Thread Tools
Nov 22, 2013, 03:39 PM
Pattern Pilot
Thread OP
Question

Forecasting too high a ripple ?


I've been surfing this forum on how a too high ripple voltage will impact an ESC. I think I understand how longer leads will impact the inductance and thus the ESC and the solution is to add caps inline.

But I also see a lot of other things like high resistance (tired, low C rating) batteries, poor connectors (again high resistance), arming systems (yet more resistance) and longer leads (resistance and inductance) causing the ESC fits.

I've seen answers for longer leads limits before inline caps are required, but my question is:

Is there a definitive approach to adding up all the resistance up stream (ESC to battery) that one could use to determine that the ripple ration could be in jeapardy?

Or, am I over simplifying a complex issue?
Sign up now
to remove ads between posts
Nov 22, 2013, 05:12 PM
A man with too many toys
As I recall someone from CC said that ripple over 10% could cause FET failure if running close to the max current limit. In that case the ESC would just stop working. I have never had any problem getting over 10% on any of my airplanes or helicopters and a couple of them have over a foot of wire. Run some tests and look at the data log – you may not need any extra capacitors.

You could calculate a theoretical ripple number but measuring all the component resistance accurately is difficult and takes some expensive equipment. Much easier to just run some short bench tests and look at the data log.



.
Nov 22, 2013, 05:14 PM
A man with too many toys
One other thing to consider is that the ripple depends on the PWM frequency at part power and commutation frequency at 100% throttle.

.
Nov 23, 2013, 10:58 AM
Pattern Pilot
Thread OP
Quote:
Originally Posted by RC Man
...

You could calculate a theoretical ripple number but measuring all the component resistance accurately is difficult and takes some expensive equipment. Much easier to just run some short bench tests and look at the data log.


.
I have been faithfully following the data log, but that is more "real time". I was thinking more a long the lines of "predicting" a troublesome situation.

Let me explain where I'm coming from. Back in Sept. I went through three Phoenix ICE2 HV80 ESCs in about a three weeks time. The first occurred the day prior to a contest at the contest site. I had just taxied out, turned at the end of the run way; no problem. As I started to throttle up to take off, I was greeted with a pop, a belch of flames and smoke. The plane a little charred but OK. The ESC toast. That was ESC #1.

A friend at the contest loaned me his spare HV 80 which worked just fine for the rest of the weekend. The friend suggested that I keep his ESC until the next contest. Which I did and while practicing the next week his ESC started going into "hard voltage cutoffs". Same result with every battery I owned, new and old. So much for #2.

An other friend let me use his HV 80. That worked for about another week, until it just flat out quit, with no warning. #3 gone.

By this time I called CC for help. After he (Bernie Wolfard) listening to my plight, suggested that it was my connectors explained the effects of ripple voltage. I had been using Anderson Power Pole 45 amp. connectors exclusively. I then replaced them all with Dean's and this is where I am currently.

I am flying with ESC #4 the replacement from CC for #1, it was still under warranty, and Deans connectors. So far so good.

I still follow my data logging religiously, but with much more attention to the ripple.

I started this thread with the intention of a better understanding of ripple.
Nov 23, 2013, 11:37 AM
A man with too many toys
The new Edge has new hardware and larger capacitors. It would be nice to hear is any of them have a ripple problem.

It would be nice to get someone from CC involved in the discussion but it seams that they don’t support the forums much any more.

?
Nov 23, 2013, 02:22 PM
Pattern Pilot
Thread OP
Quote:
Originally Posted by RC Man
The new Edge has new hardware and larger capacitors. It would be nice to hear is any of them have a ripple problem.

It would be nice to get someone from CC involved in the discussion but it seams that they don’t support the forums much any more.

?
Yes, I was hoping that Bernie would jump in here!!! It is also my desire to help nay sayers understand the issues that they themselves impose upon ESCs.
Last edited by BuffBuilt; Nov 23, 2013 at 02:28 PM.
Dec 03, 2013, 11:31 AM
Castle Support
Clintstone's Avatar
Guys I have been on vacation and Tech Support usually answers emails first which are backed up from the Holidays...

Below is information I pulled out of an Archive that Lee Estingoy wrote to help everyone understand ripple. Thanks


"My battery did what?
An exploration of ripple voltage

Mysterious events are often attributed to mystical causes and brushless power systems are about as mysterious as things get in RC. Some systems work and others don’t. Why? The usual explanation is something along the lines of “It’s a mystery!” Don’t get me wrong, the reason for a component failure was no doubt a mystery to most involved, but understanding a bit more detail about brushless systems can go a long way to helping a hobbyist enjoy outstanding reliability in an electric plane or heli.

There’s a lot going on in a brushless power system. The motor, controller and battery are truly connected to one another and they all need to be in a sort of equilibrium. Any “weakness” in one can lead to a failure somewhere else in the system. The internal resistance of a battery is one of many such factors.

Unfortunately, it is rather difficult for most hobbyists to accurately determine the suitability of a pack for an application. Wildly varying definitions of C rating from the battery vendors, rather large differences in ESC “overhead” between the brands, and the fact that few setups are truly identical, all conspire to confound the hobbyist. The knowledge and tools needed to measure the system haven’t been a part of the average RC flightbox.

Using a high internal resistance battery in an application where it is being worked too hard can shorten the life or limit the amp capacity of a controller. High internal resistance results in a bit of a “wobble” in the output voltage of the pack under load. This voltage swing is NOT the sag in voltage most users can identify with their wattmeters when the pack is put under load. It is actually very quick swings in voltage that occur every time the ESC switches a FET on/off, and this can happen up to 20K times per second! Laboratory grade oscilloscopes can capture this voltage swing and display a nice graph of the input voltage. That graph would appear to have a “ripple” in it. No wonder then, that the phenomenon is called “ripple voltage” amongst ESC designers.

How does ripple voltage affect a controller?
An ESC is fundamentally a bunch of switches that open and close electronically – there are no moving parts. All the switching takes place inside electronic components known as FETs which are basically silicon connected to legs that connect to the outside world. Silicon is alternately conductive and resistive depending on the application of a controlling voltage. This controlling voltage is applied to the gate leg of the FET and it generally must be about 10 volts higher than the output voltage of the line being switched in order to cause the FET to switch to complete the circuit.

Things get a bit foggy here in a haze of technicality, but bear in mind that FETs can also be in a state that is somewhere between completely open and completely closed. A partially conductive FET is much more resistive to current and that resistance is ultimately transformed to heat. Heat is a bad thing in an ESC; excessive heat is harmful to the components and it can ultimately kill the controller.
FETs go through the partially conductive phase every time the FET switches from on to off. This is expected by the ESC. While it generates heat, it is not likely to interrupt the operation of the ESC. The partially conductive state also occurs when the voltage on the gate leg of the FET comes close to that 10 volt spread over the battery voltage, and that can occur either by raising the voltage on the gate leg, or by dropping the voltage on the source and ground below the voltage sitting on the gate.

Ripple voltage can serve to drop the voltage on the source leg while the voltage on the gate leg takes a bit longer to drop due to some factors related to the internals of a FET. Bingo, the system is in a great position to have unwanted operation of the FETs. Simply put, ripple voltage can cause the FETs to enter the partially conductive state which can allow currents to go places that the controller doesn’t intend for them to go. The worst case scenario is that the FET can be triggered at the wrong time which creates what amounts to a dead short between the red and the black wires from the battery. This will definitely lead to a bad day at the field. More frequently, it can create much more heat in the controller assembly.

ESC manufacturers do use capacitors to even out the voltage coming from the battery. This works to some extent, but there are practical limits to the capability of capacitors to mask the underlying issue – the battery is being asked for more power than it is suited to deliver.

Explain Battery Resistance in 100 words or less.
Think of a battery with low resistance as a balloon filled with air. When the balloon is popped, the air rushes out in what appears to be a single burst. A battery with high resistance can be compared to one of those inflatable Godzillas. Puncture the dinosaur and there is a brief moment of high pressure air release and then a much longer period of much lower air pressure as Godzilla slowly crumples to the floor.

This effect goes both ways a battery that has a voltage sag likely also has a hard time recovering from that sag. Imagine a video of Godzilla and the balloon being played in forward and then reverse. The “inflation” or “recovery” would take a longer time for the Godzilla than for the balloon, the same is true for high internal resistance batteries, they take longer to return to voltage after a load than low internal resistance batteries.

Right, that’s much more than 100 words, but when’s the last time you saw a reference to an inflatable Godzilla in a modeling magazine? Anyway, this sort of rush and then slower flow happens in your electric power system every time the FETS are switched and it can lead to the unwanted partial triggering of the FETS conductive property.

Teaching old dogs new tricks
Remember when I said that the tools to scientifically measure a battery’s suitability for a task weren’t in an RCer’s flightbox? Well, now that onboard ESC data logging is available in Castle’s ICE controller line, hobbyists can gain a better measure of their system’s health. ICE controllers have the ability to measure and record battery voltage at the beginning and end of the FET cycles and display that voltage swing in terms of volts in Castle’s Graph Viewer program on a Windows PC. Users don’t have to pay extra for any of this, it’s just part of the ESC package. Not very often that a tool falls into your flightbox for free!




So, how much ripple voltage is too much?
Sadly that can’t be answered with a clear cut value. The harm to an ESC from ripple comes mainly in the form of extra heat, and a controller’s ability to function despite the additional heat depends on a huge array of factors including ambient air temperature, cooling air flow, and throttle position. Keep in mind that an ESC also generates far more heat when it is operating at partial throttle than at full throttle.
Great, so how do you use this magic number? Ripple voltage is best used as a reference point to compare a pack with itself over time or to compare multiple packs in the same vehicle. The one with the lowest ripple value is the best bet system for the application. Batteries with higher ripple voltage in one application may work just fine in less demanding applications.


Caveats:
Some motor types, particularly high inductance inrunners, may actually generate enough electricity while running to reduce the ripple voltage. This may skew the results reported in the data log. Please use the results from these motors in comparison ONLY with other motors of a similar type.
Ripple is best measured at or below 50% throttle as this is the time where the FETs are switched OFF long enough for the battery voltage to recover between pulses and therefore create a swinging voltage differential. "
Dec 06, 2013, 11:02 AM
Pattern Pilot
Thread OP
Quote:
Originally Posted by Clintstone
Guys I have been on vacation and Tech Support usually answers emails first which are backed up from the Holidays...

Below is information I pulled out of an Archive that Lee Estingoy wrote to help everyone understand ripple. Thanks


"My battery did what?
An exploration of ripple voltage

"
Thanks Clintstone/Lee. This is the kind of information I've been craving. Deep enough to get to the guts of Ripple, yet in laymen's terms that I can understand; Gorilla bags? I can see that.


Quick Reply
Message:

Thread Tools