Fundamentals of RF Power Sensors

[@AllElectronicsChannel]

Support the channel becoming a Patron: patreon.com/allelectronics

[@markonis100]

I think this is the most in-depth explanation of the diode RF power detector available on youtube so far. Thanks, Greg.

[@AllElectronicsChannel]

Thanks!!

[@ebayscopeman]

This is the most concise explanation of diode function at low currents I have ever heard. It is very interesting that the Z of the diode at small currents is very different for DC verses AC current. I read years ago a very good explanation of this where various types of diodes were used as small signal AC detectors. This came in very useful when selecting diodes for a Brune bridge type power meter design. If I come across it again I will try to post a link. Keep up the good work!

[@AllElectronicsChannel]

Thanks!!

[@igorb4650]

Good description, the only thing that would benefit it is showing actuall frequency content after a diode on SA/FFT

[@anoimo9013]

Very nice explanation and deatails not commonly available. From the graphs I think its been implied that the signal applied to the diode has already some dc bias to support its non linear but critical asymetric conduction to have a meaningful reading on the instrument

[@baghdadiabdellatif1581]

Thank you very much. This is the best explanation ever. Yesterday I was reading the instruction manual for the URV3 millivoltmeter. I found that the RF probe was used as an indicator, but when adding 20 dB or 40 dB attenuator, it reads RMS voltages with an error rate of 3% for frequencies less than 200 MHz. According to your explanation, it seems that the accuracy of the measurement is to reduce the input voltage in order to ensure that the diode works in the non-linear region. I have two requests. Please explain the use of four diodes in the RF probe and what are the maximum and minimum voltages input by using attenuators to ensure the most accurate measurement of DSBSC ( with tow tones) & DSBFC power

[@dominikgs]

Rectification can be seen as mixing the signal with itself to create all kinds of sum and difference frequencies. The following capacitor is then the low pass with very low cut off frequency to cut out the DC point in the spectrum. So for me both points of view are undivideable connected. But I think you are right that this point of view isn't as much spread. Often those changes in perspective add so much to the understanding so thank you for that nice video. Keep up the nice work.

[@AllElectronicsChannel]

🍺🍺🍺

[@allthesecontinents]

I'm not in RF whatsoever, but I watched, enjoyed and understand the whole video! Wow!

[@atakaravelioglu301]

I didn't think until now that the diode was behaving outside of the switch. It's fascinating. Thank you Mr. Gregory

[@AllElectronicsChannel]

Great!

[@yakovdavidovich7943]

Wow. This was a fantastic video. Thanks for continuing to produce such excellent content.

[@AllElectronicsChannel]

Glad you enjoyed it!

[@arashghasemi]

Looks like when the diode starts to charge the cap, it takes significant power from the source? And thus the measured power is not accurate... Why not putting an opamp in voltage follower configuration (buffer) just after the diode? It will drive the cap. A single rf npn in emitter follower should work as well. Also additional filtering can be put on the top of the opamp or NPN config directly instead of cascading?? Some thoughts 💭

[@AllElectronicsChannel]

Hummm!!

[@two_number_nines]

An opamp would be horribly unpredictable and noisy if you attempt to use it at RF. (assuming you find one without the low pass filtering internal cap). You would need a second power detector to constantly calibrate the opamp based one. The low current draw by an opamp follower would mean that the internal capacitance of the diode would dominate the main component behavior so much it would no more be a diode at this point. An emitter follower has no advantages over a diode, as both have the same input-voltage--scaled-output-current function.

[@manjudevidevi6082]

very good explanation sir

[@AllElectronicsChannel]

Thanks!

[@tsengirv]

Perfect interpretation!

[@kaleoride]

Excellent presentation, I learned a lot, thank you!

[@johnaweiss]

Is the following correct? The diode resistance is inversely related to voltage. As a result, the average voltage after the diode is not zero, as it would be if we replaced the diode with a resistor. That non-zero average is proportional to input voltage (ie power). Is this a typical RSS detector? Or, free air power detector? What determines the knee voltage of the diode V/R curve?

[@AllElectronicsChannel]

Inversely related to *current*

[@two_number_nines]

Does the diode cause any main frequency/overtone reflections back to the source? The parasitic capacitance of the diode in series with the output capacitance acts as a capacitive load after the double-100R-resistors, so wouldn't that push the impedance seen by the load-side-end of the cable to be more capacitive? Wouldn't the asymmetrical load introduce a DC level shift into the RF line? I would put 2 diodes and capacitor banks into a half-bridge voltage doubler to increase the measurable DC voltage and have a more symmetrical load on the line. Although the diode here is used in a mode far from its ideal switch model, the circuit is still based on the fact that the diode asymmetrically discriminates conductance in one half of the cycle vs the other half, which is still a valid model in generic power electronics until you get into the synchronous rectifier territory.

[@AllElectronicsChannel]

It causes. Two diode is a possible topology.. but it may increase complexity and reduce bandwidth.. more parts.. it is a big tradeoff

[@OctavMandru]

Gregory, doesn't the diode internal capacitance have something to do with this behavior? We use diodes as RF switches, but a small DC bias is needed in order to allow RF to pass.

[@AllElectronicsChannel]

The internal capacitance is actually a bad thing here! Imagine that the capacitance is bypassing the non-linear behavior, limiting the maximum frequency!

[@OctavMandru]

Ok, you are correct! I was actually considering a series capacitance, kind like we have in a PIN diodes. I was trying to push somehow the mV levels into the diode bias range.

[@OctavMandru]

My SigGen just arrived so i could actually experiment this. Once I had my setup I was amazed. Now I understand mixers much better. You really should do an in depth video about this, I don't think I have seen this shown on an oscilloscope.

[@AllElectronicsChannel]

Great!! "The best programming language is SOLDERING" Bos Pease, hahahah!

[@aduedc]

Great presentation. To improve you can use germanium diodes or schottky diodes And why not bias the diode with say 10 uA current source. The biasing would put the diode on the edge of conduction.

[@yakovdavidovich7943]

Isn't the whole point of the video that the resistance modulation occurs before the forward voltage threshold, and thus we expressly do *not* need voltages at that level? Adding bias would just move the "kink" in the curve...

[@AllElectronicsChannel]

It moves the kink but can help for detection of very very low input powers... where the modulation is too small.. but we are talking off very low bias int the range of 500nA to 5uA

[@like_cabezon3810]

Hey gregory I have following this topic since that year you started to design your Power sensor and Since then I have read a bunch of articles you have recomended in your blog entries about this topic and some others and this video condenses the useful information in this topic very well. I don’t remember very well but I think I have read some article that in it discusses about using a tappered transmission line to transform the 50 ohm impedance to a higher one so providing the same power to the sensor create a higher voltage excitation in the pn junction improving the detection in low amplitudes. IMSAI guy has a video about and old power sensor and he said in that video that he used to work with the people designing those microwave sensors and that they used also that tappered transmision lines and also a differential diode configuration with two diodes. Have you did any experiment about this impedance transformation and if it is worth it? Thank you very much for your videos your content is top in the platform about rf.

[@AllElectronicsChannel]

Thanks! I didnt test any impedance transformstion technique.. Im not sure how it behaves for largue bandwidth detectors...

[@jorgerive7335]

Do you have any pointers on how to simulate this on LTspice or any other Spice based simulator?

[@AllElectronicsChannel]

Never tried...

[@sersheva]

Gregory, if we help and polarize the diode (with rf choke in the bias and dc block), it will make it even more sensitive to even lower rf powers? or it will make it more linear? I mean we help the diode put already into conduction mode.

[@AllElectronicsChannel]

Yep! It helps for detection of very low power levels !

[@saurabhambulkar1]

I don’t think it works with silicon diode with 0.7 forward voltage. Please specify which diode will be suitable for signals of 10mv and below.

[@AllElectronicsChannel]

It works.

[@ghlscitel6714]

For low amplitudes backward diodes had been used in the past.

[@AllElectronicsChannel]

Humm... interesting! I never touched a backward diode.. only read about

[@ghlscitel6714]

@@AllElectronicsChannel It is actually a wrecked Germanium Tunnel diode. It's dynamic range however is small, perhaps only 0 to 0.3 volt.

[@AllElectronicsChannel]

It is available or is something to get from Russians on Ebay? Hahaha

[@ghlscitel6714]

@@AllElectronicsChannel General Electric made them until the 1980s. Today however I would try to get a Nano-diode.

[@noslidemais]

[@AllElectronicsChannel]

🙏🏼🙏🏼🙏🏼

[@dn0hr9879]

🎉

[@AllElectronicsChannel]

🍺🍺

[@FUNKLABOR_DL1LEP]

ok, very interisting and good explained. but audio very annoying audio, and cutting auf audio is 👎+ very frequent repetitions - waste of time :-/ but - followed . 🤫🤗73