3.0V regulator with Vin down to 3.0V?

Thread Starter

hspalm

Joined Feb 17, 2010
201
Hi,
I'm doing a project where I attach an FPV camera (camera and video transmitter) to an existing quadcopter powered by a 3.7V lipo battery. I want to use the battery as power for the FPV camera.
The motors are brushed and very noisy, ripple voltages are around 0.3V and varies in frequency with motor rpm. The FPV camera blacks out when the propellers start spinning, even though the voltage never goes below 3.0V (but very, very near), and I've tested the FPV camera so I know it functions down to right about 3.0V.

I have replaced the capacitors across the motors, they were 22nF and I put in 100nF. I put a filter cap 47uF tantalum across battery leads. This all helped very much, the ripple is just 1/3 now but FPV camera still blacks out, so I think I need a voltage regulator.

Weight is really important here, so I can't use a module, and I prefer no inductors or very large caps. I increased the weight from 12.5 grams to 13.01 grams by adding the camera and the flight time was halved... Not sure how the existing electronics are powered from this noisy source, I don't see a voltage regulator on the board so maybe they are 1.8V capable and a zener diode as regulator maybe.

Appreciate any help, thank you.
 
Last edited:

Thread Starter

hspalm

Joined Feb 17, 2010
201
Just checked again with the scope, the lowest voltage from the battery is basically at 3.0V. I forgot to tell you, the FPV camera draws 200mA. So LDO is no go :(
 

Dr.killjoy

Joined Apr 28, 2013
1,196
What about using filter circuit.. How big are you talking about with this quad copter??? What about using a large cap to filter it out ??
 

cornishlad

Joined Jul 31, 2013
242
Just checked again with the scope, the lowest voltage from the battery is basically at 3.0V. I forgot to tell you, the FPV camera draws 200mA. So LDO is no go :(
When you scoped the supply to the camera what else did you see that could be having this effect on the camera. There must be some hf at a particular frequency that's having such a drastic effect..With such a restriction on weight (.5gm halved flight time) this looks like a near impossible problem. You could try taking the camera supply wires direct to the batter terminals if you haven't already done so...
 

Alec_t

Joined Sep 17, 2013
14,280
I increased the weight from 12.5 grams to 13.01 grams by adding the camera and the flight time was halved..
Do you really mean grams? The overall weight and the camera weight seem exceptionally light.
 

MikeML

Joined Oct 2, 2009
5,444
Just checked again with the scope, the lowest voltage from the battery is basically at 3.0V. I forgot to tell you, the FPV camera draws 200mA. So LDO is no go :(
Why? Using the Mouser search page I linked you to, I found ~20 that would handle 200mA
 

dl324

Joined Mar 30, 2015
16,846
Are you certain it's power supply noise that's causing the problem?

Don't know what regulations you have in Norway, but I'll shoot at any drone I see flying over my property.
 

ErnieM

Joined Apr 24, 2011
8,377
If you are just adding a voltage regulator between the battery and the camera you should expect the problems to continue. This hookup still injects noise into the LDO and what goes in will come out.

Instead look at keeping that noise contained at the motors. Inductive and capacitive filtering should work there, or in series with the camera.
 

Thread Starter

hspalm

Joined Feb 17, 2010
201
What about using filter circuit.. How big are you talking about with this quad copter??? What about using a large cap to filter it out ??
Yes, I'm also thinking filter circuit for this. As mentioned below, I don't think an LDO alone would keep the noise out and regulate.

When you scoped the supply to the camera what else did you see that could be having this effect on the camera. There must be some hf at a particular frequency that's having such a drastic effect..With such a restriction on weight (.5gm halved flight time) this looks like a near impossible problem. You could try taking the camera supply wires direct to the batter terminals if you haven't already done so...
Sorry, are these images too large to use on the forum? I don't know how to make thumbnails when I only have the picture URL.
This is some noise at some rpm before 47uF across battery leads. Yes, the camera is powered directly at the battery supply leads.

This is the noise after adding 47uF cap right at the camera supply pins. So there is some improvement. Sorry for not adding cursors.

This is the noise after increasing filter caps on motors from 22nF to 100nF



Do you really mean grams? The overall weight and the camera weight seem exceptionally light.
Yeah it's a "nano" quadcopter





Why? Using the Mouser search page I linked you to, I found ~20 that would handle 200mA
With what dropout voltage? I practicaly can't have any voltage drop. At all. I previously said that the battery voltage will at some point swing all the way down to 3.0V exactly :(

Consider a SEPIC switched regulator. It can produce an output voltage both above and below the input voltage.
Maybe it's easier to drop the voltage with a zener and use a regular boost regulator? Either way I think this is gonna way too much, and I might have to consider finding a lighter battery or one with higher capacity vs. weight.

Are you certain it's power supply noise that's causing the problem?

Don't know what regulations you have in Norway, but I'll shoot at any drone I see flying over my property.
Hmm.. Can't be certain. But the problem occurs each time the motors start spinning. This is also when the noise starts occurring at the supply lines.
According to regulations in Norway you must first have a license to carry a weapon.

If you are just adding a voltage regulator between the battery and the camera you should expect the problems to continue. This hookup still injects noise into the LDO and what goes in will come out.

Instead look at keeping that noise contained at the motors. Inductive and capacitive filtering should work there, or in series with the camera.
Great input. But I can't keep adding capacitors at the motors because I will hinder them from reacting quickly enough to keep the quadcopter stable? Or? To me the noise looks to slow to be caught by an inductor, is this true?
 

cornishlad

Joined Jul 31, 2013
242
you could try removing the camera from the quadcopter but keep the supply wires going to the battery on it. Rule out mechanical vibration having an effect on cmos sensor.
 

Thread Starter

hspalm

Joined Feb 17, 2010
201
Do you have a star ground system to isolate motor current from other current?
It's not me who designed either the quadcopter or the camera. But the camera is grounded directly at the battery leads. Or when I come to think of it, it's grounded after the power switch. Cause the battery is soldered in and I wouldn't want the camera to be transmitting at all times.

you could try removing the camera from the quadcopter but keep the supply wires going to the battery on it. Rule out mechanical vibration having an effect on cmos sensor.
Good idea! i will try it later.
 

ian field

Joined Oct 27, 2012
6,536
Hi,
I'm doing a project where I attach an FPV camera (camera and video transmitter) to an existing quadcopter powered by a 3.7V lipo battery. I want to use the battery as power for the FPV camera.
The motors are brushed and very noisy, ripple voltages are around 0.3V and varies in frequency with motor rpm. The FPV camera blacks out when the propellers start spinning, even though the voltage never goes below 3.0V (but very, very near), and I've tested the FPV camera so I know it functions down to right about 3.0V.

I have replaced the capacitors across the motors, they were 22nF and I put in 100nF. I put a filter cap 47uF tantalum across battery leads. This all helped very much, the ripple is just 1/3 now but FPV camera still blacks out, so I think I need a voltage regulator.

Weight is really important here, so I can't use a module, and I prefer no inductors or very large caps. I increased the weight from 12.5 grams to 13.01 grams by adding the camera and the flight time was halved... Not sure how the existing electronics are powered from this noisy source, I don't see a voltage regulator on the board so maybe they are 1.8V capable and a zener diode as regulator maybe.

Appreciate any help, thank you.
A good LDO regulator has a voltage overhead of about 0.5V so you can't get exactly down to Vout = Vin.

What you may need is a buck/boost regulator; Sepic, Cuk etc.

About a year or so ago, Elektor magazine published a project for a switcher substitute for 78xx regulators, it was only slightly more bulky than the real thing and could (mostly) do the full 1A without a heatsink.

A buck/boost regulator is a little more complex - but probably wouldn't add more than about 30% to the size/weight of what Elektor did.
 

Thread Starter

hspalm

Joined Feb 17, 2010
201
you could try removing the camera from the quadcopter but keep the supply wires going to the battery on it. Rule out mechanical vibration having an effect on cmos sensor.
Tried this yesterday, still reacts in the same way even though camera is hanging loose or I hold it with my fingers with only loose wires going to the quadcopter.


A good LDO regulator has a voltage overhead of about 0.5V so you can't get exactly down to Vout = Vin.

What you may need is a buck/boost regulator; Sepic, Cuk etc.

About a year or so ago, Elektor magazine published a project for a switcher substitute for 78xx regulators, it was only slightly more bulky than the real thing and could (mostly) do the full 1A without a heatsink.

A buck/boost regulator is a little more complex - but probably wouldn't add more than about 30% to the size/weight of what Elektor did.
When you say the real thing, do you mean the TO-220 package? That is way too large, unfortunately. But If we can shrink it five times and have only a fifth of the current capabilities (200ma at 3.0v) that could be interesting.
 

ian field

Joined Oct 27, 2012
6,536
Tried this yesterday, still reacts in the same way even though camera is hanging loose or I hold it with my fingers with only loose wires going to the quadcopter.




When you say the real thing, do you mean the TO-220 package? That is way too large, unfortunately. But If we can shrink it five times and have only a fifth of the current capabilities (200ma at 3.0v) that could be interesting.
They probably have what you want in the realms of medical research - where they encapsulate a tiny camera with battery and transmitter in a pill for the patient to swallow.

I won't be browsing the prices for those any time soon.
 
Top