Time Domain Reflectometer

Yes, the type of resistor, the length and teh relative length of the traces are important. As is the width of the main output trace. With the picture you shown the resistors are far from the IC so the traces will be long and have a quite high inductance. This alone will lengthen the rise time, but if those traces are different lengths this will also add to the rise time. If you look at my design the the resistors attach directly to the IC pins then directly to a short, wide, low impedance trace that goes directly (no turns or angles) to the connector. Also the ground (in blue) is one whole side of the board and the Vcc is a copper pour on the red side of the board that is a complete un-broken 'circle' around the board. These two make for a very, very low impedance power supply improving the drive of the IC's and further shortening the rise time as low as possible.

Thanks. Your oscilloscope is the return signal sampler. I'm not sure what you mean by a sample circuit. The schematic is also posted, is that what you are looking for?

Hi, thank you for getting back to me so fast! The scope is indeed the signal sampler in your awesome example, however the circuit presented is being called a TDR which is technically incorrect. The circuit is a pulse generator and without a scope cannot be useful for TDR as it just generated a signal. I hope my tone isn't rude, just trying to be factual as I'm studying this intensely currently. I am doing a thesis on TDR devices for soil moisture measuring and looking to build a standalone tdr device that both generates signals and samples them. Your circuit and the one in which it is built upon (the square wave version) are fantastic and a cheap solution for creating fast rise times required in TDR. However I'm struggling to develop a circuit that can sample at 1GHz while also storing this data via an ADC. I'm exploring the "Standalone digital TDR (TDR-G2)" at: "https://hackaday.io/project/169386-standalone-digital-tdr-tdr-g2" and would really like to use your pulse generator and their sampler circuit to build the standalone device. Any thoughts? Thanks in advance ☺️

That's a very nice project. Not the same design goals as mine for certain. It would not be possible to utilize my pulse generator with that acquisition system in that project. At least not without a lot of modification of the generator hardware (it needs to be controlled by the MCU and needs to have precise frequency, pulse width and timing) and the control/analysis software. The relatively slow rise time of my generator would also mean a huge loss in precision and accuracy.

Hi, Thank you very much for your detailed response and for having a look at the device. I appreciate it a lot. I was hoping to ask you about your device now. I have made 4 of them. 1 with the 74AC14, which works great. The other three dont work properly but function the same, with a 74HC14. It can do the short delay pulse with the jumper in the S position. It can produce a square wave with no jumper pin, but the long jumper pin position results in an awful waveform. I used both the 74AC14 and 74HC14 for the original square wave generator, and they both performed the same. The 74HC14 actually outperformed it with rise time and ringing despite being a cheaper and lesser version. (3.9ns compared to 4.05ns) I can't understand, though, why it won't work with yours. I thought since they are functionally the same, that it should work. They have slightly different Schmitt trigger voltage levels, but I that's about it. Thanks

I'm not sure why you were only getting 4nS rise time with a 74AC14. It should be closer to 1nS. See my other little fast rise time oscillator board. The reason a 74HC14 won't work in this one is because of that 4nS rise/fall time they have. It's just too slow for the pulse generator.

Thanks.