High power voltage divider

[Avalan]

Hi,

Does anyone have any good ideas how one could build a voltage divider that can actually deliver decent current (2-3A). The traditional resistor network would not work for obvious reasons.

I want to use this on my drone, the autopilot can take between a 3 and 6 cell battery but I want to power it using a 8 cell. The reason why a normal regulator will not work is that the autopilot must see the battery voltage to allow low battery warnings and fail safe procedures. Using a regulator the autopilot will think the battery is fine all the time and give wrong battery level indications. So the idea is that if I use 8 cells and divide it in half I can tell the autopilot that its running a 4 cell. It would then be happy and correctly monitor the battery voltage.

One idea that came to mind was to use a switching regulator, normally these devices has a small resistor divider circuit connected to the output which provides a reference for setting the output voltage. This might be a long shot but maybe with some modifications I could rather connect this divider circuit to the input and have the reference follow in the input voltage. Hopefully this would create some kind of regulated divider circuit.

Alternatively I could build a custom/discrete SMPS for this purpose but I hoped there might be some easier options?

 

[tomizett]

You should be able to take a standard buck converter design and replace its voltage reference with a divider derived from your incoming supply. You could either build your own or possible modify an off-the shelf unit.

 

[Avalan]

Ah okay thanks! So its not such a far fetched idea to connect the reference divider circuit from the input of a DC-to-DC converter then. I have a chip in mind (MP2560DN), will study the datasheet and head in that direction then...

 

[MikeMl]

As tomi suggested, for high efficiency, you want a buck converter whose output is a ratiometric fraction of the input voltage.

 

[alec_t]

I can tell the autopilot that its running a 4 cell.

Surely you don't need a 2A supply to do that? I would expect the monitoring input of the autopilot to be high impedance (e.g. like a conventional op-amp input)?

 

[KeepItSimpleStupid]

I just need to add a "hiccuup" to all of this because I don;t know where the battery falls in place. LiPo or similar batteries tend to use a "battery gas gauge" IC and the voltage is fairly level for lithium batteries. https://www.rfcafe.com/references/electrical/battery-li-poly.htm

I'm not saying that the system is not voltage based.

Lead acid, however, drops off like a rock: https://www.google.com/url?sa=t&rct...ng.pdf&usg=AFQjCNEZo1yU5IydPnHrucXwJJYTIv_iOA

Basing the SMPS reference off the 8 cell battery is the first thing that came t mind.

This is tread is a good example where the "problem is well stated, a solution was proposed by the OP and that there is obviously an easier/better solution.

Battery type or voltage was not mentioned, but I assumed it to be one to the higher energy density batteries.

 

[Mikebits]

Agree with Alec on the stated current requirement. Also, in line with KISS thinking, here is some links to more battery stuff that may apply. The Battery Uni is a good reference link as well.
Battery Uni: **broken link removed**
Sadly the battery gauges I found need a uC. : https://www.ti.com/lsds/ti/power-ma...=Integrated Protector;Battery Management Unit

 

[crutschow]

As Alec and Mike noted, your current requirements don't make sense.
If all you want is lower voltage for the autopilot to measure than a resistor divider will work fine.

 

[Les Jones]

What about having a voltage detector (Set to the cutoff voltage of the 8 cell battery.) on the input of the switch mode regulator and using the output of that to adjust the output voltage of the regulator down to the point were the autopilot sees the low battery level.

Les.

 

[Tony Stewart]

SoC battery monitors use either cell voltage or Coulomb counting to determine state of charge. LiPo's tend to have fairly flat SoC voltage which varies as much with with load current, thus current sensing and counting coulombs is the preferred way.

A buck SMPS then delivers the required power to the load which is a separate task. It may have a low voltage sense on input and/or output for Power OK. indication.

 

[Avalan]

alec_t: I agree 2A might be a bit too much but I thought I would stay on the safe side. The autopilot also does not have a special monitoring input and takes the full output from the battery to power it, it again also powers a GPS, compass, IMU, etc. units. Suppose it does it's own regulation internally. With my statement of telling the autopilot its running from a 4 cell, its just configuring the fail-safes for a 4 cell battery pack.

KeepItSimpleStupid: The batteries are LiPo's. When fully charged each cell is 4.2V, normally I would start to come back when the cells reach about 3.6V.

Les Jones: I am not completely sure what you meant, but it sounds like something that could work. Could you just explain it again?

I agree, maybe the voltage is not the best way to measure the amount of "fuel" available. Measuring the current/hour might give a better estimation but since the autopilot has fail-safe features that gets triggered based on voltage levels I have to rely on them at the moment.

Thanks for the valuable inputs and links from everyone!

 

[Les Jones]

I will use an example with voltages. I will assume a fully charged cell is 3.3 volts and that when it gets down to 3.0 volts it will be considered almost fully discharged. (I do not know if these values are correct but they should explain the reasoning.) So the auto pilot would consider a low battery when the voltage drops to 12.0 volts. So we set the switch mode regulator to normally give 13.2 volts. On the input of the regulator the voltage detector would be set to detect when the voltage drops below 24.0 volts. (We will say the output of the detector goes from a logic low to a logic high.) We will say that the logic high energises a relay. The contacts would switch in another resistor to shunt the top resistor of the potential divider between the output and the voltage sense input on the regulator chip. This would cause the output voltage to drop. We choose the value of this resistor so that the output of the regulator drops to 12.0 volts. The autopilot would now sense low battery. In practice we would use an electronic way to adjust the output of the regulator rather than a relay. I hopes that this makes sense.

Les.

 

[KeepItSimpleStupid]

The simplest way is to just track. A voltage divider for the battery just "triggered a thinks" it's a smaller pack.

This may waste energy. Suppose the system created a shutdown at voltage x and suppose it has an A to B operating voltage.
The initial SMPS can always output voltage of value A and monitor the battery voltage separately.

That monitor could be a gas gauge or a particular voltage.

Then let's drop the regulation to the value A, when the battery voltage drops.

The assumption is that only the battery voltage signifies something wrong. I would expect that a user has other ways to "trouble". It's possible that the failure switch is a wired ordered connection, so you can make another device activate it

I could be "missing something" You said programming to 8 cells, but then you can;t use that for some reason. Is there something else that requires 6 so you need an 8 to 6 drop in voltage to operate that?

As I said, I would HOPE that you have an external fault input NOT related to the battery voltage to handle the other device and that you need 6 and 8 cell power.

Info to help us would help. I've fixed stuff by asking questions. e.g an Auger electron Spectrophotometer. I've also contacted companies based on threads here with positive results. e.g. You can't read the schematic on your datasheet - Response - we were unaware and are in the process converting the datasheets to the format of the company that bought us. Thank you for bring it to our attention.

We at ETO probably don;t know much about drones, but we have an active owner/operator pilot here.

Links to the controller might help.

 

[Les Jones]

Hi KISS,
If we had a switch mode regulator chip which allowed the use of an external reference voltage (Rather than just access to the comparator input that compares a fraction of the output voltage with the reference voltage.) then the output could easily be made to be a fixed fraction of the input voltage. I have not seen any chips with that option.

Les.

 

[spec]

Hi,

Does anyone have any good ideas how one could build a voltage divider that can actually deliver decent current (2-3A). The traditional resistor network would not work for obvious reasons.

I want to use this on my drone, the autopilot can take between a 3 and 6 cell battery but I want to power it using a 8 cell. The reason why a normal regulator will not work is that the autopilot must see the battery voltage to allow low battery warnings and fail safe procedures. Using a regulator the autopilot will think the battery is fine all the time and give wrong battery level indications. So the idea is that if I use 8 cells and divide it in half I can tell the autopilot that its running a 4 cell. It would then be happy and correctly monitor the battery voltage.

One idea that came to mind was to use a switching regulator, normally these devices has a small resistor divider circuit connected to the output which provides a reference for setting the output voltage. This might be a long shot but maybe with some modifications I could rather connect this divider circuit to the input and have the reference follow in the input voltage. Hopefully this would create some kind of regulated divider circuit.

Alternatively I could build a custom/discrete SMPS for this purpose but I hoped there might be some easier options?

Hi Avalan,

Could you clarify a couple of points:
(1) What type of batteries will you be using?
(2) Am I right in thinking that you would like to use eight batteries instead of four to increase the battery duration?

spec

 

[KeepItSimpleStupid]

Les:
The tl494 makes the reference external. https://www.ti.com/lit/ds/symlink/tl494.pdf

spec:
You were asleep
(1) LiPo See: https://www.electro-tech-online.com/threads/high-power-voltage-divider.147165/#post-1250887

(2) The OP said "the autopilot can take between a 3 and 6 cell battery but I want to power it using a 8 cell" Not sure if it answers your question, but I would assume so.

 

[spec]

Les:
The tl494 makes the reference external. https://www.ti.com/lit/ds/symlink/tl494.pdf

spec:
You were asleep
(1) LiPo See: https://www.electro-tech-online.com/threads/high-power-voltage-divider.147165/#post-1250887

(2) The OP said "the autopilot can take between a 3 and 6 cell battery but I want to power it using a 8 cell" Not sure if it answers your question, but I would assume so.

I didn't ask what the chemistry of the batteries is, I want to know what the specific batteries are. By the way, LiPo is a misnomer. Nearly all LiPo batteries on the market at the moment are LiIon in a flexible case. But this needs to be established.

What is the value of your second answer?

 

[KeepItSimpleStupid]

Battery university seems to define "type" more in terms of dimensions.
**broken link removed**
My device takes an A, AA C and D cell.

If I asked for transistor type, you might get "NPN silicon transistor" not necessarily a 2n3904 for an answer. It depends on context.

1. I need a battery for my flash light
2 responds: What kind?
1 responds: AA
2 responds: NiCad, alkaline, carbon Zinc?

Confusing Engrish.

What is the value of your second answer?

Not much. Food for thought? Assume makes an A$$-U-Me, so it doesn;t hurt to ask.

 

[spec]

Battery university seems to define "type" more in terms of dimensions.
**broken link removed**
My device takes an A, AA C and D cell.

If I asked for transistor type, you might get "NPN silicon transistor" not necessarily a 2n3904 for an answer. It depends on context.

1. I need a battery for my flash light
2 responds: What kind?
1 responds: AA
2 responds: NiCad, alkaline, carbon Zinc?

Confusing Engrish.

Not much. Food for thought? Assume makes an A$$-U-Me, so it doesn;t hurt to ask.

I asked Avalan the question and you were being patronizing saying I was asleep.

 

[spec]

Hi Avalan,

This is an approach for sorting your battery and monitoring that I am considering, but this depends on the answers to these questions:
(1) Am I right in thinking that you would like to use eight batteries instead of four solely to increase the battery duration and that you do not need a higher supply voltage for other purposes?
(2) What specific batteries will you be using (part number)? Note that nearly all LiP batteries currently on the market are actually LiIon batteries in a light flexible case. True LiP batteries are still at the development phase.
(3) Are all the batteries the same ie make, capacity, etc ?