RFI Mitigation

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Actually a 100mw tx is more than capable of causing rfi if it's output is close to other electronics.
I run my 440Mhz trackers down around 10 - 12 mw output which is more than enough output to give me a great range, and have never had an issue with rfi bothering the altimeters.
Can the Eggfinder be run at a lower output? That would be a lot easier of a fix than trying to shield your altimeters...
Greg

The Eggfinder is fixed at 100 mW. I have not heard any reports at all of Eggfinders interfering with altimeters, if anyone has had any adverse experiences I'd like to hear about it.
 
There were reports that GoPro and Mobius cameras writing to their SD cards were causing so much noise that they swamped the GPS receivers on some quadcopters. I'm sure the square wave noise from logic busses is more harmful than the analog output of the RF downlinks we're using in rocketry.
 
Interesting, I would think putting a GPS receiver inside a faraday cage would block GPS signals to the receiver. Shows I really don't understand RF reception and antenna theory...
 
I have never had any issues with Eggfinders interfering with other components, simply because I have always mounted them in the N/C by themselves. Since this new avbay is so big, and there is no all thread anywhere near it, I will probably just build it and check to make sure everything works. The sled with the Eggfinder will be at least 6" away from the sled with the altimeters, and I suppose I could mount the components so they are on opposite sides, that would put two sheets of 1/4" plywood between them. All the Eggfinders I have built acquire lock in 2 or 3 minutes, even on their initial power up, so if that doesn't happen, I might have to re-think the configuration.

 
BTW: In an email thread I started about GPS lock times BDale stated that the antenna on the TM isn't tuned as well as the TeleMega. He said that the ground plane in the board was too small for the antenna and it's "below the knee", meaning that it isn't as sensitive as a properly tuned antenna. So you have to do everything in your power to not detune it.

I'd try moving the Telemetrum away from the all thread any way I could.
 
Interesting, I would think putting a GPS receiver inside a faraday cage would block GPS signals to the receiver. Shows I really don't understand RF reception and antenna theory...

Chris put the camera in the cage to keep it from interfering with the GPS receiver--not the receiver itself...


Later!

--Coop
 
BTW: In an email thread I started about GPS lock times BDale stated that the antenna on the TM isn't tuned as well as the TeleMega. He said that the ground plane in the board was too small for the antenna and it's "below the knee", meaning that it isn't as sensitive as a properly tuned antenna. So you have to do everything in your power to not detune it.

I'd try moving the Telemetrum away from the all thread any way I could.


Good to know... in the future, I'll look into allthread-less solutions when using the Telemetrum.


Later!

--Coop
 
The u-blox GPS receiver used in the Telemetrum has a claimed time to first fix of 30 seconds for a cold start. If you are having times much greater than this, particularly if the receiver is outside, there is a problem. (30 seconds is the worst case time to receive the ephemeris data after locking onto the signal.)

Looking at the images of the altimeter I find a few areas of concern.

The first is the through hole connection between the GPS antenna and the GPS receiver is between the GPS receiver and micro-controller. A really good location if your goal is to provide maximum coupling to the hash from the micro-controller.

The second is the close proximity to the 70cm transceiver chip. Again, a good location if you want to maximize coupling.

If there is any chance at all of RFI there should be a bandpass filter between the antenna and receiver. SAW filters are cheap and available and since there are sources of RFI on the same PCB, one should be used. Without one it is very easy for a signal to swamp the RF AGC. Even a signal far from 1.575 GHz. The data sheet for the CC1111 (I think that is the correct part) specifies the third harmonic of 433Mhz (1.299GHz) is at -42dBm.

Then there is the complete lack of shielding.


The Telemetrum design all by itself might be raising the noise floor (reducing performance) and the addition of another device nearby might be all that it takes to render GPS completely non-functional.
 
This one started out as just a TeleMetrum in the NC, then I decided to add the camera, then I decided to add the RRC2+ as a backup. :rolleyes:

I was doing some testing in my backyard in the shade of a tree and the times were terrible. I moved the TM about 6-8 feet away and my times came back down to 3-8 minutes.

With the unshielded camera and RRC2 powered up it took over 10 minutes if it ever got a lock. The shielding brought the times back into the same range as when the camera was off, so it seemed to work.

Not that this is directly related, but I'm kinda surprised how susceptible these units are... I have mounted my Entacore AIM XTRA (similar to telemega) in a couple different locations far worse than those listed here and had no problems. Mine is currently in a 3'' electronics bay stuffed with aluminum all-thread and Li-po's directly below the unit, a backup altimeter next to it, and a mobius HD camera a few inches above it. With all this, even inside my house I get standard lock, 4 satellites, within 2 minutes, sometimes faster, and after 5 min or so peak out at 8-9. Outside at the launch, I get lock in ~1 min, and after ~5 min its peaked out at 16!
 
Here's a stupid question.
Can you replace the metal all thread with a oak dowel rod and thread the ends with a die?
You can still use metal nuts and washers.
Being a wood worker, it's just a thought.


Not stupid at all. I'd wondered if another material could be reasonably used...


Later!

--Coop
 
The u-blox GPS receiver used in the Telemetrum has a claimed time to first fix of 30 seconds for a cold start. If you are having times much greater than this, particularly if the receiver is outside, there is a problem. (30 seconds is the worst case time to receive the ephemeris data after locking onto the signal.)

The Telemetrum design all by itself might be raising the noise floor (reducing performance) and the addition of another device nearby might be all that it takes to render GPS completely non-functional.


I've never seen 30 seconds on a good day all by itself. You're admittedly over my head with most of your post. Is there anything in there I, as an end-user, could do?


Later!


--Coop
 
I've never seen 30 seconds on a good day all by itself.

Which is a sign that the design of the Telemetrum is at least partly at fault.

You're admittedly over my head with most of your post. Is there anything in there I, as an end-user, could do?

There is nothing you can do about the design of the Telemtrum. One thing that you could try to choose a telemetry frequency that keeps the harmonics as far away from 1.575GHz as possible. The problem here is that the GPS signal lies between the 3rd and 4th harmonics. 450Mhz makes them equidistant. That is at the top of the 70cm band so some caution should be used there and be sure to check band plans.
 
Which is a sign that the design of the Telemetrum is at least partly at fault.



There is nothing you can do about the design of the Telemtrum. One thing that you could try to choose a telemetry frequency that keeps the harmonics as far away from 1.575GHz as possible. The problem here is that the GPS signal lies between the 3rd and 4th harmonics. 450Mhz makes them equidistant. That is at the top of the 70cm band so some caution should be used there and be sure to check band plans.

Thanks! I will see what I can do about the frequency... I never would have thought of that, to be honest.


Later!

--Coop
 
RFI abatement comes up every now and then. Yes it does bite people and I being a victim twice. One thing is to ask around what works with what.
2 watt dog trackers and Adept 22's don't play well. The Adepts shutdown. My experience a P6K (I think OOP) likely blew charges on ascent with a 150mw tracker and then with a rebuild sustainer, a 15mW tracker blew the charges while on the pad.
This was with trackers riding in the same ebay. Boy oh boy do I ground test now.

The major problem is if the tracker has to ride in the same ebay as the deployment electronics. I've had flawless performance with the Ravens riding right next to the low powered 70cm Beeline GPS trackers. BLGPS on top and Raven sitting aft in the bay.

Twisted wires = good practice.

Deployment electronics with opto-isolators may make a very big difference. I tested out an EggTimer Qwark with 5 watts of 2 meter, MURS and 70cm frequencies over 15 minutes time each. Had the Qwark wired up to naked ematches and was squawking
the "ready to go" beep. Only thing I could get the device to do was a change in tone of the beeping sound. That was with the tip of antenna actually touching the white opto-isolator chip on the board blasting away with pulses of 5 watts Rf. Pretty extreme
mind you.

The device kept up the "ready to go" beeping throughout the test, no popping of matches, no shutdown or reset. Very encouraging if you ask me. The EggTimer flight computer is of similar design but I haven't tested one like this. I suspect it may behave in the same manner.

From what I gather from others, to test an installation with a tracker and deployment electronics, Connect everything up with contained bare ematches. Turn everything on, stand up the rocket and let it sit for 30 to 60 minutes. No match popping, no altimeter resets or shutdowns and you are likely going to have an excellent chance that the flight will not be interrupted by stray Rf. Do not assume that sticking the tracker in the nosecone is going to alleviate the problem unless you test it out

Like I said, ask around. That's a good start. Then test to satisfy yourself. Better that than this: https://www.tripolipeoria95.com/uploads/OMG5.jpg

One comment about the issues with the GPS position acquisition. Have seen some comments in the Altus group. One thing I've noted with the BLGPS and I have some old ones, is if the keep alive battery goes dead man oh man it takes a loooooong time
to get an initial lock even out in the open. Also if the unit has been off for a long time it may take longer. For that reason I replace the battery for utility when need be.
I've had no issues with the Altus Tele-GPS but Bdale had mentioned that unit seems to lock faster that the others for some reason. Kurt
 
I've had no issues with the Altus Tele-GPS but Bdale had mentioned that unit seems to lock faster that the others for some reason. Kurt

It has enough channels of correlation that it can assign one to every satellite. It can then lock onto the signal for the ones that are in view quickly and then the main driver for getting a position is capturing the ephemeris data which can take up to 30 seconds.

The previous generation didn't have that luxury so they would pick 12 (if that's how many correlators it had) and hope to find at least one. Once one was found it could take 15 minutes to capture the almanac data which gives the rough position of every satellite. Then it could figure out what was in view. This made cold starts slow.

The first generation had one correlator that had to be shared with everyone.
 
RFI abatement comes up every now and then. Yes it does bite people and I being a victim twice. One thing is to ask around what works with what.
2 watt dog trackers and Adept 22's don't play well. The Adepts shutdown. My experience a P6K (I think OOP) likely blew charges on ascent with a 150mw tracker and then with a rebuild sustainer, a 15mW tracker blew the charges while on the pad.
This was with trackers riding in the same ebay. Boy oh boy do I ground test now.

The major problem is if the tracker has to ride in the same ebay as the deployment electronics. I've had flawless performance with the Ravens riding right next to the low powered 70cm Beeline GPS trackers. BLGPS on top and Raven sitting aft in the bay.

Twisted wires = good practice.

Deployment electronics with opto-isolators may make a very big difference. I tested out an EggTimer Qwark with 5 watts of 2 meter, MURS and 70cm frequencies over 15 minutes time each. Had the Qwark wired up to naked ematches and was squawking
the "ready to go" beep. Only thing I could get the device to do was a change in tone of the beeping sound. That was with the tip of antenna actually touching the white opto-isolator chip on the board blasting away with pulses of 5 watts Rf. Pretty extreme
mind you.

The device kept up the "ready to go" beeping throughout the test, no popping of matches, no shutdown or reset. Very encouraging if you ask me. The EggTimer flight computer is of similar design but I haven't tested one like this. I suspect it may behave in the same manner.

From what I gather from others, to test an installation with a tracker and deployment electronics, Connect everything up with contained bare ematches. Turn everything on, stand up the rocket and let it sit for 30 to 60 minutes. No match popping, no altimeter resets or shutdowns and you are likely going to have an excellent chance that the flight will not be interrupted by stray Rf. Do not assume that sticking the tracker in the nosecone is going to alleviate the problem unless you test it out

Like I said, ask around. That's a good start. Then test to satisfy yourself. Better that than this: https://www.tripolipeoria95.com/uploads/OMG5.jpg

One comment about the issues with the GPS position acquisition. Have seen some comments in the Altus group. One thing I've noted with the BLGPS and I have some old ones, is if the keep alive battery goes dead man oh man it takes a loooooong time
to get an initial lock even out in the open. Also if the unit has been off for a long time it may take longer. For that reason I replace the battery for utility when need be.
I've had no issues with the Altus Tele-GPS but Bdale had mentioned that unit seems to lock faster that the others for some reason. Kurt

Thanks for the input --I plan on testing a few more things (the transmit frequency, for one, and still want to see what happens if I bridge the allthreads together with and without the chokes). I'm rather glad I decided to nerd out on this --because by the time I'm done, I should have a fairly good line on what to expect at the flight line.


Later!

--Coop
 
There were reports that GoPro and Mobius cameras writing to their SD cards were causing so much noise that they swamped the GPS receivers on some quadcopters. I'm sure the square wave noise from logic busses is more harmful than the analog output of the RF downlinks we're using in rocketry.

Thanks for mentioning that. I've not used that type of camera and it's good to know if mounting one of these devices inside a carrier bay. Keychain, 808 jobs I haven't heard of any issues yet though most folks stick them on the outside. Kurt
 
I looked at the schematic for the Telemetrum today and it is a good news, bad news sort of thing.

The good news is that it does have a SAW bandpass filter between the antenna and GPS receiver. I missed it in the photo because it is much smaller than I expected. This improves immunity to external RFI.

The bad news is that the GPS receiver shares power and ground with the transceiver with no filtering. It would be interesting to see the cold start performance of the GPS if initialization of the frequency synthesizer was delayed until the GPS locked.
 
im no RF expert but how about a metal sleeve on the all-thread
to eliminate all those angle reflectors?

john
 
Not stupid at all. I'd wondered if another material could be reasonably used...


Later!

--Coop

I was thinking, why not use a styrene tube with short pieces of allthread glued into the ends ?
 
I was thinking, why not use a styrene tube with short pieces of allthread glued into the ends ?

Would styrene have the strength necessary? I think that's really what it comes to...


Later!

--Coop
 
im no RF expert but how about a metal sleeve on the all-thread
to eliminate all those angle reflectors?

john
Metal is the problem. You need an insulator there.

I was thinking, why not use a styrene tube with short pieces of allthread glued into the ends ?
It would be much simpler to take a plastic or fiberglass rod and tap the ends to take a screw from the outside.

Bob
 
Metal is the problem. You need an insulator there.


It would be much simpler to take a plastic or fiberglass rod and tap the ends to take a screw from the outside.

Bob

The other alternative as far as getting the "Rf" out of the ebay is to put a bulkhead mount connector on the aft bulkhead (https://www.ebay.com/itm/4-SMA-male...832?pt=LH_DefaultDomain_0&hash=item2598046d00) and either screwing the antenna there or using a short length of cable to connect to the antenna. https://www.ebay.com/itm/8inch-RG58...606?pt=LH_DefaultDomain_0&hash=item27d03e6916

Can zip tie it to the sustainer shockcord so the antenna gets deployed at apogee.

If you need RP-SMA type in your desired connector and search.

In this fashion, the antenna gets deployed at apogee for optimal transmitting on the way down where it's needed for recovery purposes.

This won't of course help with GPS signal reception, but the all-thread is generally not the problem there usually. In fact, when I had trouble with a 70cm GPS tracker installed inside of a metallic painted rocket, the issue was getting the
Rf "out" of the enclosed ebay. I didn't do a "good enough" range check and had one packet received at altitude but it came down luckily within sight. The Beeline GPS tracker had seven to ten satellites locked during most of the flight as downloaded to memory. So Rf couldn't get "out" but the GPS signals could get "in".

I didn't want to repaint the large rocket so I used a cable interconnect like above and used an aft bulkhead mount. I was able to find a flexible antenna for the 70cm band similar to this one: https://www.ebay.com/itm/6-9-Golden...3453cfd&pid=100033&rk=1&rkt=4&sd=141714411052

The whole thing gets deployed at apogee for the ride down.

OK, here's the "type" of antenna I use on the aft bulkhead of an ebay on a standard DD rocket: https://www.ebay.com/itm/Golden-Sin...975?pt=LH_DefaultDomain_0&hash=item27e5d8470f ..................except in 2014 I was able to get one with an SMA base. If one has room on the bulkhead just use an SMA to BNC bulkhead mount and that will work.
The spring antenna may better survive a landing and gets deployed at apogee for the ride down. Kurt
 
Last edited:
Don't know if this has been thrown out yet, but I would suggest making sure that none of the wires in your AV bay are resonant at your RF devices operating frequency. That is, make sure that the length of the wires aren't 1/4L, or any multiple thereof. I would also reccomend checking out the RFI forum over at eham.net. There's all kinds of stories and suggestions on how to deal with RFI.
 
Back
Top