RF downconverting

Joerg, The target direction is not known. But it seems that the Optics guys went and changed the requirements. It seems that they now want to use a center frequency of 180MHz +/- 60MHz. I think there reasoning was noise from radio station(fm band). I still think the best approach is in downcoverting in order to reduce power (fast sampling, moving data around). One possible solution that has been kicked around would be to use a part such as the Atmel A/D AT84AD001 to do the conversions. Maybe the best solution would be to directly drive the A/D. I need to look more closely at the IC specs. I appreciate the help. There are only two EE's in the company.

david

[...] [...]

It depends on whether the scene is constant enough or not. If the scene contains some objects moving at fixed speeds, you can still undersample at different rates and then work out what the frequencies are based on that.

Take the frequency of 10MHz as an example. If it is aliased to 1KHz,

73KHz and 100KHz and the 100KHz measurement is zero, you know that there is no 1MHz. In fact the noise will set some lower limit on the ability to say that there is no 1MHz. Basically you have to solve a bunch of equations in the form of:

Y = X1 + X2 + X3 ...

Where the Y is the measurement and the Xes are the true amplitudes. If you have enough equations and enough SNR, you can find the Xes.

I agree that a frequency counter would be easier in this case. Since the bandwidth is many times the frequency he may want to measure, getting and estimate and then switching in a filter would give better answers.

--
--
kensmith@rahul.net   forging knowledge

Hello David,

Ahem, did they mind the fact that this puts the upper sideband smack dab into the upper VHF TV band? Out here we have a station in there that could make a small light bulb glow.

Every time a pilot announces a turn to left base we also see a big fat AM signal at 123MHz because he is usually just a few hundred yards away at that point. Same if a "big iron" pilot is on the glide path into Mather field. I could listen to that with a 1N4148 and a piece of wire (and know that my Fedex package is coming ...).

Is it one target or multiple? If one or just a few you could, as someone suggested, sweep that whole upper and lower range and see where the echoes are. Kind of like spectrum analyzer style.

Anticipating EMI from radio and TV stations I'd consider doing that at least twice with different center frequencies, to reduce ambiguity.

Lucky you. Some of my clients only have one EE.

Regards, Joerg

Was there more than one version? I have a copy dated 1970, Library of Congress card number 69-13615 on my desk.

There is a paragraph about PRF discriminators on page 29-16 that I can scan if you need it.

--
?

Michael A. Terrell
Central Florida

Hello Michael,

Yes, there are two. The first was the blue one and 1990 they came out with another in red. I had both but donated the older one to our library. Libraries are really happy when they get high-tech books and this kind of stuff doesn't really age that fast. I didn't notice much difference in terms of the fundamental concepts.

Thanks. I don't but maybe David does?

Regards, Joerg

I got lucky on this copy. I bought it for $2 about ten years ago from a used book dealer at a flea market. I talked him down from $5. ;-)

--
?

Michael A. Terrell
Central Florida

Hello Michael,

That's a steal. For the 2nd edition I had to plunk down $89.95 plus tax at the Arizona State University bookstore :-(

Regards, Joerg

I've found several other bargains, like a copy of the Radiotron handbook for less than $5, but the Skolnik book was the best deal so far.

I found it quite useful when I had to identify and make a replacement feed for a poorly designed three meter C-band dish, after the owner removed it for the summer to prevent theft and then threw away all of the paperwork. This was pre-internet for me, and the other possibility was a small outfit that built dual feed conversion kits for five meter dishes for broadcast and CATV headends. It would have taken me weeks to get the formulas from the designer, and he had a bad habit of transposing equations when he wrote them out.

I wrote a quick program in Commodore Basic, entered the measurements I had taken from the dish and it spit out the answers. I was within an eighth of an inch from what I calculated, when I had the best signal.

--
?

Michael A. Terrell
Central Florida