Cap pack?

This site may earn a commission from merchant affiliate
links, including eBay, Amazon, and others.
Sorry I am a bit late to the party here but wanted to chime in since I have built 4 cap packs and done a fair amount of data logging of ripple voltage. (also have an Electrical Engineering degree) I suspect sadlerbw above might also.

To answer some questions, I personally believe you will not "feel" any difference.
These capacitor packs cannot power more than the servo on their own without the battery. Hitting the throttle would drain the caps almost instantly. (With the exception of some monsters over 20,000uF)
All of our ESCs come with capacitors built in already and are there to smooth out the ripples in the voltage supply. The problem comes up more so related to large motors and or aggressive gearing, which create larger amperage loads on the system.

The cap-packs purpose is to smooth out the oscillations in power, the ultimate goal is to provide a smoother power level. Having an oscillating dc power supply "stresses" the ESC and will greatly shorten its life. It could also blow the caps that are pre-installed on the ESC. That is the whole point, save the ESC from power supply stress.

During a speed run with an 8s system I would often see data logs of voltage drops of nearly 1v per cell. My 8s system (4.2v * 8 = 33.6 volts turned into nearly 25.6 volts. ) Prior to installing the cap packs data logs showed oscillations in the power supply of up to 20%

I have always built my Cap-Packs with low ESR caps from the Panasonic FM line. I have used 35v caps from 1000uF up to 2200uF wired in parallel depending on the car and room available. I have found that around 6000uF total is sufficient in most cases. On larger RC's where you have more free space, fit in as many as you can within reason. It is also noted that you should wire them as close to the ESC as possible. I like to install them with an battery connector to make it easy to remove during repairs on the RC. I also use clear shrink wrap because it is easy to inspect the cap tops for swelling which is the #1 sign of a cap on its way out. Like batteries caps have a certain number of cycles for their life span and heat tolerance, so considerations for mounting location should be made.

Ideally you may want to test the capacitor pack with your multi-meter initially and then test it every 6 months to make sure it is still healthy.

-Liberty
 
I have 13200uf cap pack in my Outcast. It's increase the punch and the acceleration. My Outcast is the heaviest because I have many of T-bone parts, but still the fastest.

I noticed that the ESC and batteries temperature has been reduced after adding the capacitor " 13200uf cap pack"
 

Attachments

  • IMG_20190331_164409_319.jpg
    IMG_20190331_164409_319.jpg
    152.7 KB · Views: 132
  • IMG_20181212_233716_728.jpg
    IMG_20181212_233716_728.jpg
    479.4 KB · Views: 137
Sorry I am a bit late to the party here but wanted to chime in since I have built 4 cap packs and done a fair amount of data logging of ripple voltage. (also have an Electrical Engineering degree) I suspect sadlerbw above might also.

Thanks for the recognition, but I'm no EE. If I told you half the things I've done with electricity, you would probably take away my soldering iron! Anyway, I'm half tempted to throw one of these on my truck just as a joke:

https://www.mouser.com/ProductDetail/KEMET/ALS70A114KF040?qs=sGAEpiMZZMtZ1n0r9vR22YqjN7vGN3ZSQ36MZ6S8MYleS3WczQB4BQ==

I've got a 68,000uF one already for....reasons. But that just doesn't seem like enough! I was looking at the multi-farad supercaps, but the only ones that would fit inside an RC car had too low of a voltage rating!
 
Last edited:
Thanks for the recognition, but I'm no EE. If I told you half the things I've done with electricity, you would probably take away my soldering iron! Anyway, I'm half tempted to throw one of these on my truck just as a joke:

https://www.mouser.com/ProductDetail/KEMET/ALS70A114KF040?qs=sGAEpiMZZMtZ1n0r9vR22YqjN7vGN3ZSQ36MZ6S8MYleS3WczQB4BQ==

I've got a 68,000uF one already for....reasons. But that just doesn's seem like enough! I was looking at the multi-farad supercaps, but the only ones that would fit inside an RC car had too low of a voltage rating!

I'll shoot you a PM because series capacitors gets complicated versus parallel, and I don't want to muddy the thread. I have built a set with super caps that was 2F
 
To answer some questions, I personally believe you will not "feel" any difference.

I agree.

Hitting the throttle would drain the caps almost instantly.

During a speed run with an 8s system I would often see data logs of voltage drops of nearly 1v per cell.

Prior to installing the cap packs data logs showed oscillations in the power supply of up to 20%

Yes- my response to your 2nd and 3rd part here reflects back onto the first part- time involved. Other than a measurable but real world insignificant power drop... my guess is your average person is not going to acel or top speed any faster whatsoever with a cap pack. Especially off road where traction loss is by far the biggest loss to accel followed by the drag and power lost out the front wheels freely spinning, giving power away in a wheelie..

And actually- since we are talking minute variables- they will be slightly slower due to extra weight. Unlike real cars where the owner rides along- the lighter wallet is not helping here. :)

That is the whole point, save the ESC from power supply stress.

Ideally you may want to test the capacitor pack with your multi-meter initially and then test it every 6 months to make sure it is still healthy.

And if that is the case, what is the cost of a good cap pack, connectors and time VS the cost of a new esc taking into account the presumed lost longevity of not having the pack.

Just throwing numbers out, but if an esc is $60, the total cost of a cap pack installed is $20- then you need 33% increased life to break even.
 
It's increase the punch and the acceleration.

I noticed that the ESC and batteries temperature has been reduced after adding the capacitor


In the real car world we call that 'increased throttle response' .... ;)

If we took 2 identical cars other than the cap pack, didn't tell you which is which- you wouldn't know the difference.

It is not physically possible- everything else unchanged other than the caps- for the car to both accelerate faster and have cooler electronics. (discounting whatever minimal increase in airflow cooling at each exact moment.- which would be no measurable gain/loss)
 
I prefer Ripple Killer cap packs myself.. I make and sell RK i can tell you there are differences in what cap packs can do for a rc car. I take no credit for the design or products used i was a distributor for the guy Stuart Mutch who originally owned RK. I used to make my own caps and occasionally sell but the pcb we have are far superioir to anything i could have made. just a heads up though cap packs can help a car run better though maybe just through efficiency.
 
In the real car world we call that 'increased throttle response' .... ;)

If we took 2 identical cars other than the cap pack, didn't tell you which is which- you wouldn't know the difference.

It is not physically possible- everything else unchanged other than the caps- for the car to both accelerate faster and have cooler electronics. (discounting whatever minimal increase in airflow cooling at each exact moment.- which would be no measurable gain/loss)
I would disagree there, as certain caps can increase speeds. I have seen it first hand serveral times in testing cap packs.
 
I would disagree there, as certain caps can increase speeds. I have seen it first hand serveral times in testing cap packs.

If you could explain how, instead of just saying you saw it...that would be helpful. :) ?
 
If you could explain how, instead of just saying you saw it...that would be helpful. :) ?
I don't know how it does it. I'm no electronics guru. I'm just saying it is possible. i can only assume by making the system more efficient. one more thing is my esc cost 180.00 castle xlx so its well worth it to me. bur I'm a speed runner .
 
But you said 'make and sell' the cap packs, so you not only have a possible bias / incentive- but you should also know how they work.
 
But you said 'make and sell' the cap packs, so you not only have a possible bias / incentive- but you should also know how they work.
you are taking what you want from my reply! if you read what i said"I take no credit for the design or products used i was a distributor for the guy Stuart Mutch who originally owned RK " i follow what i was instructed! im not here to sell cap packs but to add what i have seen. And i never said what cap pack i was running did I? please don't try and make this something it isn't.
 
Speed run guys have always looked for that last 2%. Records are often achieved by small differences like 2mph. Nearly every top speed runner I know runs ripple killer.
I think many people like to run what is known to work well.

I am a DIY'er so I build my own.
 
Speed run guys have always looked for that last 2%. Records are often achieved by small differences like 2mph. Nearly every top speed runner I know runs ripple killer.
I think many people like to run what is known to work well.

I am a DIY'er so I build my own.
yeah prob so we have a great group of very fast cars on every platform, and it grows daily. nothing wrong with Diy.
 
In the real car world we call that 'increased throttle response' .... ;)

If we took 2 identical cars other than the cap pack, didn't tell you which is which- you wouldn't know the difference.

It is not physically possible- everything else unchanged other than the caps- for the car to both accelerate faster and have cooler electronics. (discounting whatever minimal increase in airflow cooling at each exact moment.- which would be no measurable gain/loss)
i agree with what you are saying here but it does help your temperatures a tiny bit because the capacitors you put on take stress off of the ones inside of the esc. and because of that they don't need to work nearly as hard and run way cooler. you know more than me on this subject though so if i am wrong i would not be surprised.
 
And if that is the case, what is the cost of a good cap pack, connectors and time VS the cost of a new esc taking into account the presumed lost longevity of not having the pack.

Just throwing numbers out, but if an esc is $60, the total cost of a cap pack installed is $20- then you need 33% increased life to break even.

I only ever used them with the XL-X Max5 and MMX all of which are in the 140-190 price range.
I recommend them if you modify and upgrade. All stock probably little ROI to be seen.
 
I wonder how to calculate the capacity required to keep the ESC happy. Has anyone done the maths? I'm not an electrical engineer but using Ohm's law and information from https://www.electronics-tutorials.ws/rc/rc_2.html I tried to calculate the correct value.

Some assumptions:
  1. The motor is 2050kV which means for 22.2 V it will rotate at 45510 rpm, or 758.5 rotations per second. Since it's a 4-pole motor the ESC has to switch current to a different wire 3034 times per second, or about every 0.3 ms.
  2. I've read that the voltage ripple should be lower than 5%, so I want to calculate which capacitor will keep the voltage above 21.09 V for a duration of 0.3 ms.
  3. In the discharge formula we also need to know the series resistance. If I assume the motor pulls a maximum of 150 A then the equivalent resistance is 22.2 V / 150 A = 0.148 Ohm
Maybe I'm already wrong with any of these assumptions, but let's use these figures to calculate the required Farads.

At first I thought the basic formula is simply T = R*C where T is time in seconds, R is resistance in Ohm and C is capacitance in Farads. So to calculate C we get 0.0003 / 0.148 = 0.002 F, 2 mF or 2000 µF. But this gives us the result for 1 time constant, and looking at the discharge curve at one time constant the capacitor will be discharged for 63%, not the 5% we were looking for.

Using the more complex formula Vc = Vs * e^-t/RC and some a lot of help from WolframAlpha https://www.wolframalpha.com/input/?i=21.09+=+22.2+*+e^-(0.0003/(0.148+*+x)) , I get a result which is about 20 times higher: 0.0395 F or 39.5 mF or 39500 µF!

None of the cap packs I know about is this large. The big Kill Mode RC pack is "only" 13200 µF so three times less than my calculations.

Did I make a stupid math mistake or are any of the assumptions wrong or unrealistic?
 
I wonder how to calculate the capacity required to keep the ESC happy. Has anyone done the maths? I'm not an electrical engineer but using Ohm's law and information from https://www.electronics-tutorials.ws/rc/rc_2.html I tried to calculate the correct value.

Some assumptions:
  1. The motor is 2050kV which means for 22.2 V it will rotate at 45510 rpm, or 758.5 rotations per second. Since it's a 4-pole motor the ESC has to switch current to a different wire 3034 times per second, or about every 0.3 ms.
  2. I've read that the voltage ripple should be lower than 5%, so I want to calculate which capacitor will keep the voltage above 21.09 V for a duration of 0.3 ms.
  3. In the discharge formula we also need to know the series resistance. If I assume the motor pulls a maximum of 150 A then the equivalent resistance is 22.2 V / 150 A = 0.148 Ohm
Maybe I'm already wrong with any of these assumptions, but let's use these figures to calculate the required Farads.

At first I thought the basic formula is simply T = R*C where T is time in seconds, R is resistance in Ohm and C is capacitance in Farads. So to calculate C we get 0.0003 / 0.148 = 0.002 F, 2 mF or 2000 µF. But this gives us the result for 1 time constant, and looking at the discharge curve at one time constant the capacitor will be discharged for 63%, not the 5% we were looking for.

Using the more complex formula Vc = Vs * e^-t/RC and some a lot of help from WolframAlpha https://www.wolframalpha.com/input/?i=21.09+=+22.2+*+e^-(0.0003/(0.148+*+x)) , I get a result which is about 20 times higher: 0.0395 F or 39.5 mF or 39500 µF!

None of the cap packs I know about is this large. The big Kill Mode RC pack is "only" 13200 µF so three times less than my calculations.

Did I make a stupid math mistake or are any of the assumptions wrong or unrealistic?

There are many more factors involved I'm afraid. The source of the issue is the batteries and their ability to provide constant power under load or small bursts. There is a significant difference between quick throttle blips and a smooth throttle input speed run. Coming up with the math for these specific scenarios is more than I have time for.

Higher quality batteries will reduce ripple and heat significantly. The problem comes up that all battery companies have inflated the C-rating of their LiPo batteries since lower quality brands starting claiming they have 40c when quality (honest) brands had 25c. People purchased the 40c because it was rated higher, assuming the details were accurate. To keep up in the business world all LiPo companies now inflate these numbers. I have seen 150c on some packs and could be higher numbers out there? I can tell you that even the best batteries likely do not exceed 60c. On this topic stick with well known battery brands. My preference has been SMC, Turnigy, and Dinogy. (there are others that are great as well)

Sorry to dodge your questions about the math, but in reality nothing beats real world testing and real data. I have seen it countless times where simulations and calculations just don't tell the full story. What people need to know, using Castle ESCs I have done quite a bit of real world data logging during runs with large Leopard and TP Power motors. Findings from the logs indicate anything 6000 uF or greater will get the ripple voltage back into a desirable range. Bigger is still better here, but that comes at both a price premium and physical size. My recent cap pack build was 12,600 uF just because I had extra caps and plenty of space in the Kraton.

-Liberty
 
Last edited:
Thank you for the reply @LibertyMKiii , very helpful!

I understand that battery quality will make a big difference. With a perfect 1000C battery an ESC would probably not need much of a capacitor at all :unsure:

My 1st assumption was obviously very wrong; there's no way the capacitor has to provide all power for a full 0.3 ms while the battery doesn't do anything at all. The actual ripple is shorter and doesn't follow a square waveform, so there's no easy way to say "the capacitor must be able to hold this voltage for this amount of time". Plus, when you're accelerating from a standstill the rotations per second start very low so the time the ESC needs to switch power to a different wire will be much higher than 0.3ms.

I understand that real world data is the only way to know which capacity is needed to get a specific result. Only ESC manufacturers might be able to run good simulations to get useful formulas.

It's cool you have some real world data and that you found the number of 6000 µF through experiments. I recently got 4 Panasonic 2200 µF capacitors to build a cap pack for my Kraton with stock electronics, happy to hear that should be more than adequate (y)

Just out of curiosity, do you, or anyone else, have any data on the maximum current pulled by the motor? I'm wondering if my third assumption of 150 A is anywhere near realistic?
 
Wanna melt your brain? Go have some fun poking through this mess: https://onlinelibrary.wiley.com/doi/full/10.1002/ente.201600154

It doesn't directly address our questions about the response time of batteries, but it DOES have some information in it that is pertinent to what we are interested in. Here is some of the more interesting info I THINK I found in there. It's fairly dense and technical though, so I may be wrong:

- First of all, they were testing at frequencies up to 10MHz, which is way faster than any of the switching regulators in our ESC's are running. At those speeds, the batteries they were testing acted like they had a higher internal resistance and impedance than normal, but they did react. Basically, 10MHz wasn't 'too fast' for the battery, although the C-rate it would be able to deliver would be lower.
- The battery was, however, quite happy delivering power at a 10kHz pulse frequency.
- There are generally three different ways a battery delivers power, and which one is happening depends largely on the frequency of the power draw. The three 'modes' all act like they have a different range of internal resistances, although the ranges can overlap. Also, these modes aren't mutually exclusive. So, if you are using a battery to drive a switching power supply (which is pretty much what an ESC is), you could be in both the high frequency and low frequency 'modes' at the same time. Why do we care? Because it means the battery isn't going to act the same all the time, so there isn't an easy answer to figuring out where it might need help to keep up with demand.
- Over about 20kHz, the impedance of the test system was driving the response more than the battery was. Basically, at those speeds, the wiring mattered more than anything the battery cells themselves were doing. So, this could mean that one of the reasons capacitors help is that they are physically closer to the switching components. Literally, just being closer than the battery might be part of the reason it helps!

Now, having said all that, PLEASE DON'T TAKE THIS AS 100% TRUTH! I'm trying to interpret some data that wasn't really the focus of a paper that is in a field I don't have experience with. I could very well be super, mega wrong. The only reason I was willing to make some conclusions at all was in the hope that it would help you understand that batteries are not easy to figure out. Without some serious science chops and an oscilliscope, your best bet is probably just to guess and check: if you have a power problem, add a cap pack and see if it helps!
 
Old Thread: Hello . There have been no replies in this thread for 90 days.
Content in this thread may no longer be relevant.
Perhaps it would be better to start a new thread instead.
Back
Top