons, I want to use a pure polymer shell and there is no glass filler on the last board.
Very cool. Use the balance between capacitance and resistance to achieve a reasonable frequency response, and make some repairs to the HV probe. But it is difficult enough to exceed 1 MHz (HV probe is very large -200 kV +...). Higher bandwidth may be the next locked item.
Just an idea.
If you put the resistor on one end of the coaxial cable "sorta", then solder a resistance wire to the core of a coaxial cable, and then heat the core of the coaxial cable with a large current? Then pull the resistance wire part into the coaxial cable to replace the core.
Or through some calculations, use a metal tube with the correct diameter and a resistance wire, and then make a coaxial arrangement.
Put one end in the vise and the other end in the (accu) drill bit, then pull on the wire and rotate the drill bit to tighten the wire at the same time to make the wire very straight.
I will not do this now. My oscilloscope stopped at 50MHz...
In my opinion, the most common resistance wire (its specifications are the same as the coaxial cable core) needs to be a few meters long to obtain a suitable resistance. When you start to study heterogeneous alloys and special structures, you are back to the problem of the probe set.
Please be kind and respectful to help make the comment section great. (
)
The site uses Akismet to reduce spam.
By using our website and services, you expressly agree to our placement of performance, functionality and advertising cookies.
Tel
+86 400-003-5559
