Revisiting the capacitor mod

[K2Freak]

Hey all,

Happy 4th of July weekend.

So I am a bit baffled by my apparent failure to significantly reduce my ripple current by adding capacitors. The first time I replaced the stock 2 x 50V 330 uf with 3 x 63V 330 uf that were a different brand. They were the same color green and looked great, but the result did not change much. 8^) I dropped from peaks of ~3.6 - 4.0V to about ~3.2 - 3.6V. I chalked this too small reduction up to possibly poorer quality caps and not enough different in capacitance between my new bank and the original.

So I bought what I thought were some good ones - Panasonic FC 63V 470 uf and put a bank of 4 of them on. Everything worked great for the process, and they were smaller so the implementation looks neater and has more strength. But I just took it out in the yard and did some pitch pumps, and saw 2.3V peak ripple. From what I have heard this is a lot closer to the acceptable range, but was expecting less this time given how much I have increased the capacitance of the bank. It is about 3 times the stock value. I put another bank on the tail for reference in the pic.

Clearly I am no EE, I am just copying what I see others doing out here, but I would really like it if somebody that can explain more about this would chime in. I understand that the batteries are part of the equation (a big part), so if that is the limiting factor then I guess I will live with it. I have seen no change over time though, so I don't think it is an indicator that the packs are wearing out. Just maybe that they are not so good. 8^) This test was my ICE HV80 with 2 x 4000 mAh Blue Lipo packs. These packs are about 25-30 cycles old and perform great.

I also have a heli that runs 5000 packs on an ICE HV120 that shows me ripple of as much as 5V. That one is my next target, but I am going to hold off until I learn more. It has flown for 9 months this way and I suspect that it will last a while longer...

Help! 8^)

[Inverse]

How are you stacking your caps?
Parallel, series?

[K2Freak]

All the cap pluses combined to the ESC red, all the cap minuses combined to the black. Guess that's parallel.

[Tisondm]

Did you remove the original Caps and replace with the ones in the pic. That's the way it looks in the pic.

[K2Freak]

I removed the originals because one day there was no spark. I assumed that meant they were shot, and decided to disassemble the 80 and do the mod.

I also really liked the idea of going to 63V instead of stock 50V, although I later read a couple of Patrick DC posts that say this is not really as good an idea as it sounds like.

[f22warzone]

you have to figure to drop my substantially I had to add 3x the amount plus the original
on an hv 160 you have 4x 470 then I went and added 6x 1000uf
original capacitance = 1880uf
new ( -1 original as it broke) = 7410
so almost 4x the original capacitance to drop it down from 3.5v to .2v plus using 65c tp packs

also you have to remember your ripple will be really high during like the first 30 seconds, then will drop down. Also you are nowhere near the 7410uF that I have so to see yours go down at all is a good sign.

another tip is increase your % esc usage, at a lower % out you will have higher ripple. when mine is up near 100% out there is almost no ripple, and that's how I have mine geared to run at about 90-95%

also yes it is worse to go with 63v caps to drop the ripple curent as they have a little higher ESR, and they are bigger compared to their 50v brothers...... but if you go 63v your caps will last longer

[K2Freak]

So maybe my expectations are inflated? I guess if it stays under 2.5V I'll be happy - after all that is a significant reduction from the original numbers. Need a few flights to see what the bigger picture is - sounds like 15 seconds of pumps on fresh battery may not tell the whole story.

So is ripple more of a "peaks do damage" thing or is it more of a "cumulative" or "long-term average" phenom?

And should I add more than 4 x 470 to the 120? What about the working caps on the 120? Remove or leave there? I have heard not to mix the voltages, but anyone have more info on what would be the best approach?

[firehopper]

what frequency is the ripple? higher freq ripple needs smaller value caps. sometimes putting a 50-63v .1 ceramic or using a smaller value tantalium cap may reduce ripple more.

[K2Freak]

what frequency is the ripple?

Sorry - I have no idea. I don't think we log that, do we?

[f22warzone]

would it not be the same as the pwm frequency? or 3x the pwm frequency for each phase

[Tisondm]

Originally Posted by Tisondm

Did you remove the original Caps and replace with the ones in the pic. That's the way it looks in the pic.

Originally Posted by sacolvin

So maybe my expectations are inflated? I guess if it stays under 2.5V I'll be happy - after all that is a significant reduction from the original numbers. Need a few flights to see what the bigger picture is - sounds like 15 seconds of pumps on fresh battery may not tell the whole story.

So is ripple more of a "peaks do damage" thing or is it more of a "cumulative" or "long-term average" phenom?

And should I add more than 4 x 470 to the 120? What about the working caps on the 120? Remove or leave there? I have heard not to mix the voltages, but anyone have more info on what would be the best approach?

From what I've read you need to add 2 additional caps per 4 inches of wire over the alotted 12 inches per CC. The caps on the HV160 are huge and I haven't had to add any caps. My ripple current starts out at 1 volt and then drops to about .6 but my pack wires are only 7 inches long.