3.7 v lipoly powering micro servo? bad idea?

[dmehling]

Is it a good idea to attempt to power servos with a 3.7 v lipoly battery?  I have read on various websites and other discussion forums that it is possible, but not necessarily the ideal.  My project will involve a small picaxe mcu, an rf receiver board, perhaps one led, and two micro servos.  I intend to use the smallest possible servos I can find.  I have found one that is about 3.7 g, but it is still 4.8-6 v.  However, the servos will only run every couple of minutes, and no longer than one second of run time.  And that is one servo at a time, not both running together.  So with very light usage of the servos and powering the other components, would be 3.7 v battery work

[dmehling]

?

[mklrobo]

  Hello, its me again! 
If you can provide me with the servo information, and the battery information that you have, I might be
able to help look up some performance stats. I assume the lipoly, is lithium poly____, ?
(Is this related to the your other project, or multitasking?  )
I will help if I can....   

[dmehling]

Yes, it's for the other project.  I just thought this question was better suited for this part of the forum.

This is the servo I had in mind:

https://www.pololu.com/product/1053/specs

The specific battery should not matter too much should it?  I have not yet decided which one I will get.  I'm just thinking of a typical lithium polymer battery.  Nearly all of them are 3.7 v.

[bdeuell]

as 4.8 v is listed as the low end of the range I would not count on it working at a lower voltage.

You could also contact the manufacturer/retailer but they will probably give the same response. If you want to know for sure buy it and try it out they don't cost too much. Likely you will find that it doesn't work (or if you're lucky turns but has very little power). Remember there is a control circuit inside a hobby servo that requires power as well.

If you only have to power 2 small servos you could look for a boost converter that can handle the current. this would let you run the servos within their specified range and would maintain constant performance across the battery life. Remember that batteries decrease in voltage as they are drained, your servo system needs to work with a range of input voltages.

[mklrobo]

  Hello!
I think I am begining to see a target of interest.
The ampere/hour of the battery will be of interest to you, in relation
to the amount of work desired, and the motor that is consuming the amperage.
Batteries Plus+ may have a good direction of what you would require.
I would base the load on the servos, doing what you want them to do for
a specified amount of time. (not forever) OR, the opposite, acquire the
amps of the servos moving, and convert that to ampere hours. The ampere
hours will determine how long you can run the servos.
The process will help you navigate the limitations of the batteries
that are avaliable, VS. time worked. What do you think?   

[dmehling]

as 4.8 v is listed as the low end of the range I would not count on it working at a lower voltage.

You could also contact the manufacturer/retailer but they will probably give the same response. If you want to know for sure buy it and try it out they don't cost too much. Likely you will find that it doesn't work (or if you're lucky turns but has very little power). Remember there is a control circuit inside a hobby servo that requires power as well.

If you only have to power 2 small servos you could look for a boost converter that can handle the current. this would let you run the servos within their specified range and would maintain constant performance across the battery life. Remember that batteries decrease in voltage as they are drained, your servo system needs to work with a range of input voltages.

Are you sure it would not be possible?  I just googled 3.7 v servo again, and found several sites and discussion forums of others who have used 4.8 v servos with 3.7 v batteries.  They suggest that it may not be the ideal, but is workable with light loads.  I have found some of the smallest servos I can find, and they will perform very light duty actions for very brief amounts of time.  The servos should encounter very little resistance.  So I should at least try it out.  I was just wanting additional confirmation of what I had read elsewhere.

I would certainly consider using a boost converter as you suggested.  If I was doing that, how would I know when the voltage is falling?  I intend to put all of the electronics for my project inside of an enclosure, except the servos.  Is there some way I could have the voltage voltage be checked internally and then have an LED light up if it goes to low?

[dmehling]

What do you think Mklrobo about this issue of 3.7 v lithium polymer battery to power light usage of a sub micro servo 4.8v?

I would really like to use just a single battery, and preferably not use the heavier NiCd or NiMH battery packs.

[dmehling]

Hello!
I think I am begining to see a target of interest.
The ampere/hour of the battery will be of interest to you, in relation
to the amount of work desired, and the motor that is consuming the amperage.
Batteries Plus+ may have a good direction of what you would require.
I would base the load on the servos, doing what you want them to do for
a specified amount of time. (not forever) OR, the opposite, acquire the
amps of the servos moving, and convert that to ampere hours. The ampere
hours will determine how long you can run the servos.
The process will help you navigate the limitations of the batteries
that are avaliable, VS. time worked. What do you think?   

Forgot to mention something else.  Based on my very preliminary estimates I am thinking that a 300-1200 mah battery should give me enough power for at least 20 or 30 hours.  I will only be running a servo for perhaps half a second every couple of minutes.

[bdeuell]

Are you sure it would not be possible?  ... They suggest that it may not be the ideal, but is workable with light loads.

see earlier post:

If you want to know for sure buy it and try it out they don't cost too much. ... if you're lucky turns but has very little power

if you are engineering a design and want to use something outside if its specified operating parameters you can: A) request additional information about the product, B) conduct your own tests to determine the performance of the product, C) take an educated guess based on previous experience, or D) build it blindly and hope it works (not really engineering or recommended by me). Unfortunately i do not have experience undervolting servos to that degree and cannot offer you direct advice on option C. keep in mind you already have very small servos so your torque is already small without undervolting the servo (not to say it won't work just a caution). Also it doesn't really matter how short of a time you need the torque 1s - 10s the torque you can deliver is the same.

I would certainly consider using a boost converter as you suggested.  If I was doing that, how would I know when the voltage is falling?  I intend to put all of the electronics for my project inside of an enclosure, except the servos.  Is there some way I could have the voltage voltage be checked internally and then have an LED light up if it goes to low?

the battery voltage will continue to decrease as it is used (a datasheet would show this discharge curve). there are many ways to monitor a batteries voltage two common methods are to use an analog input pin on a microcontroller or a comparator circuit that uses a reference voltage to determine when the battery is below a threshold.

[mklrobo]

  Hello!
(Sorry about the late entry, I had to power off for awhile)
 In regards to;
What do you think Mklrobo about this issue of 3.7 v lithium polymer battery to power light usage of a sub micro servo 4.8v?
I would really like to use just a single battery, and preferably not use the heavier NiCd or NiMH battery packs.
In reviewing  the latest posts by you and bdeuell, I would make a generalization;
Points have been make concerning voltage dips, brownouts, possible voltage mismatches,
servos that may or may not be used.
To address this, I look back at some power supplies like TVs provide. The circuits (in the old days)
provided all kinds of levels of power supplies; I was not sure why, until now.
In the interests of the aforementioned, and possible future demands, it may be prudent to
have a battery that provides 18 Volts, and can dock into your design. - Why?
Emplace voltage regulators on your design, of any voltage that you need. 5 V, 3.3V, 3.7, 4.8.
The battery that can dock, can easily be replaced, while the other is charging.
The reason for multiple voltages off of a larger voltage, is to keep regulation going to the
sub - power supplies, with no brown outs; and also gives you latitude if you need to replace
a servo with a better one, or need to replace the servo you have, but it has been discontinued.
If you need to add something that you did not expect, you have the flexibility to power that also.
It adds a little more design time, but may save you more in the future. What do you think?   
Keep me posted.....

[mklrobo]

  hello!
If the direction of the voltage supply is resolved, we can move onto
the actual battery, or other parts of the design.
In the making of the housing of the circuit, battery, and servos, I
could give an idea there. On the website, Instructables, they give
great ideas for do-it-yourself housing construction. Everything from
making a gun case, to repair your car, to special electronic housing; the
latter in this issue. I made my own fiberglass housing for one of my projects,
and  worked great. Light as a feather, strong as steel, and pretty as much as your
labor is involved. Resists corrosion, an insulator for temperature as well as electricity,
make a good case for many applications. Keep me posted!.....

[mklrobo]

  Hello!
I reviewed your post on the other thread, but could not review the
item you selected. Will try to access soon.....   

[dmehling]

Are you sure it would not be possible?  ... They suggest that it may not be the ideal, but is workable with light loads.

see earlier post:

If you want to know for sure buy it and try it out they don't cost too much. ... if you're lucky turns but has very little power

if you are engineering a design and want to use something outside if its specified operating parameters you can: A) request additional information about the product, B) conduct your own tests to determine the performance of the product, C) take an educated guess based on previous experience, or D) build it blindly and hope it works (not really engineering or recommended by me). Unfortunately i do not have experience undervolting servos to that degree and cannot offer you direct advice on option C. keep in mind you already have very small servos so your torque is already small without undervolting the servo (not to say it won't work just a caution). Also it doesn't really matter how short of a time you need the torque 1s - 10s the torque you can deliver is the same.

I would certainly consider using a boost converter as you suggested.  If I was doing that, how would I know when the voltage is falling?  I intend to put all of the electronics for my project inside of an enclosure, except the servos.  Is there some way I could have the voltage voltage be checked internally and then have an LED light up if it goes to low?

the battery voltage will continue to decrease as it is used (a datasheet would show this discharge curve). there are many ways to monitor a batteries voltage two common methods are to use an analog input pin on a microcontroller or a comparator circuit that uses a reference voltage to determine when the battery is below a threshold.

I'm willing to at least try it since some have apparently managed to do it without any serious issues.  The servos I have in mind are quite inexpensive so it would not be that terrible if they malfunctioned.

Having said all that, I would like to know the potential problems.  What would be the harm of lower than recommended voltage?  Could it damage the battery or the servo?  And what is the harm of reduced torque on the servo itself?  Could that also be something that could damage the internal components?

With regards to using a voltage booster, the batteries I want to get actually have a voltage of 4.2 v when fully charged, and they go down to 3.7 v when fully depleted.  What impact would that have on the ability of the voltage booster to output adequate voltage?  Would that make any difference?  Does 3.7 v provide enough voltage to be boosted to 4.8 v?  And is there a danger of boosting it too high?

I appreciate your patience since many of these concepts are new to me.

[mklrobo]

  Hello!
In regards to your questions;
Having said all that, I would like to know the potential problems.  What would be the harm of lower than recommended voltage?
A generic answer to this is an ohms law example; - 50 ohms at 10 volts is .2 amps || 50 ohms at 5 volts is .1 amp.
In that sense, you are loosing amperage( which motors need) and the motor slows down. Like a dimmer switch
on your lights.  (ohms is the servo load) 
 Could it damage the battery or the servo?  And what is the harm of reduced torque on the servo itself?  Could that also be something that could damage the internal components?
Sind you are dealing with small motors, the damage may not show up for awhile, if ever, depending on the
workmanship of the motor. Some motors have saftey feature built in, that prevents most damage.
Bdeuell has posted alot of info on this, on a related thread, dealing with the performance expectations of
a motor. Will have to get back with you about the power booster. 

With regards to using a voltage booster, the batteries I want to get actually have a voltage of 4.2 v when fully charged, and they go down to 3.7 v when fully depleted.  What impact would that have on the ability of the voltage booster to output adequate voltage?  Would that make any difference?  Does 3.7 v provide enough voltage to be boosted to 4.8 v?  And is there a danger of boosting it too high?

[dmehling]

The servos I am thinking about using require somewhere between 100-200 ma.  Would a 3.7 v battery be able to supply that much amperage?

[dmehling]

I have also considered using a 7.4 v lithium polymer battery, and using voltage regulators to provide the right voltage for the microcontroller and servos.  Does using a voltage regulator reduce the amount of amp hours that a battery can provide?  And by how much would it be reduced?

A related question is, do I need separate voltage regulators for the microcontroller and the servos?

[jamesd168]

is weight of the battery a key design consideration for you? if so, adding a voltage regulator will also add the weight to your device. if weight is not an issue, then your design options are wide open, you can use a booster for one cell, or a divider with two cell battery pack, etc.

[bdeuell]

Having said all that, I would like to know the potential problems.  What would be the harm of lower than recommended voltage?  Could it damage the battery or the servo?  And what is the harm of reduced torque on the servo itself?  Could that also be something that could damage the internal components?

I doubt you will cause any harm to the servo electronics. The low voltage could cause them to be underpowered resulting in higher currents when compared to the same torque at rated voltage. If run for a period of time this could overheat and burn out the motor. caution that servos are often powered even when not moving this allows them to maintain their position if a load is applied. basically at the lower voltage you still do not want to exceed the motors rated operating current for a given duty cycle.

Reduced torque should actually help the life of the mechanics inside the servo.

The battery life is often dependent on the charge discharge cycles you put it through. Typically deeper discharge = reduced life, faster discharge = reduced life, higher temperature = reduced life.

With regards to using a voltage booster, the batteries I want to get actually have a voltage of 4.2 v when fully charged, and they go down to 3.7 v when fully depleted.  What impact would that have on the ability of the voltage booster to output adequate voltage?  Would that make any difference?  Does 3.7 v provide enough voltage to be boosted to 4.8 v?  And is there a danger of boosting it too high?

A boost converter is designed to give a constant output voltage for a range of input voltages, the allowable input range depends on the converter and would be given in the specs/datasheet.

The servos I am thinking about using require somewhere between 100-200 ma.  Would a 3.7 v battery be able to supply that much amperage?

This depends on the battery specs. No battery should be discharged at a rate greater than its specified maximum and lithium batteries are particularly sensitive and at risk to blow up if discharged too quickly, over discharged, or overcharged. youtube has plenty of videos if you want to see this i would recommend avoiding it yourself.  Most lithium batteries you find will have a built in protection for these conditions but i would not rely on these to protect the battery, you should design your own protections for these conditions into your system.

I have also considered using a 7.4 v lithium polymer battery, and using voltage regulators to provide the right voltage for the microcontroller and servos.  Does using a voltage regulator reduce the amount of amp hours that a battery can provide?  And by how much would it be reduced?

It does not directly reduce the capacity of the battery but a regulator/booster will waste power and reduce the overall efficiency of your system, how much depends on your regulator type/design. However battery capacity is dependent on discharge rate so drawing more current from your battery will shorten its run time.

A related question is, do I need separate voltage regulators for the microcontroller and the servos?

Depends on the voltage you supply for the motors, microcontrollers are usually not very flexible on their input voltage. However its not a bad idea having a separate linear regulator for the microcontroller can eliminate a lot of problems by isolating the power.

[dmehling]

I have done some research since I posted my last message and I have found some lower voltage servos that will work with 3.7v lipoly batteries.  However, I'm not sure about the best way of connecting the batteries to the microcontroller and servo.  The particular microcontroller I want to use is a Moteino, which has a 3.3v voltage regulator and can provide up to 250 mA of output current.  The servo I intend to use can receive input of 3.3v and has an idle current of 50 mA and no load running current of 90 mA.  The stall current is 500 mA.  I want to know if I can connect the battery to just the microcontroller and then connect that to the servo, without the need to separately connect them to the battery.  Furthermore this will ensure that the servo gets 3.3v because of the onboard regulator on the mcu.  I really don't think that my servo will ever reach stall current levels, so I am thinking that I should be able to do this without any problems. 

Having said that, I would like to know how much resistance would be needed to really increase the current draw on the servo.  The specs say that the servo has .3 kg-cm of torque.  I really have no idea how much that would be.  The point of the servo is to gently touch a surface and then return to its original position.  So I would think that it would barely increase the amount of current it would draw.  I know this may seem like a simple matter, but I am just trying to stay on the safe side with this first project of mine

[bdeuell]

I would not recommend powering the servos from the microcontroller voltage regulator. You run the risk or damaging the regulator as well as causing all sorts of interference and dropout issues for the microcontroller. For your application i would recommend powering the servos directly from the battery, assuming the new servos are designed to work with the voltage range the battery provides.

I also would definitely not assume that you will not see the stall current. not only is that poor design practice (unless you have some other limiting control in place) it is definitely not "trying to stay on the safe side". especially as you admit you have no idea what torque to expect. Also the servo will draw the full stall current and likely more each time it starts to move from a stopped position (albeit for a very brief time interval).

.3 kg-cm (while not actually units of torque ... technically it should be in force and distance not mass and distance) is equivalent to saying that a 1 cm long horizontal arm with a .3 kg weight attached at one end and the servo attached to the other. Also equal to a 10 cm long arm with a .03 kg weight. I can't provide any guidance on the forces you should expect as you have provided no details on the mechanics.

you can assume the current draw increases approximately linearly with torque from the no load current (at zero torque) to the stall current (at stall torque).

[dmehling]

That is really helpful.  I had no idea that the servo would probably reach stall current levels as soon as it starts turning. 

[mklrobo]

  Hello!
Bdeuell give great advice. In reference to your concern of powering the
servos, Venfors do offer servo driver circuits, which may simplify some
of the issues you have. All the power goes through the servo controller,
the MCU is isolated, and programming feedback is offered in some servo
driver kits. I assume noise filtering is maintained through the servo controller,
as well as protection.
In reference to the power booster issue, those would be extra circuits and weight, Vs.
IC chips for voltage regulators, which have little weight and compact. 

[mklrobo]

  Hello!
Also, insofaras you batteries, ServoCity is having a sale on batteries,
as expressed in the Miscellanous section, in this forum.

[mklrobo]

  Hello!
Thinking through the whole design issue, and the questions/directions
in sight, maybe this would be an option;
Buy the microcontroller, servos, and accessories that you think you will need.
Buy a power supply that can deliver a main voltage, in which you break down
into subsystems with voltage regulators. (alot of seperate power supplies would work too)
Set up a test jig and fixture, to be able to look at the entire setup. You can begin
programming, then measuring the results you are looking for. If any problems come up,
you can deal with them accordingly. When all your projections are realized, and the
system is the way you want it, you can assemble the unit in the form in which it will
fit on your Kindle. What do you think?   

[wasosur]

You could also contact the manufacturer/retailer but they will probably give the same response. If you want to know for sure buy it and try it out they don't cost too much. Likely you will find that it doesn't work