I pimped my RSSI meter with that chip and an Arduino for data logging etc. Works great! Thanks!

You helped me a lot - had the same board - did not run. took it today out, tied Vset and Vout together. and Bam! a winner. took some measurments at 100Mhz, 500MHz and 1Ghz in steps of 10dB from 0dBm down to -50dBm. Works fine! 73 de DL1LEP

I have this same exact Analog Devices evaluation board currently on order. Thank you for going through the steps you did, as it will make my journey that much easier/clearer. The eval board is scheduled to arrive in another week or so. Fred

Another project? Great, more fun.Thanks for the look.

I was curious and did the exact same measurements using the RF Power Snitch, which is a diode detector. Turns out that for the range above -30dBm the Power Snitch does even better. So a simple diode detector can do well under certain conditions I think (this chip isn't cheap). The only thing that needs attention is temperature compensation. I was looking into this chip series for a new project.

Datasheets are your friend. They explain everything about the chip to stupid details. Enough to design circuits around it.

@KCLIBURN-mj9qx

2 ай бұрын

Without a datasheet............I wish you the best of luck and hopefully something doesn't smoke check!!!!!!

I tend to buy some of the more interesting things you review. I wouldn't get one of these without your review, but I'm getting 2 of them.

today my Rf lab got 2 other instruments to calibrate the power meter board, one Farnell PSG1000with frequency 10Mhz reference GPSO, and one HP436A power meter both calibrated, are fun to see the tinySA reading frequency and power properly.

I've had good luck using a plain vanilla AD8007 feeding the 10 bit arduino analog input. Within 1dB -60 to +8 dBm. I got a bit better linearity by using a lookup table for every 10 dB and interpolating between points. I used a precision voltage reference for full scale on the Arduino. My application was a remote power meter for an antenna range that transmitted the received level back on 433 MHz.

Hello MAESTRO, today I got the HP POWER METER 436A (very clean luck new little used) the reference from the HP 435A and 436A spot on 0dBm I'm happy I waiting for the Marconi 6950 is analogic and I go to start with, The chopper system on the detector is similar I think have some fun have ordained the N FET and I go to build the chopper inside the powerhead and put the dc reference in if work. This is the first step. (I think you doing very well and fast of me with your experience) Your videos are one inspiration for this project. Thank you!

@IMSAIGuy

3 жыл бұрын

the detector is built with no solder joints in front of the chopper. everything is gold on gold. This eliminates bimetalic voltage sources. The HP sensors also include a thermistor for temperature monitoring I believe.

I built one with the AD8307 chip into a box about four years ago; the spec was up to 500 MHz; characterized it from 1 - 650 MHz and found it to be quite linear from -74 dBm up to +18 dBm. A simple chart yields as you say, a really nice power meter for next to nothing. It actually surprised me with a very stable slope of 32.9 mV/dBm. Sufficient for most work around the bench and I don't have to pull out all the cables & sensors for the HP437 to do a quick measurement!

@IMSAIGuy

3 жыл бұрын

Did you use an A/D? If so how many bits do you think are required

@PapasDino

3 жыл бұрын

@@IMSAIGuy Boy, got me there...I simply used the board that looked exactly like yours but with the AD8307 chip; made a big table and went from -74 dBm to + 18 dBm for frequencies between 1 and 650 MHz at 50 MHz intervals (yep, crude, but it did work). Then plotted manually on graph paper (yeah, I know I'm a dinosaur!) and found the quite linear curve. The board did come with a 100 Ω input so just paralleled another 100 Ω to make it around 50. That's all I did...if I measure something I guesstimate in the table and arrive at a power level which is close enough for government work! ;-)

@IMSAIGuy

3 жыл бұрын

@@PapasDino Let me ask another way... Arduino? just ran the voltage into an analog pin?

@PapasDino

3 жыл бұрын

@@IMSAIGuy Nope, just measured the voltage from the PCB with a DMM! (yeah, it was a lot of changing frequencies and input levels with the sig gen).

@Gengh13

3 жыл бұрын

@@IMSAIGuy if you want a cheap, available everywhere, good enough ADC easy to use with Arduino I can recommend an ADS1115 board, it is a 15bits I2C ADC with a PGA and voltage reference that can be powered from 3.3 or 5v, and in the 2.048v range has a resolution of 62.5uV, more than what's needed to accurately capture the output of the sensor. The only disadvantage is that it is max 860sps, but if you are measuring a relatively constant power it won't matter. If you also need faster response you can always connect it to the internal ADC on the Arduino, it is also good enough for this application, in most Arduinos you have a resolution of 1mV or less.

Was just looking at the AD8307 last night , those critters are going for $13+ from Mouser/Digikey price has went way up on them since I last checked, those prebuilt China modules are cheap if they work then great!

@ruhnet

3 жыл бұрын

@Seth Lavinder You can get the AD8307 chips for free from Analog Devices under their sample program. They let you order I think 4 different chips every month or two weeks or so. I've gotten several different neat ICs to experiment with that way. There is no limit as long as you don't abuse the program.

i know it's been a year but how did you get the voltage values for the decreasing values of dBm scale?

"I'll add that to my list of projects." I've lost track of the number of times I've said that :) Great video as always but you didn't include the link.

@akashtanwar1939

3 жыл бұрын

No issues. I have a list but no progress.

The AD 8318 8Ghz the 8317 reach 10Ghz is perfectly feasible to put on the front end one of this detector and use the meter (some use one of this detector and one mini Babbuino (sorry Arduino) and make a USB power meter.

I didnt know that chip, recently i bought the AD8307 and the AD8313, and just yesterday while i was checking my diy dbm (AD8307), i found in the tinySa that while doing a LO output at 10 Mhz there is a Jump of 5db extra attenuation (-50dbm) when doing -45dbm and repeat from then on, it happens too in other frequencies...and it is not the dbm meter i checked it with an attenuator and its Ok.

Thanks for the video! Did you ever put the AD8318 in the project box? Do you have any other videos which show the project box you mentioned at around 17:45? I'd be interested in seeing the construction techniques you use.

@IMSAIGuy

2 жыл бұрын

kzread.info/dash/bejne/e3eK1rBxe6fLkrg.html

How do you use this board/RF detector, it would have been nice if the video showed at least taking one or two measurements.

Hello friend, I take your suggestion" to make one power meter need one power meter" I buy one HP 8484A powerhead sensor very clean from one well-reputed UK company at a very reasonable price to use with the HP436A , this means I go forward to build one other use this for calibrate.

The big problem is I'm not a designer on the RF land I work on the RF field installation radio relay and broadcasting station (I starting on the battleground 50year ago)

Can you please try to measure the differential input impedance?

How about doing your project box as suggested in your video. Season Greetings, thanks for all your interesting videos.

Hi I.G. maybe do a video on calibration, offset and gain etc...

So that others don't make the mistake that I did, (I bought 2 of these, I'll keep them because you never know when you might need them,) they also sell an AD8313, and the PCB photo is identical to the AD8318 board and it's also a logarithmic RF power metering circuit. But it only covers much higher frequencies.

I'd be interested in learning more about your process for generating the calibration values you were showing and how they'd be used to convert back to dbm.

@IMSAIGuy

3 жыл бұрын

I did gloss over that. The RF generator with calibrated signal levels was an HP 8921A. I created a table of volts vs dBm. This could be used to look up a voltage and convert to dBm. The next step was to curve fit the data using a linear fit. This is expressed as Y = A * X + B. or X = (Y -B) / A where A is gain and B is offset. In my video I had an additional offset. so my formula was dBm = ((volts - 0.955) * -42.2) - 16

@nvrumi

3 жыл бұрын

@@IMSAIGuy I think you're saying that you used linear regression to estimate the slope and intercept. You won't need to do a p-test on the regression parameters given how linear the relationship is between the variables. Well done.

@IMSAIGuy

3 жыл бұрын

@@nvrumi yes, but I try to avoid using terminology that beginners will not understand.

@richardphillips2405

3 жыл бұрын

@@IMSAIGuy Could you explain linear fit and offset. Does Y = dBm and X= Volts?

@IMSAIGuy

3 жыл бұрын

@@richardphillips2405 I gave you the formula: dBm = ((volts - 0.955) * -42.2) - 16

I admire your optical ultra-expert pick of an Eye Loupe. What type of lens, or lens system is it? Did you Buy, Make, or Recycle it?

@IMSAIGuy

3 жыл бұрын

In my office I have: www.edmundoptics.com/p/edmund-6x-deluxe-wide-field-magnifier/4705/ In the shop (on my videos) I use a DIY/Recycled magnifier. I saved big old camcorders and ripped the lenses off. The lenses contain internal elements that make great magnifiers.

@Subgunman

2 ай бұрын

@@IMSAIGuywow, I remember Edmond Scientific and their Optics division. A go to for science projects in the 60’s and 70’s!

Did you try AD8319 Power Meter?

Interesting video. Did you build your power meter? I am looking into something similar (to buy or build myself) to protect the input of my RSP1A, when using as Spectrum Analyzer, or when maybe even combining on the same antenna for Rx/Tx for the purposes of pan adapter for my analogue rig. RSP1A can handle only 0dBm on the input (very short +10dBm). So I would like a circuit that fast decouple the input if RF Power>0dBm. Also maybe to combine with a DC-block circuit, as that can be also a problem for the input. Similar thing for my nanoVNA. I am not sure yet what would be the limit for the nanoVNA, but similar story, you don't want to blow up your sensitive equipment if something stupid accidentally happened. And we all know, sooner or later, that we WILL do something stupid, haha. But the circuit would need not to degrade the input properties of the equipment of course. Any suggestions?

@IMSAIGuy

Жыл бұрын

higher end equipment will use this: kzread.info/dash/bejne/aGafmbaxdNK3e84.html

One thin that might be of a concern is RF noise levels that we live in. Made a DIY level detector whit MAX2015. Noise levels are on the order up to -40dBm on some bands, since that one GSM tower is just 300 meters from work place. On spectrum its insane level of noise. As a amature astronomer, this is worse than light polution and light polution is on insane level. Tin foil hat, aneyone? 😄

Yes, the sensor inside got two parts. first, the detector itself (diode or thermistor) is gold stuff, end the second part is the Chopper (two N FET), and one transistor the first amplifier on PCB the Zero, the feedback ecc.. The second stage is the way to connect the sensor to the meter and is reproducible if I reach this part to connect the second stage to the meter the calibration is with D.C. I think there is a different way for the detector itself (remember this whose old technology) instead of gold stuff can put one IC detector personally 3or 6 Ghz is good to me. Is creasy the meter £50 pounds, the connector cable £150, and the powerhead £800 is crazy!

@IMSAIGuy

3 жыл бұрын

Should be interesting. The calibration is tricky. It is calibrated over a range of dBm and calibrated over a range of frequencies. If the sensor head is designed correctly, DC to 3GHz might not need frequency adjustment. The key is very low capacitance. The 436 is just a chopper amp with a selectable filter. It could easily be replace with an arduino and software.

@IMSAIGuy

3 жыл бұрын

Agilent/Keysight now has a sensor with USB output which does just that.

@frankreal2258

3 жыл бұрын

Yes I read the service manual, the calibration, HP has one calibrator that is injected different DC value to calibrate the meter the detector itself is the hard part special diode or thermistor very carefully chosen to reach 20Ghz 40Ghz

@IMSAIGuy

3 жыл бұрын

@@frankreal2258 No that is not all of it. In the old detectors the frequency error calibration was done at the factory and printed on the detector. the user had to turn the dial on the front panel to match the cal data. Then they started to put the cal data into an EEPROM in the sensor, the user still had to tell the instrument what frequency was being used. the sensor can not tell. your detector will also have a frequency error and you will have to deal with that someway.

@frankreal2258

3 жыл бұрын

@@IMSAIGuy Yes I agree but the HP436A takes the power sensor from the 435A is one hybrid the calibration factor is on the front panel the same of 435A is not like yours, the power sensor needs to be calibrated at the factory and still the calibration factor percentage manually on the front panel. Someone tray to DIY one power sensor for the HP436 and partially succeed only use 2 diode BAT62 and get a limited frequency range and not terribly accurate, the problem whose the detector, the other part of the circuit mate with the power meter instrument the chopper work the zeroing work 220 kHz give the back to the meter the meter reading the power. the model after gets a more sophisticated digital feature.

Hello Maestro, the missing link is theAD8318 out is positive and I need to transform on negative some suggestion diagram, please

@IMSAIGuy

3 жыл бұрын

www.electronics-tutorials.ws/opamp/opamp_2.html

I have heard that Chinese suppliers often use chips that are "seconds". Is there any concern that the chip used on these boards might not be a first quality part?

@IMSAIGuy

3 жыл бұрын

that is always possible. you can have peace of mind by getting them from a real supplier

I have a 6GHz dummy load. How can I get a proportional sample into the TinySA from the transmission line to get an accurate power measurement?

@IMSAIGuy

4 ай бұрын

kzread.info/dash/bejne/ka5608adYseTndI.htmlsi=qbsUoepKYPxIOkBZ

@mrtechie6810

Ай бұрын

​@@IMSAIGuy Thank you! ❤

Hello, the mesure in millivolt are DC?

@IMSAIGuy

3 жыл бұрын

yes

@frankreal2258

3 жыл бұрын

@@IMSAIGuy Thank you

Watching ED TECH PH

How did you collect your data?

@IMSAIGuy

2 жыл бұрын

output from a HP8921A

Wavetech in Indy?

@IMSAIGuy

2 ай бұрын

Sunnyvale

So where is the link for these ?

@IMSAIGuy

2 жыл бұрын

search ebay

@KD0CAC

2 жыл бұрын

@@IMSAIGuy I ask becasue of the usall - products from china being copied by many - so using a tested source - I ask ?

Hello MAESTRO yesteday I buy 2 power meter one HP (Hewlett Packard) POWER METER 436A and one Marconi instruments RF Power Meter 6950 cosmetical very clean, my dilemma is the power head /sensor (very expensive for my pocket) the project is to put on function. my first question is: Make one DIY or " I find some sensor from General Microwave POWER HEAD N436A is compatible with the HP meter? the is other compatible? There is a way to put on the front end one Analog Devices RF power detection chip directly. Your great expertise is for me one incuragement. I hope not to bother you.

@IMSAIGuy

3 жыл бұрын

Sorry. You can't make a sensor. You can't substitute a sensor.

@frankreal2258

3 жыл бұрын

@@IMSAIGuy, there is a compatible substitute for the HP 436A

@IMSAIGuy

3 жыл бұрын

@@frankreal2258 The HP sensors that work on the 436 will also work on the 437 and 438 and maybe later models too.

Or you could just look at the AD8318 data sheet...

I found one power sensor Head COLINE DC- 2.0 Ghz 5 watt the only information sed OUTPUT 1mV/Watt some help or idea MAESTRO.

@IMSAIGuy

3 жыл бұрын

It is probably a diode detector with a built in attenuator. It will give a relative output but not absolute. You will need someway of calibrating.

@frankreal2258

3 жыл бұрын

@@IMSAIGuy Hello I open de sensor, I found this information: C Coline limited. 50 OHM POWER SENSOR Input impedance 50 ohms. BNC male connector. SWR 1.1 from DC to 2 Gh Output. 1mV/watt impedance. Output. 470 Ohms. Use 200mv range on D.M.M to read on watt. Accurance better of 5% DC to 2 Ghz True rms mesurament. 5 watt continuos rating 10 watt for 5 minutes MAX Calibrable from a DC source. inside one temperature sensor attached to the BNC. some heating system and one resistor plus one resistive trimmer to calibrate to the banana socket output. (the god side of the box is 80x30x20 god quality BNC male metal body)

@IMSAIGuy

3 жыл бұрын

Interesting device. Looks like it might be a bolometer design

@frankreal2258

3 жыл бұрын

@@IMSAIGuy to reach 3.5 millivolts I need take the ptt 1 minute

@frankreal2258

3 жыл бұрын

hello, I get one other Chinese power meter with the AD8318 RF Power Detector this one read 0.2 dBm on the 0 dBm HP 50 Mhz calibrator 0.224 millivolts on the oscilloscope with 50 ohms load, the tinySA is 2dBm high,