Strange behavior (RF stuff,Radar cantenna)

[AwArD_RzD]

Weird thing here, and i can't figure out what the problem.

I have these three part, a vco ZX95-2536C+ (http://www.minicircuits.com/pdfs/ZX95-2536C+.pdf) ,an amplifier (http://www.minicircuits.com/pdfs/ZX60-272LN+.pdf) and   a 3 db attenuator  (http://www.minicircuits.com/pdfs/VAT-3+.pdf).

I use 8 aa battery, a 7805, two 100nF and two 10 uF cap for the input output filtering and a LMR-400 cable (http://www.ame-hft.de/pdf/lmr400.pdf) to connect the circuit to the HP8594e.

The problem is when i connect these part together VCO--3dBATT--AMP i have so much crap showing on my SA,

at first i found that my power supply added some noise and random frequency and my neon over my bench added some modulation on the tune port of my vco so i shut off the light and powered the VCO and AMP with the 7805 on battery. Now the peak frequency look good and i see much less modulation of the tune port on the 100 khz span when i look at my VCO freqency. But when i look at the 9 khz to the 2.9 ghz span of the SA, it's a mess (first picture)..  peak everywhere and all thing move when i play with the tune port so i'm sure this is generated by the circuit.

I tested the circuit with the VCO and the 3dB attenuator alone (second picture) well thing are better i have some peak around 340 mhz and 1.6ghz nothing to worry about i think.

I tested the amplifier with my e4421b at 2.4 Ghz, (third picture) the noise floor is great and i don't have any other frequency expect the 2.4 Ghz generated by the signal generator.

So there is my problem, everything work alone but when i put them together it's not working, i tried to add capacitor directly on the VCC/Ground input of the VCO and the Amplifier but nothing change, i even closed all the breaker in my house except the one of my bench to be sure no computer or other appliance generated these frequency and nothing helped. The only thing i didn't tried is to put the circuit in a closed box to clear all the EMI.

These part is for the Radar cantenna course (http://ocw.mit.edu/resources/res-ll-003-build-a-small-radar-system-capable-of-sensing-range-doppler-and-synthetic-aperture-radar-imaging-january-iap-2011/projects/MITRES_LL_003IAP11_proj_in.pdf) . I did exactly what explained in the PDF, and well the result was awful when i recorded the Doppler shift and if found all these frequency on my Hp8594e.

So this is a normal behaviour for these part ??  I'm a complete noob for this rf stuff. i had some experiment with radio/tv frequency in my CEGEP class but i never encountered something like that. I have two amplifier and the two give me the exact same result....

Now my battery are dead and i need to wait for a full recharge before attempting other test. If anyone have an idea of what causing that i welcome your advice !!!

[ConKbot]

Check your power levels, you have a > 0 dbm signal (+6 dBm in the datasheet), and 10 db of attenuation dialed in on the SA.  If youre overdriving the mixer, you get a bunch of harmonics/mixing products/ other crap going on that makes that measurement useless. 

I havent found anything cut and dry, but it looks like the 8594, but a lot of the specs are for -10 dbm at the mixer.  Your reference amplitude is 0 dbm, you have 10 db of attenuation(also alluding to the -10dBm max input of the mixer) , which means a 0 dBm signal would have the mixer maxed out, and anything going over would be overdriving it.  Change your input attenuation from auto to manual and up it to 20dB, and all the crap on screen should clear up.  Or leave it on auto and change your reference level to +10 dbm, you will probably hear some relays clunk and attentuation will change up to 20 db.

[VK5RC]

It's hard to track issues,  power supply de-coupling really important,  over driving makes harmonics (sometimes deliberately),a lot of guys put 3db attenuators between each module to moderate reflections/standing waves.  Check ground issues / loops.
Remember at 2Ghz you can couple a signal very well between 2 bits of copper next to each other as if they are soldered. Good luck!

[KJDS]

If a signal is above the top of a spectrum analyzer screen, then it is likely that some spurious products will be visible on full span. Given that the signal is above the top line, then if you add an amplifier it will get worse.

Once you've made the measurements again with the ref level set correctly then you'll have some results that will either be ok or show that you have a decoupling or stability issue.

[tonyarkles]

Check your power levels, you have a > 0 dbm signal (+6 dBm in the datasheet), and 10 db of attenuation dialed in on the SA. 

I'm not the OP, and I have no idea if you're right that this is causing the problem, but I just want to say "nice catch". That's a piece of advice that I'll definitely be tucking away into the knowledge bank for some day when I get weird crazy signals on the SA. I would have probably encountered this exact problem the other day if I hadn't been lucky enough to put a 20dB attenuator on. I knew exactly what kind of power I had though and was worried that I was going to go over the "do not exceed" limit written on the front.

[yramgu]

Yep, looks like you're overdriving your SA. Increase your ref level until the fundamental spike can be fully visible. The other spikes should disappear.

[AwArD_RzD]

Damn, i used two wifi antenna to send the signal from the circuit to the SA and the -10 dbm signal i get  look very good and nothing show elsewhere so i was over driving the mixer... i never played with any signal over 0 dbm so i learned something today !!


but what can explain the Doppler result i have there in the picture, at first when i saw the crap on the SA from the ouput of the TX amplifier i was sure that was all the mixing up frequency getting back on the RX Antenna mixed with the TX signal. When i do some walk test in front of the RADAR i see something happening on my scope but alot of noise/crap and the result is not good in MATLAB when i record it...

I understand the circuit can be very sensitive, when i move on my chair the peak move so everything around can have an effect, but no where the MIT guy mention some problem like that and they have all the thing mounted on a protoboard with wire everywhere !! Maybe it's my pcb board now who causing this issue (in my test everything looked the same with the over driving) so i never suspected my pcb... Well i will try to rebuild the whole TX/RX circuit and i will put the two wifi antenna and see what the circuit mixer send me if it's a good flat line since there no doppler shift i will be sure the RF stuff work so i will start looking at my pcb.

The Modulator/Video amplifier circuit are in the MIT PDF i linked in the first post, if anyone can check the schematic to be sure i didn't missed anything (trap for young player), i did exactly the schematic on my own PCB. I'm sure it's not a Matlab problem since i can reproduce the same result they did with the wav file record in the project folder.

So PCB or my laptop sound card....

[ConKbot]

The only problem I see on a quick glance in his circuit is that he's not AC coupling the output of the LPF, despite using a +5v virtual ground.  The sound card probably has an AC coupling cap which saved his bacon on this one. How are you powering it? Are you battery powering it? Or a power supply/wall-wart?  And is its output isolated or ground referenced? The PC input will be referenced to mains ground, so if anything else is, then you could be getting ground loops causing problems. 

[AwArD_RzD]

The only problem I see on a quick glance in his circuit is that he's not AC coupling the output of the LPF, despite using a +5v virtual ground.  The sound card probably has an AC coupling cap which saved his bacon on this one. How are you powering it? Are you battery powering it? Or a power supply/wall-wart?  And is its output isolated or ground referenced? The PC input will be referenced to mains ground, so if anything else is, then you could be getting ground loops causing problems.

Yeah i saw that too but i knew my laptop had a ac coupling cap so i was not worried, the RF stuff is powered by battery, the laptop is floating and the video amp is on my bench power supply. For now i had some test done, the whole RF circuit is powered and i use the Video amp (GAIN,LPF) and thing seem to work (see picture). Now i need to connect the modulator to see if this cause the "ripple" when i record the Doppler shift.

Update :

The modulator seem to cause the issue, the same ripple is recorded when i use it to modulate the frequency but still don't know why, i will try to modulate the frequency with a func gen to see if it's something wrong with the modulator circuit.  I did a first test with a pot to manually change the frequency and the result don't have any ripple so i will check that tomorrow.   

[LeoUCDavis]

Hi

You can use a microcontroller to generate the modulation signal. It's simpler than the analog circuit. Check this report from my students:

https://drive.google.com/file/d/0Bzbq9fPL-_ZRY1huOXk1TnFsaTg/edit?usp=sharing

The design for the microcontroller modulator part can be found here.

https://docs.google.com/document/d/1ux03N-yWF2Y-B5jhtcrvC7al3iGTmjGLSG7Sg5OJcFY/edit?usp=sharing

Best
Leo

The modulator seem to cause the issue, the same ripple is recorded when i use it to modulate the frequency but still don't know why, i will try to modulate the frequency with a func gen to see if it's something wrong with the modulator circuit.  I did a first test with a pot to manually change the frequency and the result don't have any ripple so i will check that tomorrow.