ESC Amperage Ratings: Does Continuous Current MEAN Continuous? Name Brand ESC Tested ZTW Beatles 30a
ฝัง
- เผยแพร่เมื่อ 16 พ.ย. 2024
- This video is a continuation of previous testing of no-name cheap ESCs found on Amazon. In prior testing we tried to establish whether the peak continuous ESC rating was accurate and the test revealed the ESC could not maintain it's stated peak continuous current rating.
The obvious follow-on question came up: "Will name brand ESCs achieve their stated continuous current rating?"
I have an answer....in the video.
![ร้อยพ่อพันแม่ - WanMai [Official MV]](http://i.ytimg.com/vi/XY2pNYASEbg/mqdefault.jpg)
This is the result I was expecting as I have attempted to keep my throttle down below the amperage setting and have had ESC fail when I realized that the amperage rating is at full throttle if you have the max amperage rating at half throttle. You would theoretically need a 60 amp ESC this was my understanding and you have proven it correct! Thanks for the lesson knowledge is always beneficial.
This was definitely a cool test. It seems obvious in retrospect to design for peak power at peak current, but it is very neat to see theory manifest itself on the bench. Thanks for the comment.
Very interesting series of tests.
As you came to your conclusion the obvious dawned on me that the power rating of an ESC can only be archived be at full throttle.
In an actual model set up you specify each component ie motor, esc, prop and battery to demand no more power than the weakest component in that chain.
So by attempting to run the esc at 30Amps at part throttle you have in effect over propped the power system thus making it out of balance.
With regard to the Mosfet partial duty cycle I would suggest that the Fet's would only be expected to handle around 10-15Amps at around half throttle so anything over and above that would cause overheating.
Great thought provoking test series though, which I really enjoyed... not as much as your Open Tx series though! :)
This was definitely a learning experience. I'm going to put the cheap ESC back on the stand with this new understanding to see how it does.
I was glad to see the ZTW meet its specification because I use them a lot.
Awesome video!... I've been a ZTW fan and will remain. Thanks for the video John.
You bet. Thanks for watching. I'm glad we got some closure on this.
Well look at what the prop dragged in!! ?Tim was on target. Wide open throttle at 30 or so amps(with the right prop). Can't wait for the re-run. Interested to see what prop would be used with a larger ESC. Cool 😎 stuff. AirHammer out!!
Yeah, I'll do the cheapie tomorrow. The prop was ok, it just had the effect of moving the throttle curve to peak @ about 80% vs 100%.
Actually, there is a little more to it than just the MOSFET duty cycle. When you are at about 50-75% duty cycle, the MOSFET switching is causing "ripple voltage" in your battery voltage bus which is causing the capacitor(s) in your ESC to heat up and can actually blow an ESC over time. Let me explain.
All batteries actually resist the flow of electricity though themselves. This property is called Internal Resistance (IR) and will cause the voltage in your system to sag under load. You can calculate this voltage sag with Ohm's Law (V = I * R). A "typical" mid-grade LiPo battery will have an IR of ~10mOhms/cell (the best available will be in the 1-3mOhm's/cell range). Since you are drawing 30A for this test, a typical 3S LiPo will have an IR of ~30mOhms (3 * 10mOhms) and you can expect a voltage sag of 30A * 30mOhms = ~0.9V under constant load.
BUT, your MOSFETS are actually switching on and off at an incredibly high rate to command the motor to spin, so you are not seeing a constant load until the duty cycle approaches 100%. If you give the ESC a 30A load at around 50% duty cycle, the battery is actually seeing and "instantaneous load" of 60A while the MOSFETs are "on" and 0A while they are "off", so your batteries will try to sag ~twice as much while in "on" periods and recover to their resting voltage during "off" periods. Any capacitors connected to your ESC (or to the voltage line in an external Cap Pack) will try to smooth this out by dumping extra electrons into the circuit during the "on" periods which takes some of the instantaneous load off the battery, and then they will act as a load during the "off" period as they recharge. This helps the battery see a more consistent load over the entire active period.
ESCs are generally designed to see max load at or near max throttle where they have ~100% duty cycle and relatively low ripple voltage. However, if you have a prop on your plane that is too big (or gearing that is too tall for an RC car), the ESC can start to see a really high load at lower throttle duty cycles which will cause the on-board caps to kick in to fight the system's ripple voltage. Caps are "different" electrically, but they have a "resistance" term just like batteries do called their Equivalent Series Resistance (ESR). As the caps kick in, charging and discharging to try to fight the ripple voltage, they start to heat up based on their ESR and the current they are being asked to handle by the circuit (Heat_Power = Current^2 * Resistance). If the ripple voltage in a circuit is too high, the caps (and the rest of the circuit board) will heat up very quickly and the MOSFETs can actually pull even more current during the low points of the ripple cycle, causing even more heating. If left unchecked, the ESC could kill itself. I'm into RC car Speed Runs, and guys kill ESCs this way all the time (mostly noobies).
So, overloading at part throttle is really, really bad for ESCs. Well designed units from quality makers will usually have some way to protect themselves from damage that will not leave you with a dead-stick and near-certain crash, though, like the pulsing you saw on the ZTW. I'm sponsored by Castle, so I'm thinking I may want to do a test to see how their ESCs handle the same test. 🤔
BTW, I have that same watt meter. I LOVE that thing. It's works great and is AWESOME on camera.
I hope this was helpful.
What set of circumstances in this test specifically created an "overload"?
Current was (margin of error) ~30a. As long as the current draw didn't exceed specification what is the definition of overload? If the answer to that is: Drawing maximum current at < 100% duty cycle (throttle) is an overload, that brings up a whole different series of questions like: If overload occurs anytime max continuous amperage is attained before reaching 100% throttle input, then why would any ESC have a burst rating? That would imply the stated continuous current is attainable at some level of power < 100%.
I want to get some clarity on the use of the term "overload" because it gets bantered about. So I'd like to know precisely what condition occurred that created an "overload". From the ESC/Battery point of view, the load never changed. It was 30a for both tests. The only variable was throttle percentage or duty cycle yielding the additional downstream impacts like ripple voltage and more active capacitors.
I know capacitors are used to smooth voltage flow on circuits, but the real question is: "What was the specific trigger that resulted in pulsing?"
Did the pulsing trigger have anything to do with capacitor activity? Is the controller monitoring capacitance cycles? I don't believe it is. I think this was a thermal control based on the fets getting hot as a result of their duty cycle. I think the controller saw an unsatisfactory temperature on the fets and the logic is designed to pulse the motor significantly reducing the duty cycle while the fets recover.
I never knew it was a problem to run an ESC at capacity at a throttle level less than 100. I've never designed a system that way on purpose, I stumbled onto this topic because I wanted a test environment that would let me create a very specific current draw.
PS The range on the servo tester is wonky from ESC to ESC even with a calibration. In this case, the effective range was 83 - 210. The test run resulting in pulsing saw the sweep at 193. That is 110 steps of 127 or 87% throttle.
@@RCVideoReviews I never meant to imply that your test purposely "overloaded" the ESC. It's just something that happens when a brushless ESC is asked to produce an output near it's max power rating at less than ~75% throttle. I'll explain the cause of this in more detail below, but it all comes down to the "I-squared" heating losses (P=I^2 * R) during the "on" portion of the 50% duty cycle. Unfortunately, it's really easy for guys (including myself) to do this in the real world and not notice it until something overheats... unless you happen to have a data logging ESC.
Also, I actually think you performed a great series of tests. I was simply trying to clarify some of your findings based on my experiences as a driver/pilot sponsored by Castle Creations & Venom Power.
BTW, I'm actually an Aerospace Engineer by trade, even though my main RC discipline is in ground-based RC drag and speedrun applications. It's really common for us on the ground-side to overload our systems and blow ESCs and motors by over-gearing. I review a LOT of castle data logs for community members (which is how I first got Castle's attention) and have a number of tutorials on TH-cam and Facebook trying to teach guys how to use their data to tune their setups and one of the first things I tell folks is to watch their ripple voltage and temps like a hawk to make sure their ESCs are properly loaded along with providing guidance about how much ripple is too much.
Castle's data logging ESCs produce high fidelity graphs of your ripple voltage along with all the other key ESC parameters including throttle position, current output, real-time voltage, load %, etc. Max ripple usually occurs in the 50-75% duty cycle range and falls to almost nothing once you get to ~85-90%. Since Amps = the number of electrons flowing per second in the circuit, pulling 30A when the ESC is at ~50% duty cycle actually means your Watt meter is just seeing an "average" of 30A, but the ESC is alternating between 60A and 0A really, really quickly. Since heat is generated with the square of current, (P=I^2*R) this means that the ESC will actually be creating 4X its rated waste heat output while the MOSFETs are "on" and 0 while they are "off". This averages to 2X more waste heat output during the 50% duty cycle 30A test than at the 100% duty cycle 30A test. This is why the ESC is much more likely to overheat or damage itself if it is highly loaded while the duty cycle is still low (and the ripple voltage is correspondingly high). From talking to my contacts at Castle, dealing with high loads at 50-75% duty cycle is one of the most stressing thermal conditions (as your test showed), so the ability to achieve a "burst output" above the rating at 100% duty cycle is probably due to the extra margin designers have to factor into a system to survive the lower duty cycle cases.
Here's an example from my world, Castle's Mamba XL surface ESCs will deliver 300 - >600A on 8S power. Castle experienced a serious problem in their last 2 generations of Mamba XL ESCs (the XL2 and XLX) because guys were setting up cars to pull near full power at ~50-75% throttle. This caused the on-board caps on the XL2 to quickly get overwhelmed by ripple voltage under certain conditions and blow which caused the ESCs to blow immediately afterward. They addressed this by installing ~7,000uF of caps on the XLX (which is a LOT of capacitance) which greatly improved the ripple voltage issue, but that allowed the ESCs to literally pull enough current to melt the circuit boards, so they had to do a complete redesign on the just announced XLX2 to basically handle as much current as any set of high output LiPos can provide.
I think you are correct that the pulsing was triggered when the ESC temp exceeded a preset limit. I mentioned the capacitors because they are put into the power system to smooth out ripple voltage swings, but they are generally completely passive devices. The heat they produce when they are working hard just added to the mainboard temperature and helped to trigger the "high temp" state which probably used the pulsing as a way to reduce the load on the ESC while giving you enough power to stay in the air. I think this "pulsing" is actually a really good idea vs the "all off" mode my surface based ESCs go into when they overheat.
BTW, I'm a big fan of your videos and think you use a very good, methodical approach to testing. I try to do the same thing on my tests. You've actually taught me a TON about my new TX16S, so I'm really appreciative for what you do.
I appreciate the well reasoned reply. I'm simply trying to get a definition of "overloaded" because people can define this to mean the prop was too big for the ESC. In a practical sense, of course it was, but this was a test. The prop was simply a load generator designed to draw current at a defined rate. As it relates to current, the ESC has no idea if it's a 10x10 prop, a 16x12 prop, or a 20x5 prop, it merely sees and attempts to satisfy load governed by operator input.
The real question here isn't strictly about current or the load, the question is the relationship between the current and the duty cycle + associated electrical realities of high current and less than peak duty cycles. I'll stipulate there are multiple additional factors as you've graciously pointed out (thanks for that), but I believe the conclusion remains in-tact: The negative side effects of a suboptimal duty cycle resulted in a failure of the ESC to maintain rated continuous current.
I appreciate all of the extra theory and am truly interested in understanding. I think we have common ground with agreement on a thermal shutdown due to the negative effects of peak current at suboptimal duty cycle--yes?
It's been suggested I "overloaded" the ESC because the prop is ridiculous but I disagree with the claim the prop had anything to do with "overloading" the ESC. Rather, the duty cycle ran hot because I hit the peak sustained current ahead of a 100% cycle.
Your definition in this instance is: "when a brushless ESC is asked to produce an output near it's max power rating at less than ~75% throttle." You clearly are not referring to "overpropping" per se, rather hitting peak current @ less than full throttle regardless of prop size.
In your response: "Max ripple usually occurs in the 50-75% duty cycle range and falls to almost nothing once you get to ~85-90%. "
I think the throttle was ~87%, so that brings up new questions.
PS Glad the radio videos helped you out :)
@@RCVideoReviews Are you on Facebook? I can send you some plots where you can see how the ripple varies with load/duty cycle. This behavior probably varies a bit with manufacturers and specific designs, but I believe it's generally true to at least some degree for any PWM modulated motor driver. Some of Castle's designers are active on the Castle Creations Fan Facebook page. I'll share this video there and see if any of them would like to chime in.
Part of testing is setting conditions that will show you the edges of the performance envelope. I think your test was great because you both verified the claimed steady-state performance and uncovered how the ESC behaves when pushed really hard. You may consider making that a standard test so viewers know which ESCs have hard-cutoffs and which provide a "limp home" mode.
I would probably define "overloading" as any thermal condition where a system cannot achieve steady-state thermal equilibrium. You could test for this by loading up a motor or ESC and measuring its temp over a short interval. If the motor or ESC runs up to a certain temp and levels off, it's hit a stable operating point and not overloaded. If the temps rise beyond a safe zone (with some safety margin added in), it is not at thermal equilibrium and therefore "overloaded". How does that sound?
Great test and some extremely informative reply’s in the comments section. At the end of the day it helps to clarify the importance of a watt meter when setting up a speedy/motor/prop combo. I know it’s saved me many times. Thanks for video. 👍
Ding Ding Ding--we have a winner! Great comment, and you're definitely on why I take the time to post this stuff.
Thanks for watching.
J
4:49... It takes time for the heat from the Mosfet die to go through the Mosfet casing to one side of the Heat Sink material and to the outer side of the heat sink and through the ESC Package Heat Shrink to the outside of the Heat Shrink after you remove power. Heat propagation...;-)
Yup.... Now is the time to re-test the Cheap Yelllow ESC's again. You now will compare Oranges with Oranges on the additional test.
I will. Gotta be fair.
I have some multirotor Super RACERBEE 30A ESC Supporting 2-6S -30A in NO Airflow that I use in a twin motor Skysurfer and soon in a DIY E-flite Twin otter " Replacement parts" build. In fact I zip tie the two escs together, they are inside the fuse and I run longish motor wires and short battery wires. I am wondering what gives? I have had even 60amp airplane esc that cant handle what these "micro" esc can. I also recently bought a Volantex S1600mm motor glider with a known poor choice of 1400kv 2210 with a 9 in. prop and a 30amp esc. Blazing heat in both the motor and esc. I chanced to a 40amp multirotor esc and a separate bec which solved the esc heat. So on most of my planes I pull the red wire from the esc and use a separate bec.
Pull the ground too to avoid ground loops.
I just got back from a flying session this morning with my Sky Surfer with the Sunnysky 2212 2450kv motor and 6x4 prop, Hobbywing 30A ESC and running 3 cell batteries. Yes it howls like a banshee on that combo, but full throttle runs usually draw about 24-25 Amps on the bench so in the air its a bit less no doubt, but on full throttle sustained the ESC doesn't overheat (its reasonably warm after tho) and I usually run into battery sag where the voltage drops off rapidly under load after about 4-5 mins on a 35C battery pack. I can bounce off the sag point as I have a battery voltage warning set at 3.7 volts per cell (and at 3.6 volts I have a critical warning) which the value is called out on my TX16S. My results pretty much match yours, except yours are bench based hence current draw is likely to be higher as a result base on motor / prop load. Thanks for sharing the MOSFET duty cycle info, it completely makes sense!
Awesome! Glad to see practical correlation. It's neat to see all of this tie together.
Great video.
Btw, is there a test for HobbyWings ESC?
Actually, for safety reasons, in most cases I always used an ESC with higher amp rating for the application; usually having 30% extra as the buffer.
I haven't done one. This was more of a follow-on to a prior pair of videos on cheap ESCs.
The amp of the esc is only 20 or 30ah or is there also 5a, 10a, 15a, .... 55a, 95a, ...? And the amperes of the motors should usually be less than the amperes of the esc?
This is a 30a ESC. I think ZTW has a few different options, like 40, 60, 80.
The second statement is argumentative. Most in-the-know types will say a static motor test shows static load, not dynamic load. Dynamic load is generally less than static because props unload in flight. In a static test they cannot because the stand is anchored to the bench.
My personal practice that has served me well is to have minimum 10% overhead on an ESC over peak static demand.
So, it looks like you only get a continuous 30A under specific conditions? I always try to have excess capacity in my Battery and ESC so I assume under those conditions this would not happen. Thanks for all the testing as it's very informative. See you in the Air!
Bottom line is design your ESC to reach peak at peak throttle and you'll be good to go--at least on ZTW. I still need to retest the cheap yellow 30a.
Excellent series. Does this mean that the “Burst Amp rating” is a fairy tale? The “Continuous Amp rating” can only be exceeded for a limited time (10 seconds) with the throttle near max to avoid problems.
Using the setup in the last test shown, could you reach the “Burst Amp rating” at all?
I don't think so, because remember the pulsing started at about 60 seconds at partial throttle and max amps. The burst rating is almost always around 10 seconds. In fact on the 1st cheap ESC I ran it to burst for longer than that and it didn't die.
But you do bring up an interesting point: If you want to take advantage of burst amps, then your max continuous will have to be lower than 100% throttle.
I think just to keep things simple, I feel the best bet is to design for max continuous or lower. Then you don't have to start wondering if the duty cycle is going to cause bad behavior from the ESC. This thing ran straight up at 30a for over 7 minutes.
Would more voltage run it at the same speed with less amps?
There's a direct relationship between voltage and Kv. More volts = more spin with everything else being equal.
@@RCVideoReviews what if you add more volts but less amps?
The power consumption of mosfets does not change with duty cycle. It changes with frequency. With higher frequency you get switching losses. Driving motor with same current at higher throttle will make no difference.
I'm not suggesting power consumption changed. I'm saying at slower RPMs the mosfet duty cycle is increased causing temperature increases. In a partial throttle state the fet is in linear mode. In linear mode they heat up and become more resistive. This manifested itself quite clearly in the test with pulsing at partial throttle and 30a while running clean for 7+ minutes at full throttle and 30a.
dude first time you gote more then 1 A more and it got in safe mode