How do I prevent video interference caused by electric motors

Connecting a large capacitor across the output of a power supply will cause a brownout on any power supply. Connecting a heavy resistive or resistive/inductive load will not, if the power supply is adequate.

A brown-out indicates that the power supply (and regulator, and the circuit connecting all this to the load) isn’t capable of supporting the load being placed on it.

This is not necessarily so, as already indicated several times. If the power leads are long and
have too much inductance, then you get large swings in voltage - at the load - with switched
load currents. That’s what the info about the “Main Capacitor” was talking about.

Many good power supplies deal with this problem by running sense wires from the P/S straight
over to the load, so they can measure the voltage fluctuations and compensate. Negative
feedback stabilization.

@oric_dan(333)

Yea but that doesn't explain why the normal batteries have the same problem. I understand the use of the Main capacitor, to get rid of any voltage spikes or dips. But if the power supply doesn't have enough current to supply to everything as it is, then the capacitor won't have enough charge to do its job.

Now if he had access to an adjustable voltage supply, then he could rule out the power supply he is currently using. All he has to do is set the voltage to 12 volts and adjust the current, and keep monitoring it until he gets no brown outs or interference. Then get a proper power supply based on what he measured.

Yea but that doesn't explain why the normal batteries have the same problem. I understand the use of the Main capacitor, to get rid of any voltage spikes or dips. But if the power supply doesn't have enough current to supply to everything as it is, then the capacitor won't have enough charge to do its job.

Yeah, if the power supply can't supply enough current, that's certainly needs to be fixed. I thought OP was using a huge old PC supply.

Also, the business with the Main Capacitor is an illustrative point. It works best IF your controller is competently designed, and then helps deal with the battery leads, which may be longish for any #of reasons. OTOH, if everything in the ckt has long leads with nontrivial inductance, then the cap probably won't do much.

Good design is a systems-level solution. Would be interesting to know how auto manufacturers deal specifically with the brownout problem when the starter motor cranks. I imagine there's a lot of brownout protection cktry inside the computer box.

Yeah, if the power supply can’t supply enough current, that’s certainly needs to be fixed.
I thought OP was using a huge old PC supply.

No its like a laptop supply

I imagine there’s a lot of brownout protection cktry inside the computer box.

Im sure, but will they share it with us, probably not.

HazardsMind:

the power supply isn't the problem from what my experiments have shown. Even when using batteries, the same action still causes the camera brown-out.

Because you are not supplying enough current! The problem is you dont know what everything is rated, (what needs what)

The LCD is small. It's only about 1A. The camera is only about 1A. I don't know what the cap draws when it is added. The power supply is 90W/12V=7.5A Batteries on the other had can output a heck of a lot of amperage but the problem still persists.

This is a better image of the test system: It was suggested by Magician in another thread to put a inductor in series on the positive lead going to the camera. That makes sense to me as it would help smooth out current fluctuations caused by the sudden added load. What Henry value should I look for though?

BTW, I found these pretty pictures showing a lot of fun designs,

http://www.google.com/search?q=star+grounding&hl=en&tbo=u&tbm=isch&source=univ&biw=990&bih=824&sei=4k_rUOvZFaaZiQK1koGYBQ

David82: Batteries on the other had can output a heck of a lot of amperage but the problem still persists.

I don't know what type of battery you're using, but if you're talking about conventional AA sized cells they're going to struggle to provide a couple of Amps never mind having anything in reserve for spikes in the demand.

HazardsMind:

the power supply isn’t the problem from what my experiments have shown. Even when using batteries, the same action still causes the camera brown-out.

Your power supply can ONLY produce 12V @ 7.5A, whereas your componets may need 8 - 8.5A.
It worked correctly last time when you had 2 power supplies because you split the current. Your single power supply can not handle everything.

His powersupply may only produce .75a at 12v, who knows. Go back to the below and check that it is a “dual” supply.

http://arduino.cc/forum/index.php/topic,140951.msg1059054.html#msg1059054

Ok then, I tap out.

You will need to use shielded cable from the motor drive ICs (or board) to the motors. Ground the shield in the cable to ground on your board. Leave the shield open on the motor end. Also, from looking at you diagram, you may need to clean up the wiring mess and/or use shielded cable in other places. Try shielding the motor cable first, and let me know.

Patrick

I did some tests. I need help interpreting the results.

measuring equipment: Fluke 287 history graph feature.

Theory: 90W PC power supply isn't good enough because it can't supply enough current when needed Test: use car battery Result: same video flickering problem

Theory: a drop in voltage when a load is applied is what causes the camera to flicker Test: apply 4700uF 35v cap to circuit (which will also cause the video to flicker) and read voltage data from DMM. Result: voltage only momentarily dropped from 12.1v to 11.8v Conclusion: voltage drop is not significant enough be the cause of the problem.

what else should I test?

what else should I test?

You need to observe the minimum voltage drop point with an o-scope instead of a DMM. Most DMMs cannot accurately show rapid voltage changes due to their design.

but an o-scope only shows instant voltage fluctuations. It would be gone, off of the screen before I could get a chance to look at it..

David82: but an o-scope only shows instant voltage fluctuations. It would be gone, off of the screen before I could get a chance to look at it..

I would think an o-scope would have a trigger/hold feature for rapid events. Perhaps you should take your project to an electronics shop for a professional opinion. Nobody here knows the construction methods (soldered or twisted wires, wiring size, etc.) you used. If the cam requires a regulated 12v power source, any drop below 12v may cause issues. No amount of whining, fantasy thinking, praying, etc will change the way your project behaves. If you have an arduino, you might be able to make a data logger setup to detect/record how low the voltage drops.

my fluke 287 records a graphical display of whatever it is measuring. I don’t know if an o-scope has that feature.

I have a new experiment to try. I need some component that only allows amperage to be drawn gradually as if you were turning the knob of a variable pot. what would that be called?

Your fluke should have some features to allow you to see how low the voltage drops either ysing the max/min capture (100ms resolution) or duty cycle (which may have the shortest measurement interval).

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&frm=1&source=web&cd=1&cad=rja&ved=0CEMQFjAA&url=http%3A%2F%2Fwww.myflukestore.com%2Fcrm_uploads%2F287_289_umeng0100.pdf&ei=65XwUP28B-Ww0AHvyIGIBQ&usg=AFQjCNFvxUdKELcescSQ-fr3XYfgVGD0Uw&bvm=bv.1357700187,d.dmQ

Capturing Minimum and Maximum Values The MIN MAX Record mode captures minimum, average, and maximum input values. When the input goes below the recorded minimum value or above the recorded maximum value, the Meter beeps and records the new value. The Meter stores the elapsed time since the recording session was started at the same time. The MIN MAX mode also calculates an average of all readings taken since the MIN MAX mode was activated. This mode is for capturing intermittent readings, recording minimum and maximum readings unattended, or recording readings while equipment operation precludes watching the Meter. The MIN MAX mode is best for recording power supply surges, inrush currents, and finding intermittent failures. Response time is the length of time an input must stay at a new value to be captured as a possible new minimum or maximum value. The Meter has a 100 millisecond MIN MAX response time. For example, a surge lasting 100 milliseconds would be captured but one lasting only 50 milliseconds may not be captured at its actual peak value. See the MIN MAX specification for more information.

To activate the MIN MAX mode, press M. As shown in Figure 6, the Meter displays e at the top of the measurement page, and the MIN MAX start date and time along the bottom of the page. In addition, the recorded maximum, average, and minimum values appear in the secondary display with their respective elapsed times. Restart 119.81VAC Stop Start : 06/07/07 7:00 pm Maximum 127.09 119.50 110.23 Average Minimum Auto Range VAC 500 VAC VAC 00:03:17 01:10:09 00:59:59 VAC 8:10pm 06/07/07 Min Max est42.eps Figure 6. MIN MAX Record Display To stop a MIN MAX recording session, press M or the softkey labeled Stop. The summary information in the display freezes, and the softkeys change function to allow saving the collected data. Pressing M again or the softkey labeled Close exits the MIN MAX record session without saving the collected data.

Measuring Duty cycle Duty cycle (or duty factor) is the percentage of time a signal is above or below a trigger level during one cycle, as shown in Figure 25. The duty-cycle mode is optimized for measuring the on or off time of logic and switching signals. Systems such as electronic fuel injection systems and switching power supplies are controlled by pulses of varying width, which can be checked by measuring duty cycle.

zoomkat:
Your fluke should have some features to allow you to see how low the voltage drops either ysing the max/min capture (100ms resolution) or duty cycle (which may have the shortest measurement interval).

It does but why would use that instead of the historical graph? Does the min/max have a much faster cycling time or something?

David82:

zoomkat: Your fluke should have some features to allow you to see how low the voltage drops either ysing the max/min capture (100ms resolution) or duty cycle (which may have the shortest measurement interval).

It does but why would use that instead of the historical graph? Does the min/max have a much faster cycling time or something?

I don't have your fluke, so I can't do actual test like you should be doing. The bar graph apparently updates 30 times a second, but I assume you will have to be able to visually determine the low voltage value during a very short interval. Where in your circuit have you measured using the bar graph method and what were the results? I would think the duty cycle might be interrupt driven able to capture the shortest interval (the voltage trigger level would have to incremently lowered for each test to find the lowest value).