Tuning RF circuits without a Network Analyzer?

[Mikebits]

Yes Robert is a very nice guy, and helpful too. For the sig gen, just be aware that you will have 2nd and 3rd harmonics in your signal. At the higher freq ranges the harmonics will be far enough out that they should be easy to filter out with a LPF. I am attaching a FFT of the sig gen output at 35MHz.
Let Robert know you heard about his sig gen from Mike in San Diego

 

[Mosaic]

Bob said 'Hi' and that the lower frequency waves are 'squarish' which implies significant harmonics.

 

[Mikebits]

I hope he did not think I was bashing his design, for $150 bucks its quite nice. His sig gen would make a great LO for a radio design, and with some additional filtering it would make a nice general purpose signal source. I also thought about a broadband push-pull amp which would reduce the even order harmonics. App note for a amp is here:
**broken link removed**

 

[Mosaic]

Well I plan to get in RF early in 2016 once I set up the work space for it. Got to build more vertical eqpt racks.
I have some recommended texts on order to browse thru as well. It's really the waveguides, antennae, SWR and impedance matching issues I have to come to grips with. Also PCB features and details to reduce noise and cross talk or use trace capacitance, PCB thickness and permittivity and trace inductance to 'manage' HF signals.
I hope to read the theory and try it out in practice, etch some PCBs and use the instruments as signal sources and analyze my results to come up with some fabrication techniques I can do reliably to perhaps a couple Ghz.
Microstrip for hobby design work. Stripline is possible too I suppose.

 

[Mikebits]

I mentioned some of the issues of using stripline in your other thread on precision references, as you had first mentioned it there but I think it went unnoticed, so I just ported over to the RF thread.

I don't think you would want to go stripline as that would force you to buy a 4 layer board (3 layer is not standard-mo money) since the stripline is sandwiched between two layers. Not only that, but most of the times the transmission line impedance will not be matched to the load so some sort of tuning is required usually adding a matching circuit at the load like series L, shunt C. You really have a tough time tuning Stripline as it is a inner layer.

Also if you look at a standard 4 layer board the stack up will make the stripline calculations more difficult.

MicroStrip is done on the top layer and you can get by with a 2 layer board, not to mention you can tune the trace to match the load if need be (assuming you can get your hands on a network analyzer).

Well anyways, that is my thinking on the topic, and is fresh in my mind as I been doing a lot of reading on RF layout this past week, but I am in no way a Jedi RF master like Radio Ron or Toni Stew, so they would be much better resource on this subject and I.

Grounded Coplaner Waveguide (CPWG) is another layout option you have, and the nearby ground makes soldering tuning parts easy. I have avoided this option as the math looks complicated and I am slightly a dunce when it comes to math (I had to take calculus 3 times before I got a C), so I just stuck with microstrip. Take a look at the CPWG math equations below:


Just to give you an example of what I mean about the tuning, below is a snip from a tuning recommendation from an Appnote provided by a chip antenna manufacturer. You can see where they provided pads on the transmission line for tuning.

 

[Mikebits]

I wrote down some of the math stuff for Microstrip. It is not to difficult since I understand it. In case your interested I scanned the page out of my notebook. I got pages and pages of this stuff, and only scratched the surface. I believe these numbers are just an approximation, and some fine tuning is usual.

 

[Mosaic]

The stripline concept comes from using flex pcb stacked (one single & one double sided) and superglue bonded (after tuning) under a perfectly flat (glass plate) clamping force (using my drill press (verified perpendicular) and a sprung plate to clamp) . It's just something I want to test out from a home fab POV, the permittivity etc. will have to be measured based on the response of the wave guide. So it's a rough calc and then empirical testing to see the results. Hopefully the eqpt I will have such as the microwattmeter with 1.3% accuracy to 300Mhz or 4% to 12 Ghz and the leveled sine wave sg504 will help with the measurements along with 50 ohm loads.

 

[Mikebits]

I use to work with RF about 20 years ago, and just recently got interested again. I have really forgotten so much, but it is coming back to me slowly. I really enjoy RF for some reason. Due to recent health problems, I can no longer do outdoor work. I want to do some home based business. I have seriously been considering a Network Analyzer purchase lately, used ones can be bought for around $3000 dollars. For the small design firms that have a occasional need to test or verify a portion of their design I could do it for them, saving them from having to purchase one on there own. Heck, just renting a VNA cost that much. I could also do custom RF circuit components using PCB design technique. Perhaps assist in prototype design etc. I have not suggested it to my wife as it is a big purchase, and most of my brilliant plans backfire in my face. Has anyone reading this ever tried to do such a business, any do's or dont's you might offer? Do you think this is a folly idea all together? I am not looking to get rich or anything, just earn enough to stay at my current modest standard of living. With Wifi, IoT, and the other wireless areas I think there is still a market for RF design. (Some may say, all the RF is in a chip now, this is not entirely true)

 

[Mosaic]

Ok...after looking at those VNAs...from $1k (just powers on 'appears' to work HP8753a, 3Ghz) to close to 6 figures.....clearly the cost benefit of owning these things is important.
So I'm gonna have a read about these matters.
**broken link removed**

 

[Tony Stewart]

Harmonic generators with VCO's driving ECL XOR gates with one side delayed to create pulses are common for markers and might work as a sweep gen.

UHF TV Tuners are often used for tuning and free.,

863 MHz to 865 MHz, is sometimes referred to as 'Channel 70' and tweeking the Tuner VCO can be used to shift to 868MHz which is a common interference to Ch 70 to sweep the tuner over a suitable MHz range with 5MHz harmonic markers. 10MHz /2 with <<1ns pulse width time with PECL XOR and delay line. Then directional coupler for tuning return loss of antenna and output impedance of driver for matching with a hybrid splitter modified form 75 Ohms to 50 Ohms changing 150R to 100R. inside then use as a DC3, Using the AGC on tuner also you can modulate output by 3dB .

One trick I ahve used is using the dual port hydbrid splitter and open cct on port B reduces insertion loss of Port A due to 0dB return loss from port B. Energy returned and forwarded to port A. This modulation can be used to calibrate AGC among other methods.

 

[Mikebits]

Bob said 'Hi' and that the lower frequency waves are 'squarish' which implies significant harmonics.

I suspect the higher frequency waves are squarish as well, only the limited BW of scopes does not produce a good representation as it most likely rolls off the distortion products. The best way to check would be to use a spectrum analyzer, which I also don't have. Xmas is around the corner, so maybe my wife will win the lotto and buy me my dream MDO tek scope

 

[RadioRon]

I use to work with RF about 20 years ago, and just recently got interested again. I have really forgotten so much, but it is coming back to me slowly. I really enjoy RF for some reason. Due to recent health problems, I can no longer do outdoor work. I want to do some home based business. I have seriously been considering a Network Analyzer purchase lately, used ones can be bought for around $3000 dollars. For the small design firms that have a occasional need to test or verify a portion of their design I could do it for them, saving them from having to purchase one on there own. Heck, just renting a VNA cost that much. I could also do custom RF circuit components using PCB design technique. Perhaps assist in prototype design etc. I have not suggested it to my wife as it is a big purchase, and most of my brilliant plans backfire in my face. Has anyone reading this ever tried to do such a business, any do's or dont's you might offer? Do you think this is a folly idea all together? I am not looking to get rich or anything, just earn enough to stay at my current modest standard of living. With Wifi, IoT, and the other wireless areas I think there is still a market for RF design. (Some may say, all the RF is in a chip now, this is not entirely true)

Hi Mike
I came at this business of being a one-man RF design shop from another direction and must admit that I had a fairly easy time putting a lab together compared to most people. I retired early, from a firm that made wireless products some years ago. About a year before leaving, I started buying surplus equipment from the company whenever it came available. Because I was an insider, I think they treated me more than fairly on pricing, so I got some reasonable deals and over time got the key equipment that I wanted. Since then, the gear sat for several years collecting dust and then I slowly got into it again in the effort to start a small product business with a partner. We didn't succeed in making that business viable, but having the equipment led to other opportunities that worked out well. Over these last few years, my impression is that it isn't too hard to find work as an RF specialist because there is such a large market out there and not so many RF-savvy guys available. Often, there are companies in their early-growth stage that have too many project ideas and not enough people, and they will hire good contractors to handle their overflow. You have to look around the entire country to find these companies, not just locally. I think that having you own gear makes you especially attractive, as long as you don't price yourself too high.

I know of a couple of other RF engineers that make a living doing design work on rf circuits and antennas with their own equipment in their basement. It isn't easy to sell yourself though, and depends a lot on how much demand you can find and how good you are at talking them into trusting you to get the job done. One thing to remember, customer service (ie. good communications) and hitting deadlines is often more important than the performance of the device you are working on.

By the way, I bought my latest VNA on ebay and while it was a lot of money I was lucky that it worked fine and so I'm happy with it. I spent more on this VNA than all the rest of my gear combined, but there's nothing like a good VNA for RF work, eh? Oh, also worth mentioning that I worked hard to find a good low cost set of calibration standards and still ended up spending $750 on those alone. If you are working below 3GHz, there are some cal kits on ebay these days at around $200 that are adequate.

 

[Mosaic]

RadioRon, can u point me to such a sub 3Ghx cal kit?

 

[RadioRon]

RadioRon, can u point me to such a sub 3Ghx cal kit?

What connector type are you calibrating with? I will assume SMA.

I have not used any of these, so I am only going by what I read on ebay. This one looks good:
https://www.ebay.com/itm/7-GHz-Low-...hash=item27d9fb624a:m:mKD2h6vycJ6bNovc3UaG73w

The seller appears to know what he is talking about and provides cal coefficients on disk for some popular VNAs. These cal coefficients are important. This one is a bit more expensive than some, but seems worth it. The seller offers a few different kits suited to different VNAs.

Here is one that appears to be more like a small step up from making your own, so I wouldn't trust it to 6 GHz, but the price is modest. I especially don't like that the OPEN is just a "through" and unshielded. Probably ok below 2GHz, perhaps 3GHz. Just my opinion.
**broken link removed**

 

[Mikebits]

RadioRon, can u point me to such a sub 3Ghx cal kit?

So were you the one that outbid me on my VNA on Ebay?

 

[RadioRon]

I see there was only one bid, so I gather the other bidder jumped in before you had worked up sufficient will to commit to the $2000. Keep looking, I have seen other D versions sell for down to $1500. I recommend an E version, but its hard to justify the higher price for hobby work.

 

[Mosaic]

Well , it seems that a 8753c with the s parameter kit & cables floats around $2500 these days.

Well, 2 much for me at this stage. How much better is the D or E versions? Are they worth it?

 

[Mikebits]

The thing I like about the D model is that the S-parameter test set is built in, where the other models it is a separate box. When looking online at the A-C models, be sure the S-test set is included.
Much of the RF wireless stuff seems to be hitting up to 6 GHz these days, that is another option you may want as most only go up to 3 GHz.

 

[RadioRon]

Its been a while since I compared the C, D and E versions, but the main thing I remember is that the internal processor is a fair bit faster on the E which makes it a bit easier to use especially with automation. The E also has a bit more screen capability (four plots on one screen vs two?). The E version also has some additional calibration features (they call it adapter removal). The jump from C to D was more signficant than D to E as far as features go. Hard to say which is worth the money, there are numerous detailed changes that you need to justify against your needs. If you are doing things for hire, go for the E version.

 

[Mosaic]

The thing I like about the D model is that the S-parameter test set is built in, where the other models it is a separate box. When looking online at the A-C models, be sure the S-test set is included.
Much of the RF wireless stuff seems to be hitting up to 6 GHz these days, that is another option you may want as most only go up to 3 GHz.

Why is the s parameter bundled here?
Ebay # 141815918369

Also the description mentions S parameter box compatibility for the D model here:
161867969532

Yet this sheet says S parameters are built in??
https://literature.cdn.keysight.com/litweb/pdf/5962-9770E.pdf?id=1000080939:epsg:dow