Quote:
Originally Posted by Pdelcast
Jeff,
You have to realize what we (as a company) are up against. There are battery manufacturers that sell batteries we would rate at 15C, but they label them as 35C. And then there are other companies who sell batteries that would easily rate 40C, but they label them as 30C. And there are companies that sell 30C cells that should be rated 15C, and 35C cells that are just reasonably rated.
So how are we, as a company, supposed to inform the customers? If we tell them "Don't use brand A, because they lie about their ratings" and we have a lawsuit (and yes, we can be sued by a company based in Hong Kong, but we can't sue them...) If we tell them, "Use brand B, because they are good quality" and we have OEMs who sell their own batteries at our throats. If we tell people how to determine if their batteries are good quality, their eyes glaze over and we lose them to some Chinese company that sells a crap speed control, but doesn't tell the customer that they have to research/test/whatever their batteries before they use them.
So we give a nice general guideline, in a slightly humorous pamphlet. If the customer has a problem because he insists on using junk batteries purchased online from China, we give them a warranty replacement and a lecture on using good batteries and reasonable pinions, and warn them that the next failure caused by bad batteries will not be warranty.
Meanwhile, we continue to improve the controller as we see different failure modes, and make it more and more bulletproof. Nobody even knows that we are now shipping Version 4 of the MMM -- with a few minor changes to make them more reliable. Version 5 should be shipping the end of the year, with even more changes to improve reliability. Version 6 is in layout now, and will bring some more features, and more idiotproofness (that should be a word in the dictionary!) to the MMM.
So, our policy has been to be more "hands-off" about batteries - - support the customer, tell them to use good quality, name-brand batteries, but we won't call out brands specifically. And continue to make the controllers more able to deal with low quality batteries. In the long run, the market will drive the high quality batteries for performance reasons. Eventually even the cheapest batteries will have to perform adequately to compete on the open market.
What would really help is if people would stop encouraging the cheap junk entering the hobby market from direct-from-china internet based stores. They have no accountability, and can supply low quality goods with no chance of liability for the damage their junk causes.
Just my $.02
Patrick
|
Patrick, I recently picked up a Turnigy pack (4s 5Ah 30C) to see what all the hubub is about. I made sure it was balanced and all set, then ran it in my 8th scale buggy geared for around 35-40mph. I was quite surprised to see the results were better than I had hoped for. It literally puts certain other lipo manufacturers/resellers to shame. No, they aren't the "best" out there, but still VERY good, especially for the price. And, how does your tech staff know the ESC was run on inferior batteries vs simply good batteries pushed too hard? I would think the result would be the same.
I agree with you though; you can't name names in this sue-happy world we live in, even if it is true. Also, it is true you cannot cover every possible setup issue in the manual; doing so would make it so large that no one would ever read it (except maybe in the bathroom
). Personally, on the next generation of ESCs that contain data logging (and therefore hall effect sensors), the firmware should be set up to continuously monitor current flow and voltage sag. If the voltage sags below a certain % of the nominal value, the ESC would automatically dial in punch control to lessen the bursts. Likewise, it can also monitor reverse current (provided a bi-directional HE sensor is used) and voltage on braking, and if the voltage and/or reverse current is too high, have the ESC automatically dial down brake force. Yeah, would make driving a little more unpredictable, but I would rather err on the side of caution. I suppose you could make this feature defeatable, but provide a disclaimer that any warranty claims are denied if disabled (and could look at the ESC log files to verify on return). Would give people the choice, but discourage random setting changes just because Joe Sixpack wants to do double backflips with inferior batteries, or is using good batteries, but is geared unreasonable.
Like Jeff and ticklechicken have asked: how does one determine their batteries are good or not using a logger? I personally look at voltage drop under load, and temps after a run as a guide. Loggers don't have sampling rates fast enough to catch ripple current, but I would imagine if v drop @ high currents is low, ripple currents will be low as well.
Quote:
Originally Posted by J57ltr
...Case in point, as I am sure you have read at least once we manufacture a small conveyor (among other things). It uses a stepper motor with integrated encoder and amplifier (read ESC), it also has digital and serial I/O. Basically we send it a position command (infinite in our case), Speed and Acceleration. Over the years we have had a few come back with the board damaged. About a year ago the manufacturer made a board revision that put a fuse in line with incoming power. On the output of the fuse is a Zener diode so if the voltage rises above a certain point it conducts and blows the fuse (there is also a TVS in there as well). Well I got in 4 units that had the fuse blown and the Zener shorted. On 3 of them I was able to replace the SMT fuse and everything was OK. The fourth though had a FET shorted as well. After speaking with their tech support I was told that a brand new motor only 17 days old (including shipping from Canada 2 Houston then to Ohio) would not be covered since the Zener was bad as well. I asked him why and what would cause this, he told me it happens during Regen. Basically the conveyor has solid SST rollers weighing about 12 pounds. We have a maximum Accel/Decel rate of 25 rps/s, which is quite slow. But if the conveyor gets jammed and then let go, it will speed up over our maximum speed and try to “catch up” with the position it’s supposed to be in. So basically it runs real fast then wham almost comes to a stop as it gets to position. As you can already tell the amount of force required to stop 12 pounds of weight almost instantly creates a large overvoltage and can take out the FET’s before the fuse can even react (besides fuses are S L O W, even fast acting ones). I asked what the fix was (besides their new motor we are working on the algorithm for), and was told I should need a “slave board” that has this protection built in. We never needed it in the past, since we run at a constant rate it hasn’t been an issue and we have used this motor for nearly 6 years. And you know what it’s in their book. The motor that we use is more suited for pick and place than the application we have it in now (we need something a little dumber).
Jeff
|
A little off-topic here, but I think the "ESC" should be programmed not to stop on a dime from the "catch up" time, but rather slow down over time. Same idea for the acceration part too IMO. In process controls, that is known as integrating. Regen voltage/current is a function of the amount of magnetic flux "stored" in the motor, the weight of the payload being forced to stop, and the stopping time. Since the first two aren't easily changed, that leaves the time factor.
Quote:
Originally Posted by J57ltr
Patrick,
I understand your point about not naming names, but it wasn't until recently that the little piece of paper came out about C ratings. Also what's the deal with NiMh batteries wouldn't those have the highest resistance out there and cause damage to the controller? I haven't seen anyone mention that.
And how does one determine if their batteries are of good quality? My eyes don't glaze over so easily. If you are refering to a test to perform, I would be interested in that for sure, but if it's the C rating paper then well I already read it.
Regards,
Jeff
|
As far as the NiMH issue; it's true they have higher resistance, but remember they simply cannot provide the current levels that any decent lipo can. I'm sure there are ripple currents with NiMH too, but since overall current is lower, so is the ripple. Of course, since NiMHs have a higher resistance, they heat up a lot more, which brings other issues, but that's not the point here. IIRC, a good NiMH can output 100A, but the voltage drops substantially. So, on takeoff, current peaks and voltage sags a lot. As the vehicle nears the set speed, current dwindles and voltage climbs back up. The effect is a more gentle acceleration. Kind of like a forced punch control when you think about it.