MinimOSD dramaticaly decreases UHF RC range

[GoogleCodeExporter]

Steps to reproduce the problem

1. Connect MinimOSD to the APM in parallel to the Xbee or any other modem, 
following the instructions of the tutorial. Remember to do not connect the APM 
RX pin while using a parallel modem; 
2. Connect video input (camera) and output (video TX) to the MinimOSD;
3. Do not tie Analog and Digital VCC;
4. Power up all systems, prefer using two batteries, one for video + MinimOSD 
and another to APM + Servos + ESC + Motor. I'm using three separated batteries, 
one just for ESC + motor, but it's not necessary to reproduce this issue.
5. Put your RC transmitter in "range test mode". I'm using Dragon Link, so the 
procedure is to use that 50 Ohms knob on TX to limit irradiation.
6. Execute range test and take note from the results;
7. Now turn off video battery and redo range test.  

Expected output? What do you see instead?

The range decrease dramatically when MinimOSD is on. To be sure that the result 
is not compromised by video interference, you can also do the following test. 
To bypass MinimOSD connecting video TX at camera directly and power it up. Redo 
range test and you probably will see the issue.

Version of the product are you using? 

MinimOSD V1.1 with APM 1.0

Original issue reported on code.google.com by mpsou...@gmail.com on 27 Mar 2013 at 1:20

[GoogleCodeExporter]

Please correct me if I'm wrong, but the analog power line is regulated by a 
switching circuit, right?

If this assumption is true, the cause is found. This switching circuit may be 
generating a considerable amount of electromagnetic interference at 433 MHz, as 
GoPRO camera do in the same kind of UHF RC systems.

Some other theory?

Original comment by mpsou...@gmail.com on 27 Mar 2013 at 3:03

[GoogleCodeExporter]

I am having the same issue.

Setup is
 - TSLRS UHF link
 - Generic 12v camera
 - 1W 1.2GHz Lawmate video transmitter, 60 cm away from the rest of the electronics
 - APM 2.5 with MinimOSD 1.1
 - Separate 3 cells LiPo battery for video, powering camera, video transmitter,  MinimOSD and microphone
 - Analog and Digital VCC not bridged on MinimOSD

With recommended range test procedure (UHF emitter antenna removed) I get 20 
meters+ range. When I power the video link the UHF range drops to 1.5 meters.

If I connect the camera directly to the video transmitter I get full range back.

Isn't the camera side of the MinimOSD electrically disjoined from the APM side?

I'll try later to:
 - add a ferrite between the APM and the MinimOSD, or between camera and MinimOSD, see if there is any noise propagated from either the video transmitter or the camera
 - power the MinimOSD from a separate 5V linear regulator supplied from the 12V video battery to test the possible issue with MinimOSD onboard switching power supply.

Any other recommendation?

Original comment by yves.rig...@gmail.com on 7 Apr 2013 at 3:31

[GoogleCodeExporter]

While checking for the best place to connect an external 5v AVdd I found out 
that I have electrical continuity between AGND and DGND. I did not bridge those 
on the board, but I have zero electrical resistance between the ground solder 
pads. AVDD and DVDD are electrically disconnected.

Is this normal? That would defeat the purpose of separate power supplies, 
methink.

Original comment by yves.rig...@gmail.com on 7 Apr 2013 at 4:05

[GoogleCodeExporter]

Ok, another input: I added 12 turns ferrites near video in and video out of the 
MinimOSD, and range went from 1.5m to approximately 6 meters. Not good enough, 
but much better.

I am assuming that I have a noisy camera or video emitter, and that the noise 
is crossing to the UHF receiver through the AGND-DGND bridge.

2 questions, then:
 - Is it normal that AGND and DGND are bridged on my v1.1 board, or do I have a faulty PCB?
 - If it is faulty, where should I look to repair it (unbridge AGND and DGND)? I'm in EMEA and sending the board back for repair would cost too much, I think.

Original comment by yves.rig...@gmail.com on 7 Apr 2013 at 4:25

[GoogleCodeExporter]

I still do not know if AGND and DGND are supposed to be connected, but I assume 
this is a bad idea.

I added an optocoupler on the digital input side, powering the minimOSD digital 
side from the video battery. I still had range issue. I then cut the +12v input 
to the analog side and bridged AVDD to DVDD and AGND to DGND (voluntarily, this 
time).

I now have full range back (20 meters), exactly the same as when the camera is 
directly connected to the video transmitter.

My conclusion is that the noise radiated from the analog side switching power 
supply is interfering with the UHF receiver.

Anyway, I now have great range AND a great OSD, so I am happy.

Original comment by yves.rig...@gmail.com on 13 Apr 2013 at 4:19

[GoogleCodeExporter]

As I suspected! The step down regulator is messing everything.

Yves Riguet, could you please provide a wire skematic of your solution for 
everybody? 

Original comment by mpsou...@gmail.com on 14 Apr 2013 at 10:22

[GoogleCodeExporter]

Hi Marcos,

I attached the wiring diagram that allows me to power the minimOSD and still 
isolate it from the rest of the aircraft's electronics. I am using an 
optocoupler, a bidirectional one because that's what I had available - this is 
what I usually use for any data transmission to a power transmitter, for 
example to a UHF downlink (RFD900).

I am using a UHF-friendly SBEC (any good SBEC, really). You could use a linear 
BEC instead or make your own from a 7805 linear regulator.

Original comment by yves.rig...@gmail.com on 20 Apr 2013 at 6:06

Attachments:

• [External supply for minimOSD and isolation of inputs.png](https://storage.googleapis.com/google-code-attachments/arducam-osd/issue-54/comment-7/External supply for minimOSD and isolation of inputs.png)

[GoogleCodeExporter]

Hi Ives,
Thanks for the heads up on the optocoupler. Would you have used a 
unidirectional one if it was handy? Which one. Have you used this isolation 
method in other areas of the aircraft? Thanks.
Vik

Original comment by losthori...@gmail.com on 24 May 2013 at 12:50

[GoogleCodeExporter]

any update on this? has anyone successfully been able to use minimOSD with a 
UHF system?

Original comment by raphaelk...@gmail.com on 19 Sep 2013 at 2:45

[GoogleCodeExporter]

Very interesting the opto-coupled input solution. However, the noises on the 
OSD would rely on the quality of the filtering on the external uBec.
The interesting part is that even using an ordinary uBec you still get the 
benefit of mounting it away from the other parts, with the plus of a ferrite 
ring as LP filter on the dc wire.

I'm changing this thread to "Enhancement" type.

Original comment by sbeni...@gmail.com on 19 Sep 2013 at 10:05

• Changed state: Accepted
• Added labels: Type-Enhancement
• Removed labels: Type-Defect

[GoogleCodeExporter]

Yves,
Thank you so much for the workaround. Unfortunately I am not as technically 
inclined as you, and would really appreciate some clarifications, or even some 
photos of how you have isolated the minimosd. I am having horrible interference 
with my UHF system, crippling my range and basically making it no better than 
using 2.4 for RC.

If possible, could you contact me via email? raphaelkrengiel@gmail.com. I will 
make it worth your time!
Best,
Raphael

Original comment by raphaelk...@gmail.com on 17 Oct 2013 at 1:36

[GoogleCodeExporter]

I've found v1.1 produces a strong RF spike around 459MHZ when using both power 
rails or just the 5v supply.  Isolated the issue to the board by only 
connecting a 5v power and using different power sources.  No matter what power 
supply is used the interference is the same, would this be the on-board 
regulator?    How can we reduced the noise or change the frequency? 

Original comment by pox...@lim.com on 4 Nov 2013 at 11:42

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

If you want to bypass the on-board regulator and get rid of the noise, here's 
another way of doing it on a v1.1 board: 
http://www.rcgroups.com/forums/showthread.php?t=2175802

Original comment by nordin.r...@gmail.com on 29 May 2014 at 8:33