Antenna question

FWIW the manufacturer of my SAW based CPCA 434 MHz transmitter recommends using a 50 ohm antenna on the receiver. My transmitter fits inside of a fob and does not have an external antenna.

Thank you,

I assume you mean this thing: I don't have a schematic on that particular receiver, but did find this data sheet for the PT2272 decoder chip that includes a receiver schematic. See Pg 4. Notice that the receiver is a TRF (tuned-RF) type. I happen to have an XD-RF-5V PCB, which is similar to the eBay item, but without the PT2272 decoder. (Incidentally, don't believe the RF specs on the above page).

Notice on both the PT2272 data sheet schematic, and the XD-RF-5V schematic, that the RX antenna is either coupled to the front end tuned circuit through a single small value capacitor or directly connected to the tuned circuit. The antenna impedance is quite a bit higher than 50 ohms. If I put a 50 ohm load across the parallel resonant tuned circuit, the Q would drop, the tuning would change, and the sensitivity would decrease. If it were 50 ohms, there would be a capacitive divider or tapped inductor on this tuned circuit. On most such products, the PCB is supplied without an antenna connected because it works better with no antenna than with one attached.

I could measure the input impedance of my XD-RF-5V PCB, but I'm doing some last minute firewood today and tomorrow and don't have much time. However, I can guess(tm) the RX RF input impedance at resonance as roughly equal to Xc of the 2pf tuning cap at 315MHz. Xc = 1/(2Pi*f*C) = 1 / (2 * 3.14 * 315MHz * 2 pF) = 253 ohms.

Incidentally, the SAW oscillator transmitter has the same problem with the 50 ohm antenna. Typical data sheet (for 433.925MHz): Notice the drop in Q from 11,600 (unloaded) to 1,850 (50 ohms).

Nobody has made a super-regenerative RC receiver since the stone age of radio. Such a receiver would probably not pass FCC Part 15 incidental radiation rules, be horribly unstable, and vary horribly in performance over a production run. What's common these daze are key fobs that use SAW (surface acoustic wave) oscillators for transmitters, and TRF (tuned RF) receivers because they are cheap, simple, stable, and have no local oscillators to irritate the FCC. The real advantage is that since most key fobs use AM modulation as OOK (on off keying) for low speed digital data, the generally low RX sensitivity of a TRF receiver provides some degree of noise and interference immunity. Were the industry to produce an RC receiver that actually meets their published specs, it either wouldn't work well due to all the noise and interference. Were it a proper receiver, with good sensitivity, good selectivity, functional AGC (automagic gain control), noise limiter, etc, it would cost too much to use in a key fob type RC system.

I was prepared to walk you though the exercise, starting with the systems design, but apparently you don't want to design something. (Last time I checked this was still sci.electronics.design despite the preponderance of politics and wildly off topic postings). So, before I burn more time trying to answer you questions, are you planning to design something, or are you shopping for an eBay or Amazon solution for the original problem? I'm not sure I want to get involved in that, which is why I ignored your original posting. The problem is that most if not all of the specs for such devices posted on eBay and Amazon are either wild exaggerations or outright fabrications. I can't calculate whether it's going to work for you without having possession of the devices and making measurements and I don't want to recommend something that doesn't work or that I haven't tested.

You seem to have a fixation with the antenna as being the solution to your range problem. In my limited experience with such devices, I found that adding antennas usually makes things worse, unless I retune the receiver to compensate for the antenna. That also means that touching the antenna, or using it near large objects that can add capacitance, will also require re-tuning. If you are going to add an antenna, it has to be rigid, firmly attached, and unlikely to be leaning against something when in use.

Had you decided to continue the design exercise, I would have offered a better example using a typical key fob and typical receiver such as the XD-RF-5V PCB I mentioned previously. The likely resulting maximum range for 20dB fade margin would be around 5-10 meters. Trying to do your 100 meters, plus go through a brick wall, through a window, and make two right angle turns, is not going to happen with such equipment. You'll need to look for something bigger than a key fob, that includes a proper receiver.

If you're planning on buying a solution and only need one pair of radios, I do have one suggestion that should work. Take a pair of cheap FRS radios. Hang a tone encoder on one and designate it as the transmitter. Setup the other with a matching tone decoder and designate it the receiver. Depending on FRS radio quality, you'll get about 300 meters range. Don't use CTCSS tones to trigger the door opener as it's prone to noise falsing.

I don't want to invest time and money in learning to do this from scratch, and wanted to warn you off before you spent more time on this. So I've ordered the stuff from Ebay, and if I can get it to work, fine, and if not, we'll continue to take turns going down to the lobby to let people in.

The receiver on the Ebay device has one adjustable part, which I assume is a variable capacitor used to change the frequency between 433 and 315 MHz. When you say retune, are you talking about tuning frequency or changing impedance and other such stuff?

I've found several Youtube videos showing people adding 1/4 wavelength antennas to this type of device. Here's one that uses the even cheaper devices that don't encode at all. In open air (literally out in the country), with antennas at each end, and adding his own encoding, he gets up to 300 meters:

However, back in Berlin he gets much less range, even with antennas. Interference, I assume. Anyway, I think it may still be worth a try to add antennas of 1/4 wavelength at both ends, or at least at the receiver if I want to continue to use the transmitter's built in telescoping antenna, and I can experiment with antenna length.

I'm not opposed to changing out other parts, but in this day of SMD stuff, that's usually not practical for me. I saw the schematics you mentioned in your other posts, and will compare them to what I actually get from Ebay.

Yes. And that reminds me that one other thing I need to test is to see if mounting the antenna parallel to the board works better than sticking up perpendicular to the board. All the examples on Youtube are parallel. But I guess with essentially no ground plane, it may not matter.

Thanks for the idea. We could fall back to that if the RF remote doesn't work.

Another possibility is running a wire part of the way. AIUI these meetings are after hours when the door is locked and no one else is there to open it, so a wire would bother no one.

The "capacitor" changed the tuning, but not of the front end parallel LC filter. If you look carefully at the capacitor, it's actually a parallel coil and capacitor combination in one package. It tunes the amplifier stage between the front end and the demodulator or decoder chip. When you add an antenna, this LC combination will NOT change its tuning. Do not try to adjust it unless you the proper test equipment. What will change when you add an antenna is the input LC filter, which has no tuning capacitor. That's the coil in the upper right of the above URL. You get to spread the turns of the coil to compensate for the added capacitance.

Gone to town for lunch, oil for the chainsaws, a replacement gas filter, cardboard boxes for the kindling, and a quick computah fix. Switching my brain between chain sawing and Usenet is more difficult that I originally thought. More comments, if I survive the afternoon.

No, and I'm sorry I didn't make that more clear. What I tried to reference was the standard garage door receiver that is hanging from the ceiling in many peoples garage.

I have a Crawford Model 41A3039, I assume this is the same as a Craftsman, because I use two Craftsman remotes, I picked up somewhere. The wire sticking out is about 6" long.

Mikek

Not true. They are available today.

And I'm not going to do your searching for you.

Sniff.

Egads, you're right. There are super-regenerative receivers.

I found one by Chamberlain from 2001. Probably used in their garage door or motorized gate openers: Schematic: I would expect to see an adjustment that sets the receiver at the bitter edge of oscillation, but I don't see the pot. Q1 looks like an oscillator so I guess it's regenerative. 54 test points. No tuning in the front end and only one LC adjustment for receive frequency and regeneration. It passed FCC certification as HBW1614.

Also: which looks the same except for the operating frequency range.

Here's a collection of super-regenerative modules that claim to be FCC and EU certified. Oddly, I don't see an FCC ID number or DoC (Declaration of Confomity) for the receivers.

Googling the XD-RF-5V receiver for super-regenerative, I find this: which says that it's a super-regen receiver. Oddly, when I tested it on the bench, I didn't see anything that looked like super-regeneration, but then I wasn't looking for it. I'll try agin this weekend, after I recover from chain sawing.

Grumble. Now you know why I keep asking for clarification.

It's a Chamberlain Liftmaster 41A3039. According to this chart: it was made between 1986 and 1989. I couldn't find a schematic so no guess(tm) on antenna impedance.

In fact, the XY-DJM-5V receiver in my Ebay remote set is also super-regenerative, at least according to Aliexpress:

But that would be expected I think since the front end of that one is probably the same as your XD-RF-5V. The same except that one has the decoder, but the other doesn't.

One more time. Usually making a tuned ( 1/4 wave length ) antenna for a receiver is wasted effort.

One thing you can try is connecting a wire to the receiver and running it o ut the the window of the lobby , along the outside of the building , in th e window and end it close to the transmitter. Think of it as having the re ceiver antenna very close to the transmitter antenna.

Dan AD7PI

They're probably both the same. I compared the construction between my XD-RF-5V and the XY-DJM-5V and they look the same, except for the decoder chip.

Ok, so let's assume that the data sheets are correct. Let's see how well it plays. This page says the receiver sensitivity is -105dBm (without bothering to supply a corresponding BER (bit error rate). The eBay listing is slightly less insane at -90dBm: I'll just take it at face value and try it on the bench later (after I clean off the mess this weekend). Incidentally, in looking around at other sources for this receivers, I'm finding an amazing variety of numbers for sensitivity, which suggests that the numbers might be cooked for the occasion (much like lumens output on LED flashlights).

The associated transmitter seems a bit difficult to identify so I'll use the one that I have: That says +16dBm transmit power, which seems high. Lacking a better number, I guess I'll use it.

Antenna gains are a bit of a problem because of matching problems and the lack of a counterpoise. If it were a proper dipole, I would give it +2dBi gain. For a monopole above an uncontrolled ground, my guess(tm) is maybe -2dBi gain. I forgot where I stole the building materials (brick) attenuation:

Starting at the transmitter: TX power +16 dBm TX coax loss 0 TX ant gain -2 dBi Distance ???? Brick losses -5.5dB (for 2 brick walls) RX ant gain -2 dBi RX coax loss 0 RX sens -90 dBm Fade margin 20 dB and ending at the receiver end. All these numbers are my best guess.

Adding everything up, I get -63.5dB path loss. Plugging iteratively into: at 315MHz and 0.07 miles, I get the required -63.5dB path loss.

It might work, barely. I suggest you consider the FRS radio contingency plan, figure out a way to install a reflector, run a wire down the hallway or out the window for the duration of the meeting, or try something using power line networking (X10 or Insteon):

It was a mistake for me to use antenna theory in the first place. The radio control guys say that a given receiver is tuned (base loaded) to use a predetermined antenna length. They say that even a 1 mm error in length will degrade receiver performance. Too short is just as bad as too long.

My receiver is a 434 MHz super-regenerative circuit that uses SMD technology. It's predecessor used through hole components:

This page contains useful information:

Thank you,

Do you think that 1mm figure is right? How do the RC guys measure it? I don't think they can measure signal strength down to 1 mm.

What about antenna placement wrt local metal construction, and random wire grounds?

Then you have the metal door. Isn't that going to block some of the signal?

A local surgeon came to me with a problem. Both his garage door openers had quit working. He said if I couldn't fix them, he was in for a $1200 bill to replace the system.

The problem was obvious as soon as you opened the case. The leads to the 9V battery were too short and had broken. I installed new battery clips with longer leads and new batteries and gave them back. I didn't charge him for the repair. I hope he will remember me if I have to go in for surgery:)

But these things are built cheap. Real cheap. There is probably plenty of signal lost both ways. Which is probably good. You don't want someone else's opener on the other side of town to open your doors.

That remark suggests a misunderstanding of how super-regenerative receivers actually work. The trick is that the start-up time of an oscillator is (comparatively) hugely affected by the injection of a bit of RF energy at the right frequency. So you make an oscillator stop and start repeatedly at some high rate and the averaged start-up time is the demodulated signal. You couple the oscillator to an antenna through a single-transistor RF stage to prevent it from radiating its own RF.

This can be amazingly simple and it doesn't need any critical adjustments.

Jeroen Belleman

Agreed. I think Jeff L was thinking of plain "regenerative" where the detector has to be set just on the verge of, or only very gently, oscillating. Not the version which repeatedly crosses the critical point at a supersonic rate. Super-regens are magic.

Not the first time this confusion has been made on SED.

piglet

Trying to hit a precise antenna length reminds me of trying to hit the maxima of a resonance curve. At some point you must fish or cut bait.

A table on page 7 of my Cushcraft vertical antenna manual [1] shows that antenna length can vary from 70' 3" to 61' 6" for the 80 meter band. My strategy is to set it at the mid-point and suffer any degradation at the extremes. The antenna still works throughout the entire 80 meter band, it just doesn't work as well, in theory.

Note.

1. formatting link

Thank you,