3505 views|23 replies

85

Posts

0

Resources
The OP
 

Why does this affect the analog signal? [Copy link]

 
I am making a GUI board with steering detection. The video signal is an analog signal. However, when turning, the square wave signal generated by the car will affect the board and the video will be disturbed. The steering detection circuit used to use a voltage regulator solution, but the effect was very poor. There were many horizontal stripes on the screen. Now it has been changed to the following figure, using the optocoupler isolation method. It is basically normal under normal light, but the screen image will still fluctuate in dark light (after the camera infrared is turned on). The reason for the jitter in the actual measurement is that a spike signal appears on the GND. I added an inductor to isolate it, but the effect is not good. I would like to ask for a solution. Thank you!

无标题.png (10.62 KB, downloads: 0)

光耦检测电路

光耦检测电路
This post is from PCB Design

Latest reply

Since I want to solve the problem, of course I have to correct the wrong design first.  Details Published on 2018-8-19 19:45
 

374

Posts

3

Resources
2
 
It should be the spark interference of the engine coil, and the circuit board needs to be shielded
This post is from PCB Design

Comments

No, it is the interference caused by the square wave signal generated when the steering is triggered.  Details Published on 2018-8-15 11:42
 
 

85

Posts

0

Resources
3
 
Is it the screen or the camera that is being interfered with? We need to confirm. There are many places on the ground that are connected. We need to add comprehensive isolation.
This post is from PCB Design

Comments

It should be that the camera signal received interference when it was processed by the board. The isolation is the inductor added in the picture, which seems to have no effect. . .  Details Published on 2018-8-15 11:44
 
 
 

85

Posts

0

Resources
4
 
wenyangzeng posted on 2018-8-15 10:53 It should be the spark interference of the engine coil, and the circuit board needs to be shielded
No, it is the square wave signal generated when the steering is triggered that causes interference
This post is from PCB Design
 
 
 

85

Posts

0

Resources
5
 
szy123 posted on 2018-8-15 11:13 Is it the screen or the camera that is interfered with? You need to confirm it. There are many places on the ground that are connected. Add isolation to all areas
It should be that the camera signal is interfered with when it is processed by the board. The isolation is the inductor added in the picture, and it seems to have no effect. . .
This post is from PCB Design

Comments

Are you sure the image is normal when there is no turn signal?  Details Published on 2018-8-15 11:49
 
 
 

85

Posts

0

Resources
6
 
liuzhiying666 posted on 2018-8-15 11:44 It should be that the camera signal received interference when it was processed by the board. The isolation is the inductor added in the picture, which seems to have no effect. . .
Are you sure that the image is normal when there is no turn signal? In the dark field, is there something wrong with the camera image itself?
This post is from PCB Design

Comments

No problem, I've tried all kinds of situations.  Details Published on 2018-8-15 15:00
 
 
 

9703

Posts

24

Resources
7
 
Check the power supply.
This post is from PCB Design

Comments

It has little to do with the power supply. Now it is the external trigger that interferes with the internal circuit.  Details Published on 2018-8-15 15:01
 
 
 

85

Posts

0

Resources
8
 
szy123 posted on 2018-8-15 11:49 Are you sure the image is normal when there is no turn signal? In dark scenes, will the camera image itself have problems?
No problem, I have tested various situations
This post is from PCB Design
 
 
 

85

Posts

0

Resources
9
 
It has nothing to do with the power supply. Now it is the external trigger that interferes with the internal circuit
This post is from PCB Design
 
 
 

1048

Posts

1

Resources
10
 
There is a spike signal on GND, and an inductor is added to isolate it. 1. What reference level is the spike on GND relative to? 2. Inductor isolation, but it looks like the parameters of a magnetic bead on your diagram? For this kind of low-frequency interference, the effect of using a magnetic bead for isolation should not be good. In fact, the biggest possibility is still the layout, the layout and routing of components to prevent interference current from flowing through the circuit.
This post is from PCB Design

Comments

Oh, thanks, I'll try the inductor.  Details Published on 2018-8-16 07:55
 
 
 

85

Posts

0

Resources
11
 
This post was last edited by liuzhiying666 on 2018-8-16 07:59
topwon posted on 2018-8-15 15:10 A spike signal appeared on GND, and an inductor was added to isolate it------- 1. Which reference level is the spike on GND relative to? 2. Inductor isolation, but it depends on you...
Oh, thank you, then I will try the inductor, it is indeed wrong above, it should be a magnetic bead
This post is from PCB Design
 
 
 

1w

Posts

142

Resources
12
 
The magnetic beads in the original poster's circuit are useless. If the two sides of the optocoupler share the same ground, the optocoupler is also useless. The interference problem needs to be analyzed globally. The local part shown in the original poster's diagram is meaningless.
This post is from PCB Design

Comments

I didn't design the circuit, I just solved the problem.  Details Published on 2018-8-16 14:01
Personal signature上传了一些书籍资料,也许有你想要的:http://download.eeworld.com.cn/user/chunyang
 
 
 

85

Posts

0

Resources
13
 
chunyang posted on 2018-8-16 13:11 The magnetic beads in the original poster's circuit are useless. If the two sides of the optocoupler share the same ground, the optocoupler is also useless. The interference problem needs to be analyzed globally. The local part shown in the original poster's picture is meaningless...
The circuit was not designed by me. I just came here to solve the problem
This post is from PCB Design

Comments

Since we want to solve the problem, of course we must correct the wrong design first.  Details Published on 2018-8-19 19:45
 
 
 

1048

Posts

1

Resources
14
 
After the 12V trigger signal passes through the light-emitting diode of the optocoupler, do not connect it to the ground of the signal board (do not install the inductor), but pull it back separately, because the optocoupler only needs the current provided by the trigger signal. This can reduce interference between systems.
This post is from PCB Design

Comments

Then reversing requires two lines, but customers don't want that.  Details Published on 2018-8-17 09:49
Then reversing requires two lines, but customers don't want that.  Details Published on 2018-8-17 09:49
 
 
 

2w

Posts

341

Resources
15
 
If there is a spike signal on GND, it should be low-frequency interference. You can consider using common-mode filtering, such as common-mode inductors. They may be large in size, but you should first test the appropriate parameters to see the effect.
This post is from PCB Design

Comments

I found one that cost over 500, tried it, but it didn't work.  Details Published on 2018-8-17 09:50
 
 
 

85

Posts

0

Resources
16
 
topwon posted on 2018-8-16 15:48 After the 12V trigger signal passes through the light-emitting tube of the optocoupler, do not connect it to the ground of the signal board (do not even install the inductor), but pull the wire back separately, because the optocoupler only needs to trigger...
Then reversing requires two wires, and the customer is not willing to do that
This post is from PCB Design

Comments

You need to first confirm where the problem is, and then think of a solution. If ground separation can eliminate the interference, then there is at least one direction.  Details Published on 2018-8-17 14:20
 
 
 

85

Posts

0

Resources
17
 
topwon posted on 2018-8-16 15:48 After the 12V trigger signal passes through the light-emitting tube of the optocoupler, do not connect it to the ground of the signal board (do not even install the inductor), but pull the wire back separately, because the optocoupler only needs to trigger...
Then reversing requires two wires, and the customer is not willing to do that
This post is from PCB Design
 
 
 

85

Posts

0

Resources
18
 
qwqwqw2088 posted on 2018-8-16 16:18 A spike signal appears on GND, which should be low-frequency interference. You can consider using common-mode filtering, such as common-mode inductors. They may be large in size, but it is better to test them first...
I found one that costs more than 500 yuan, but it didn't work.
This post is from PCB Design
 
 
 

2866

Posts

4

Resources
19
 
This may be caused by the overall voltage distortion caused by the large current generated instantly. Test plan 1. First, disconnect the power supply of the entire system and supply power separately. 2. Put the system with separate power supply into the system for testing. 3. If there is a problem, just as "qwqwqw2088" said. 4. Add a large filter capacitor and a small decoupling capacitor at the power supply entrance. 5. If the problem persists, reduce the system power.
This post is from PCB Design

Comments

It is indeed possible. The instantaneous current of the car's flasher is more than 3A.  Details Published on 2018-8-21 17:26
 
 
 

1048

Posts

1

Resources
20
 
liuzhiying666 posted on 2018-8-17 09:49 Then reversing requires two wires, and customers don’t want that
You must first confirm where the problem is, and then think of a solution. If ground separation can eliminate this interference, then at least there is a direction.
This post is from PCB Design

Comments

Yeah, thank you.  Details Published on 2018-8-21 17:25
 
 
 

Guess Your Favourite
Just looking around
Find a datasheet?

EEWorld Datasheet Technical Support

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京B2-20211791 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号
快速回复 返回顶部 Return list