Jump to content
Sign in to follow this  
Wavess

Airwave rx634 Issues

Recommended Posts

Hi all, I´ve recently made a pcb board for my airwave 634rx. Something weird is happening, I made the pcb according to airwave datasheet, including rf width. When I power on my rx, there is a intermitent but continous noise in the audio signal that sounds like, "toc toc toc" :rolleyes: . But the weirest thing is on the video signal, sometimes suddenly lost it at all, with a black screen. I attach a picture of my pcb´s to help with the description of the problem. There is a white horizontal line crossing the screen downwards only sometimes too. (ground loop problem here?)

Do you guys have some thoughts of what´s happening¿

18615515178.jpg

Share this post


Link to post
Share on other sites

Not that this is specifically a AWM634 issue, but some folks have noticed noise on the audio when some types of OSD's are used. So, be sure to perform your tests with a typical CCD camera that does not have bright text data overlayed on the video signal.

I made the pcb according to airwave datasheet, including rf width.

From what I can see, the RF feedline's copper trace does not seem to be correct. It appears a bit skinny for a single sided board. Even with a thin (0.8mm) FR4 dielectric and an opposing ground plane, the trace width will be a bit fat (somewhere around 60 mils).

To avoid significant RF matching losses, it might be safest to eliminate the copper trace and solder the SMA directly to the RF output pin and its surrounding gnd pins. To put this in perspective, my gut estimate is that you will lose ~1/2 of your RF signal with what I see in the photo (I could be wrong though). It would be a good idea to double check your math (just ignore my comments if it confirms you are OK).

Share this post


Link to post
Share on other sites

The rf path´s got a 75 mils track. This is what airwaves recommends for a double sided pcb. What do you think it would be proper for single side pcb? Do you know some website impedance calculator?

Edited by Wavess

Share this post


Link to post
Share on other sites
What do you think it would be proper for single side pcb?

From my experience, it cannot be done with a a single sided PCB, at least not without bonding a copper groundplane to the opposing side. With single sided PCB's, the trace is effectively a primitive antenna rather than a controlled impedance feed line.

I doubt this issue related to your audio noise. But, it will cause a loss of RF signal to the antenna. So, once you get the other issues fixed it would be good to optimize the antenna path.

Share this post


Link to post
Share on other sites

From my experience, it cannot be done with a a single sided PCB, at least not without bonding a copper groundplane to the opposing side. With single sided PCB's, the trace is effectively a primitive antenna rather than a controlled impedance feed line.

I doubt this issue related to your audio noise. But, it will cause a loss of RF signal to the antenna. So, once you get the other issues fixed it would be good to optimize the antenna path.

Ok, I´ll go step by step here, first I´ll try with a convencional ccd camera without an osd feeding, and then would try to fix my pcb. I really like to put the sma conector on the pcb, and if I have to go with a 2 layer pcb I´ll do it. Kind of disgusting all the time I´ve inverted in doing something that doesn´t work well, but that´s the way rf goes :lol: Anybody succesfully mounted airwave modules with a single layer pcb?

Share this post


Link to post
Share on other sites
Anybody succesfully mounted airwave modules with a single layer pcb?

If you mount your SMA on the bottom (solder side) of the board, directly to the module's pins, then your existing layout can be used. Another method would be to use a short RG-174 coax feedline to the SMA (replacing the existing copper trace), but coax is easy to bugger up too.

Share this post


Link to post
Share on other sites

Ummm, after studying this a little bit, you are so correct, to match the required impedance for a pcb, it´s kind of hard for those of us that don´t order them and buy at our local electronic store one side positive pcb´s. Something curious is this,

http://www.emclab.umr.edu/pcbtlc/microstrip.html

putting down what airwave states to be 50 ohm trace (assuming 4.3 Er) give us a result of 48.81 ohms, but with 72 mils trace width give us 50.2 ohms, closer than airwave´s 75 mils. Now how could I now the "H" and "T" dimmension for my custom board, since I don´t a pcb datasheet to know it?

Share this post


Link to post
Share on other sites

The H dimension is extremely critical. You can use micrometers to measure yours (it is the thickness of the non-conductive dielectric). Also, don't guess at the Er value; it is important too. Without a PCB data sheet your results will not be predictable.

Sometimes it is difficult for a hobby etched board to provide reliable microstrip results. The PCB material specs are often not provided at the local supply store and the etching processes are not well controlled. That is why skipping this and using a directly soldered SMA is so much simpler. But, what you are doing is not impossible, and decent RF performance can be achieved if you have ways to measure your results so that you can tweak the processes.

Share this post


Link to post
Share on other sites

Yes I have some ways to measure thickness with a micro meter, what I see harder is to try to measure relative permittivity of the dielectric. I´ve seen some methods like this one:

Dielectric calc method I guess I don´t have the necessary equipment for this.

And this one, Dielectric theorical calc

What do u think about the last one? Could it be a nice aprox?

When I solve all this, I´ll end up with an "optimized" trace width for my pcbs and will make a new one. Then I´ll post results to see how much the improvement is. For people like me here in Spain we call them "hard head".

I´ll be back :lol::lol:

Share this post


Link to post
Share on other sites

I have always used the specs from the mfg. They vary quite a bit, so I have no idea what a good guess would be for your Er. However, this piece of data is less critical than some of the others, so maybe a good educated guess might work out for you. I have no idea what you should use, so it is all up to your dart board. Just keep in mind that efficient Microstrip RF designs require critical layouts and well controlled PCB mfg processes.

I suggest that when you design another board that you place the Airwave module next to the SMA so that the feedline to the SMA is extremely short. As close to zero-length as you can get.

And I will try one last time: Solder the SMA connector directly to the Airwave module's signal and Gnd pins. With this method, you can even use the board you have already built. :)

Share this post


Link to post
Share on other sites

Some more news on this. Conected without an osd makes no difference in the audio & video problems. Conected directly to the tv there is no video signal lost at all, it´s much more stable than with my archos, and that kind of noise dissapears and seems to be noise from a poor rf signal. So I guess all my problems are rf signal. I´ll do one new pcb to attach the sma directly to the receiver, and the other one with the rf path as accurate as I can to match impedance.

Share this post


Link to post
Share on other sites

Some more guesses on the pcb´s matching impedance. I wrote yesterday night an email asking my pcb´s manufacturer for the data I had missing. Surprise!!! this earyly morning he replied to it giving me the data I needed. Putting down the values he gave me on the calculator, shows that my actual impedance for the rf line is 64,79 ohms :o:o , way to far from 50 ohms. To achive 50 ohms I need a 113 mils width rf path (way to far from what I´ve got 75). So I assume that making another pcb with the rf path fixed to that numbers, it´ll work well. Mr. RC-CAM, do you think macthing all this for a single layer pcb is worthless? In other words, is it absolutely necessary to make a 2 layer board? I only have a single layer insolator, and would be very hard to try insolating 2 layers with that stuff.

EDITED: It´s necessary in order to have an rf path thinner than 113 mils. With 113 mils the sma conector ground legs and the rf line, would touch each other, so has to be a lower value, that´s imposible to fit with a single layer pcb.

P.D. I´ll do one pcb with the conector directly soldered to the receiver as you said, and the other one the way I explaine above and will compare then. Kind of liking this easy math B)B);)

Edited by Wavess

Share this post


Link to post
Share on other sites
In other words, is it absolutely necessary to make a 2 layer board? I only have a single layer insolator, and would be very hard to try insolating 2 layers with that stuff.

I'm not at all of an RF wizard and Mr.RC-CAM will probably correct me, but from what I remember you don't need to actually draw anything on the backside. You just need to have it there as a ground plane (i.e connected to GND). You can just etch your top side keeping the bottom protected, and connect it to ground properly after that.

Share this post


Link to post
Share on other sites

yes the ground behind the microstrip is just as important !

we always add alot of vias to make ground as good as possible,

at 2.4GHz it is close to impossible to do uptimal on home made boards,

I would go for a SMA direct solution, or as short as possible track,

the width must be right, and much more connections from top gnd to bottom gnd,

every 5mm or more

Edited by ThomasScherrer

Share this post


Link to post
Share on other sites

I'm not at all of an RF wizard and Mr.RC-CAM will probably correct me, but from what I remember you don't need to actually draw anything on the backside. You just need to have it there as a ground plane (i.e connected to GND). You can just etch your top side keeping the bottom protected, and connect it to ground properly after that.

Yes that could be done, I didn´t think about it before you´ve mentioned it :lol: . I´ve found a high quality pcb boards with lots of dimensions available (fr4 thickness, copper thickness, 2 positive lawyers and so on) would also make quite a few vias from layers. So let´s do some pcbs. :)

Share this post


Link to post
Share on other sites
To achieve 50 ohms I need a 113 mils width rf path (way to far from what I´ve got 75).

If you use a good quality 0.8mm FR4 material the trace width will be much smaller than 113 mils (should be closer to 75mils).

As suggested by Kilrah, the "backside" can be just unetched copper. However, as mentioned Thomas S, this groundplane must have extensive vias to the opposite side's RF groundplane. So, plan on drilling a several dozen holes to attach the two ground layers together.

Share this post


Link to post
Share on other sites

Yes, with 1 mm fr4 is around 64milis if I remember well.

I made a new pcb to solder directly the sma conector to the rf out pin and ground legs to ground neighbor pins. The audio is ok now, but with video signal is almost the same or even worst. The image jumps often and there are some white horizontal lines crossing the screen downwards (not always, but quite a few times). I´m doing all my tests at home and haven´t tried yet at an open field, but plane is 2 meters away from receiver. I´m getting mad with this stuff. :(

Share this post


Link to post
Share on other sites
I´m doing all my tests at home and haven´t tried yet at an open field, but plane is 2 meters away from receiver. I´m getting mad with this stuff. :(

Do try outside. You'll clear yourself from tons of multipathing and other interference sources. I have already had some equipment that didn't like the TX too close to the RX either, or it was overloading the latter. Required either going further away, or removing the TX antenna.

Share this post


Link to post
Share on other sites

I agree with Kilrah. Open air microwave RF tests from inside (or near) buildings are generally not very productive. Should be outside, in a large open area, far away from structures. Be wary of nearby wireless routers and cordless phones too.

Keep in mind that working with RF is often not as obvious as it seems. For sure, with microwave RF, everything matters. :)

Share this post


Link to post
Share on other sites

Something weird was that the first time I soldered the sma conector with a 10W solder, I needed quite a few heat to solder sma´s legs and didn´t seem to solder properly. I tried the receiver and was working quite better than before but there were some issues still. So I thought that soldering with a 30W solder the ground legs properly would work better, also I removed some lead from the rf soldered pin that I thought was excessive. I tried again and the thing was quite worst than the 10W soldering.

Share this post


Link to post
Share on other sites

With a 40W soldering iron and a wide soldering bit, at 340c, i have troubles with some types of SMA connectors. Some need an awful lot of heat before they will "stick"

Share this post


Link to post
Share on other sites

Be careful to remove the leftover solder flux from the connections. That's never good when it touches signal and ground.

Indeed, use an iron that is hot enough (mine is a temperature-controlled 60W one, I set it at the max 450° temp, it works fine even to solder on the case).

Share this post


Link to post
Share on other sites
Be careful to remove the leftover solder flux from the connections. That's never good when it touches signal and ground.

Indeed, use an iron that is hot enough (mine is a temperature-controlled 60W one, I set it at the max 450° temp, it works fine even to solder on the case).

I have a 30W JBC quality solder in 3 secs the sma´s ground legs are melting and the lead solders quite good. I do have to buy some solder flux, I´ve never used one, any recommending of it?

Yesterday night I removed the sma connector and did some dremel work with it. I cut down the center pin and also made it half the diameter thinner than factory. Also trimmed down the ground legs in order to have it after soldering with the same dimensions that had before. Worked quite well, here at home there are almost no significant av interference, but keeps doing some.

Today I went to my r/c club and tested the receiver, it worked great, and I could say that has even more range than my lawmate one. But not all are good news, now I´m having issues with my video splitter, when I turn on my glasses and my archos, the image in the glasses are worst than conected directly to the receiver, and archos records on black & white. I tried adding a 75 ohms resistor to the receiver without success. With lawmate I don´t have this problems at all.

The receiver is intended to be used as second receiver for Oracle diversity, so I guess I would not have this problems with it, but is used alone I can´t record my flights while doing fpv.

Also did some testing with Spectrum DX 7, three folks turn on their transmitter and receivers at the same time, they put the DX 7 really close to my receiver and noticed some av interferences when they move their sticks only but nothing too bad. They didn´t notice a problem at all with their planes close to my av transmitter. I guess that in the air I could cause them more problems than the dx7 transmitter to my av receiver ;)

Share this post


Link to post
Share on other sites
I do have to buy some solder flux, I´ve never used one, any recommending of it?

There's already some in any good electronic solder wire.

Do you have the coupling cap in the video signal output path on your receiver?

If video is crappy even through the splitter it will most likely be the same with Oracle. The splitter (I assume you've got the usual metal cased one?) already does everything nicely so should already be able to restore biased or unterminated signals. If not it must really be pretty wrong.

I guess that in the air I could cause them more problems than the dx7 transmitter to my av receiver ;)

Actually no.

The farther you fly the lower your video signal will be, while the pilots will still be at the same distance. So the interference level will be higher with regard to your video signal. Be sure to install yourself a few meters away from them, and avoid pointing an antenna in their direction.

Share this post


Link to post
Share on other sites
Do you have the coupling cap in the video signal output path on your receiver?

If video is crappy even through the splitter it will most likely be the same with Oracle.

Yes I do have a 470 uF cap on the video sginal path.

I´ve been using this splitter almost a year now without any problems with my lawmate receiver&glasses and is not the classic metallic one, it only has 2 outputs. :rolleyes:

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×