|Dec 11, 2012, 06:55 AM|
Should the ground be common?
Hello, I have made a video switch which has power from 2 different batteries.
5v (from rx) = ICs, status leds
12v (from aux battery) = cameras, vtx
I am getting weird noise in the video signal when I move sticks on my radio, I am thinking it could be caused by a bad ground? or conflict between the 2 batteries? is it possible?
What's weird is that sometime I don't have any noise, I think it depends of the order I power up things.
|Dec 11, 2012, 07:40 AM|
Yes, the grounds should all be tied together to give you a common reference and to help dump any stray signals to ground. However, it is important that each ground circuit be tied together only at one end of the leads. If connected in more than one place you can create what are called "Ground Loops" that can themselves become a source of noise.
There are a couple of questions that may help point to other ideas. Take note that most of what I am working toward is all interrelated.
First is that I am trying to see why you have two batteries in place of a 5-6V BEC from the one 12V source. This would eliminate the weight and complexity of a second battery, already common tie all the grounds AND, unless you need it to control some function on the camera, disconnect it from the rest of the radio system that can be acting as a source for spurious signals when you move the sticks.
One thing in that regard is to wonder if, along with the main aircraft radio system and the camera, any other leads are reconnected before powering up. I ask this since any extras, such as aileron leads/extensions, etc, can wear over time and not make good connections. Any resistance in a plug lead or slight break in the wire can create resistance and be the reason it is intermittent.
There are some very good switching BEC's available, but you will notice with a lot of them, (and ESC's) the signal or feed lead is curled around a ferrite ring before being plugged into the receiver or other electronics. You also see this technique used outside RC for a lot of electronic comm or power leads. You'll notice that as a strange "Bulge" near one or both ends, some covered with a plastic housing. This acts as a damper. The key is to have that ring as close to the end of the lead as possible to minimize any signals from reinserting into the lead.
If yours does not already have this installed you can add this as an experiment. It should not be harmful and the benefits can be worth the extra weight. The only concern is it takes a bit of extra lead length and may require an extension. If that's the case install the ring on the extension as close to the end where it is plugged into the device as possible.
NOTE: I save these as standard policy when dismantling old gear. If you are in a panic for one give me a PM and we can make some arrangements to get one to you.
The 12V input may already have an internal voltage control to drop it to what it requires for the electronics and drivers. This should be well designed and constructed as the camera has to work with it from the start. It is all the extraneous gear that can be causing you issues.
Finally, as an easy test, if the system does not need to be tied to the receiver, power the 5V portion with a second flight battery independent of the radio system. You may also reorient or relocate any antennas away from servo leads or extensions, main power leads and/or the two separate power inputs for the camera system itself.
|Dec 11, 2012, 07:54 AM|
Not a problem. Along with nitro and electric planes I also fly electric helis. Even ignoring the vibrations, these always seem the most difficult due to the number of RF sources, the close mounting of every device and the high current flows to the drive system. You learn a lot along the way.
|Dec 11, 2012, 06:05 PM|
Cougar, sorry for taking so long to respond.
To answer your first question, I am getting 5v from the servo cable connected to my circuit. This cable is also feeding the PWM signal I need for switching my video cameras. I am using this 5v source for ICs and some leds.
I am using 12v for the cameras and vtx, which is also connected to my OSD device (RVOSD) in the aux power port.
I think I will try to install a new ubec that would be connected to the 12v source and then regulated to 5v. That way, both power voltages will be from the same battery thus ensuring no ground loops right?
I will also try adding some ferrite rings just before connecting the servo cableto my circuit, didn't think about that.
I was also thinking about not connecting the ground wire from the rx servo cable, also ensuring no ground loop I think.
This plane is quite complicated and has a lot of electronic components, here is the build thread.
Basically there are 2 batteries on this plane:
Power system is 14.8v (main bat)
VTx, cameras and OSD is 12v (aux bat)
RX, servos are 4.8v regulated from 14.8v to 4.8v via ubec.
I am pretty sure its a ground loop problem... I will let you know about my findings here.
Thanks a lot for the help, appreciate it.
|Dec 11, 2012, 06:21 PM|
Rather than try to explain it to you, I'll just refer you to the Wikipedia article.
|Dec 12, 2012, 05:28 AM|
Miami Mike. In life I find and try to keep things a bit simpler than that article, but have experience with what we call "ground loop" problems in both automotive and industrial machine wiring apps. If you notice the pic in the top right each ground leg IS connected at each end. Disconnect the outer end of each leg, (or remove any resistive loading, such as from a worn or bad connector) and the potential difference is eliminated.
Anyway, I am trying to avoid going off on a tangent and help solve the original problem.
Metalfred, and I thought I put a lot of work into my aircraft! Yumpin Yimminy! Curious to see how those HK legs work out. I was looking at a set of those for my Harvard, (instead of the hacked up Robarts).
If you don't have any rings, give me a shout.
|Dec 12, 2012, 08:42 AM|
My main point was not your terminology, it was the fact that your advice was wrong. In general, components that need to be grounded can be grounded in as many places as you like with no ill effects, and in fact, it's generally true that the more grounds you have, the better off you are. In the case we're discussing, there's no reason whatsoever that "each ground circuit be tied together only at one end of the leads."
A ground loop can occur when two devices communicate with each other via signals of some sort, whether it's audio (i.e. a component sound system), video (i.e. a TV fed by cable), digital (i.e. a computer with external peripheral devices), and so on, and the two devices do not have a good common ground. When this happens, and the devices are at two different ground potentials, the connection between them that is meant only to carry the signal can also carry current caused by the difference in ground potentials, which can corrupt the signal, especially if the devices are powered by AC and connected to different grounded outlets. In audio systems this is often characterized by a 60 Hz hum.
Often the cure for a ground loop is to connect the device grounds together using the best and lowest impedance connections possible, which means that in many cases extra common ground connections can prevent a ground loop. But in no case that I can think of is it possible for extra common ground connections to cause a ground loop.
|Dec 12, 2012, 09:11 AM|
I guess we're both attacking this from different directions, but the common thread has to be the fact those grounds have to be at as low a resistance as possible. That is what I was trying to point out when recommending the grounds be only connected at one point. By that method there will be no potential differences regardless of any bad connections.
In automotive this has become even more critical as electronics invade and overwhelm their systems. A vehicle is a horrible environment, especially once corrosion starts its nefarious attack. On a personal note, I once watched a fluorescent trouble light completely fry the digital dash on a mid 80's Corvette just from RF injection into the interior. There was still a lot to learn about hardening the systems.
In the industrial reference I work with many different makes of CNC machines, and again this is not a very sublime condition with many high power AC and DC drive systems. Networking them sometimes gives us fits.
Added to that are the number of older and less efficient AC motors all tied to a common buss. I believe this is the reason the new Tx rules have a common ground point to balance out the phases.
Anyway, I hope some of this advice helps solve his noise issue.
One thing I forgot to mention and ask is if the interference is encountered only when a certain control is moved. If so then you may be able to trace which circuit is the problem. I have seen long and/or bad extensions cause system wide instability, as well as one bad or noisy servo.
|Dec 12, 2012, 10:24 AM|
Sorry, but you're not making any sense at all. Adding parallel connections cannot possibly increase net resistance, it can only decrease it. I guess I might as well drop this.
|Category||Thread||Thread Starter||Forum||Replies||Last Post|
|Discussion||Vbar gov, Hobbywing sensor, common ground?||stevehof||Castle Creations||1||Oct 06, 2012 12:43 AM|
|Discussion||Charge SLA (8v and 10v output, common ground)||-james-||Batteries and Chargers||2||Aug 01, 2012 06:46 PM|
|Discussion||About to be grounded - motor help (P-51, Trojan)||whichwaysup||Micro Ready-to-Fly||2||May 14, 2012 08:24 PM|
|Question||Commonly wearing Trex parts that should be changed and how often ?||Happymcc||Mini Helis||9||Oct 13, 2005 03:40 AM|
|Alert||AAAH! Nobody has it! But it should be so common!||Will403536||Micro Helis||6||Aug 13, 2004 10:12 PM|