supercurio Posted July 21, 2022 Share Posted July 21, 2022 On 7/12/2022 at 4:14 PM, RagingGrandpa said: My poor-man's FEA says:Â I worked with your suggestions in mind and here's what I got! Thinking about starting to print the first test tomorrow: Â Â 2 Quote Link to comment Share on other sites More sharing options...
Popular Post supercurio Posted July 25, 2022 Popular Post Share Posted July 25, 2022 The first print is mostly a success, and the result seem really strong, probably thanks to its combination of design and being solid plastic where it matters. I managed to shove the expected 25mm steel profile in, but now I can't get it out because tolerances.. (one side is a bridge and sunk during the print, so it's not as wide as it should) I'll try to figure out how to fix that and print another one. But it's promising, I think that anybody owning a 3D printer with a 300x300mm bed, and some patience will be able to print these. 5 Quote Link to comment Share on other sites More sharing options...
Popular Post supercurio Posted July 30, 2022 Popular Post Share Posted July 30, 2022 (edited) Some good and bad news to report today! The good The design is validated! The bracket appears to be way overbuilt, and could probably be made with half the material or less. It's great to have loads of headroom on strength, but ergonomics might get better by making the brackets less wide. On the first try, I didn't really press on the pedal and the tire spun out. Next, I learned to put weight down to increase the tire's grip. It was possible to press either on the bracket itself or on the pedal. I ended pressing on the pedal, which could be made easier by a less wide bracket. Either way it was fine. The 1.5m 25mm square steel profile used allow provide plenty of leverage with ease, so that's also proven. The bad I'm very sad to report the first casualty of this pull force test, my Veteran Sherman original mainboard 😢. Since there was a successful measurement already made I decided to try the setup with it, and only try short bursts. Alas, that was still too much and it gave out, leaving me stunned and feeling pretty stupid to be honest; that was a bad call. Unfortunately I was not recording much during this concept evaluation, and have only a screen recording showing the phase current displayed in my app EUC Alarm. Since I was adjusting things.  Aftermath At least 1 MOSFET burned, and a small component nearby lost a leg. Both 30A inline fuse also blew which I was at least satisfied to see. It allowed to prevent the mainboard to burn further and avoided damage to the battery packs BIG FAT WARNING On 5/10/2022 at 6:54 PM, RagingGrandpa said: Will this hurt my EUC? If your EUC is working properly, no. If your EUC has defects, this could reveal them. It is a stress test. Mechanical or electrical failures could occur, so be sure to perform this test in a safe area outdoors, away from people and fragile objects. Your EUC needs to be able to tolerate its maximum current for a brief pulse, because this is exactly what can happen when riding the EUC. Better to discover a defect during a test like this, than by crashing in traffic! @RagingGrandpa proved that it's possible to complete a successful pull test with a Sherman with the right technique. I cannot stress enough (pun unintended) how brief must the pulse be. Tho not so brief that the sample rate of the scale and bluetooth data couldn't capture the peak values. I guess there's a good amount of practice involved, and I would recommend doing so with a wheel known to handle this case safely. But known that there is a risk of burning the mainboard, probably for most wheels depending on the testing technique. Since I designed this primarily for Adam (Wrong Way), I recommended him to complete this test only at the very end of his review process as it can certainly be destructive. Then after obtaining all the footage and experience needed for a full review, it's okay to run this test. Reviewers are usually sent wheels with prototype firmwares, sometimes prototype hardware as well which increases the risk of failure to this test. At the same time, if it ends up being destructive well it's still very much a valid test. Why? Because it means that the wheel evaluated will burn and cut-out in some real-world riding scenarios, like climbing steep hills, climbing not that steep hills for a heavier rider as well failing suddenly if the tire rotation is halted by a rock or mud. All things that we want the firmware to prevent happening by having reasonable current limits and a failsafe if the wheel can't rotate. Next I still plan to test the V12, V10F, 16X but I will start next week, once the frustration and self-reproach settled and I am fully rested to make better decisions. Then I'll send the gear to Adam! Edited July 30, 2022 by supercurio 9 4 5 Quote Link to comment Share on other sites More sharing options...
Popular Post Freestyler Posted July 30, 2022 Popular Post Share Posted July 30, 2022 (edited) Sorry to hear that Francois. How much is that little stunt going to cost you? Glad to hear you are committed to testing your other wheels as well, I just hope it won't be damaging. Â I have some updates of my own. I wanted to find out if my assumption of the max phase current in firmware was correct. I prepared a custom firmware for a forum member that reduced this value from the stock value of 150a (mcm5) down to 50a. He did a pendulum test and he immediately overpowered the wheel. So we have proven that this value controls the upper limit of phase amps. Â We then increased this value to the max value of 250a. Initially he didn't notice a difference in the pendulum test. On a follow up report he states that he feels that wheel is more responsive and it's like a 4th mega hard pedal mode was added. Frankly this could be a placebo effect, so take it with a grain of salt. Â We could perform a more scientific test RagingGrandpa if you want, since you have the setup. Edited July 30, 2022 by Freestyler 1 2 1 Quote Link to comment Share on other sites More sharing options...
supercurio Posted July 30, 2022 Share Posted July 30, 2022 (edited) 1 hour ago, Freestyler said: Sorry to hear that Francois. How much is that little stunt going to cost you? Thanks, 300-350 Euros it seems. The wheel just ran out of the 1y warranty. I hope this costly mistake will prevent others to experience the same! 1 hour ago, Freestyler said: Glad to hear you are committed to testing your other wheels as well, I just hope it won't be damaging. Yes. I'll be a lot more careful: Only short peaks, less than half a second. Never consecutive Turning off the wheel for 1 min in between each measurement. 1 hour ago, Freestyler said: I have some updates of my own. I wanted to find out if my assumption of the max phase current in firmware was correct. I prepared a custom firmware for a forum member that reduced this value from the stock value of 150a (mcm5) down to 50a. He did a pendulum test and he immediately overpowered the wheel. So we have proven that this value controls the upper limit of phase amps. That makes sense, kudos for finding and confirming this control! 1 hour ago, Freestyler said: We then increased this value to the max value of 250a. Initially he didn't notice a difference in the pendulum test. On a follow up report he states that he feels that wheel is more responsive and it's like a 4th mega hard pedal mode was added. Frankly this could be a placebo effect, so take it with a grain of salt. We could perform a more scientific test RagingGrandpa if you want, since you have the setup. Uh-ho, 250A is a scary big number, if you see what I mean. Certainly, measuring the pull force vs max phase current will allow to draw conclusions quickly. Does the pull force (as well as torque) increase linearly with phase current? Then, same for maximums. Edited July 30, 2022 by supercurio 1 Quote Link to comment Share on other sites More sharing options...
RagingGrandpa Posted August 1, 2022 Author Share Posted August 1, 2022 On 7/30/2022 at 1:32 PM, supercurio said: I'm very sad to report the first casualty of this pull force test Dang. Of those 17 wheels I've tested thusfar, none were harmed... and I've yoinked my batch-1 Sherman on multiple occasions. My personal opinion is your Sherm had a manufacturing defect, which you just hadn't discovered yet (likely because of a lovely, respectful, experienced-rider riding style). You could take it as a success story: your Sherm could have failed during a sudden high-torque maneuver while riding (pothole, etc), and because of this test you avoided a crash. Perhaps contact your replacement board seller and ask if they agree that a new board should tolerate overlean, and then test yours immediately after installation. I firmly believe all defect-free controllers should have functional current limiting, which is not damaging when the overload is brief and occurs from a normal (not overheated) starting condition. I think all new controllers should be stress-tested for safety reasons. My favorite method is repeatedly climbing a short grade (with rest time between attempts), until the controller reaches its alarm temperature. It should get hot, beep, cool down, and work normally. If that alarm system doesn't work, it's not doing what the specification promised, and is therefore defective.  And you left me in hanging suspense- what was the maximum pull force you managed?   On 7/30/2022 at 2:15 PM, Freestyler said: mcm5 increased to the max value of 250a. We could perform a more scientific test RagingGrandpa if you want, since you have the setup. Only if you supply a spare MCM5 controller to sacrifice   On 7/30/2022 at 3:39 PM, supercurio said: Does the pull force (as well as torque) increase linearly with phase current? Nearly, yes. But I think Gotway (for example) sets the firmware current limit based on expectations for durability. And Gotway has a reputation for brazen (reckless?) design choices, so I doubt they're holding back. I wouldn't trust those controllers to survive when "hacked" to bypass the original current limit. 1 Quote Link to comment Share on other sites More sharing options...
Freestyler Posted August 1, 2022 Share Posted August 1, 2022 2 hours ago, RagingGrandpa said: Only if you supply a spare MCM5 controller to sacrifice  My recommendation was for 180-200a, but the guy said he doesn't care if he burns his motherboard and wanted the max value. 2 Quote Link to comment Share on other sites More sharing options...
Leskont Posted August 6, 2022 Share Posted August 6, 2022 On 8/1/2022 at 11:14 PM, Freestyler said: My recommendation was for 180-200a, but the guy said he doesn't care if he burns his motherboard and wanted the max value. Yes, that's right)) And thanks to the Freestyler for agreeing to this adventure. Even I wouldn't do 0km/h tests with such firmware (250A - MCM5), because I understand that these tests have nothing to do with the actual operating conditions of the wheels. And I want to have a wheel under me that will pull me out of any situation or die, rather than hope that the engineers correctly calculated the phase currents and set the correct limit, after which, in a critical situation, the wheel will simply turn off))) 1 Quote Link to comment Share on other sites More sharing options...
Popular Post supercurio Posted August 10, 2022 Popular Post Share Posted August 10, 2022 I just measured V10F, 16X, V12 HS (two firmware versions) and S22 (two firmware versions)! Interesting results, and yes the 16X is so good at this 😄 I'm gonna look at the footage and compile all the results in a video. 6 Quote Link to comment Share on other sites More sharing options...
Popular Post mrelwood Posted August 10, 2022 Popular Post Share Posted August 10, 2022 (edited) First of all, I love data, so I applaud for the people who have participated in this. But unless I'm missing something here, I feel a bit iffy about the validity of this test.  On 6/1/2022 at 2:52 PM, mike_bike_kite said: I also added a small report showing fastest accelerating wheels I think it's a bit bold to call it "acceleration" when only the stall force at 0km/h was measured. Please read on.  On 6/1/2022 at 5:29 PM, RagingGrandpa said: And I truly believe stall current is the maximum motor current the EUC can deliver. Though you continue with: On 6/1/2022 at 5:29 PM, RagingGrandpa said: Only if they use the same firmware. Because firmware determines the current limit, and current creates force.  Isn't it known that the S22 limits the maximum power at the start, by a varying degree based on the firmware version? So at least the firmware version should be included in the results.  Back in the day, burning mainboards were a known issue whenever the tire got stuck by a stick in the wheel well or otherwise. Sometimes even when trying to start behind a stick. Wheels  have gotten a lot more powerful since then, but it didn't only move up the threshold, they actually no longer burn as easily even when the tire gets stuck. Which made this test possible in the first place. Any manufacturer that wasn't completely blind was well aware of the not uncommon issue of burning the board at zero speed. If I had been asked what to do about the issue, limiting the max current at zero speed might've been my first suggestion. So I'm confident that they have thought about it as well. I don't think anybody knows if the stall current is additionally limited or not on any wheel. This test is the first that touches this aspect of the power output, so if they are limited, it wouldn't be known from before.  On 6/1/2022 at 6:24 PM, yoos said: You can also use it to roughly predict low-speed hill-climbing capability. And this is exactly what made me doubt the results. I've off-roaded many thousands of kms with the 84V MSX as well as the V11. The difference in performance isn't large, but in my experience the V11 has been a better climber before overlean. And this is in the exact same places, and even with the same exact tire. I've even gained a bit of weight for the V11, which still  remains a winner. This is why the values of 174 for the MSX and only 141 for the V11 doesn't quite make sense to me. And the V11 compared to 150 for the V10F doesn't sound quite right either. Or the V12 HS vs Nikola/16X. To me it looks like the V11 and V12 might have a tighter stall current limit than the other mentioned wheels. In the tested firmware versions that is. (Edit: Also, S18 (162) has been called a bit weak by several reviewers, yet the V11 (141) hasn't been.) If I understand this comment correctly, @yoos agrees: On 6/1/2022 at 6:24 PM, yoos said: Though many EUCs seem to have a plateau at the beginning (due to an imposed limit on current)  On 7/12/2022 at 10:19 AM, dycus said: Acceleration numbers are I think more accurate now (and the image is updated). I agree they're not very useful as absolute values, but should be good for comparing wheels to each other. I don't think they are. The power output of the firmwares takes several aspects into consideration, but I don't think scoring well in a stall test would be one of them. Not burning up when not riding might be though. If someone would bolt down the rims of a car and measure the maximum force that the car can give out, I don't think anyone would extrapolate it into acceleration values either.  On 7/14/2022 at 4:28 AM, alcatraz said: I'd love to see that in the graphs. Like, how much force was applied to rotate the wheel to produce a certain (baseline) force forward. Doesn't sound doable though. Newer Inmotion wheels have 202 riding modes, which sometimes change in behaviour after a firmware update. Other manufacturers have only three, but they too change the behaviour every now and then. Edited August 10, 2022 by mrelwood 4 Quote Link to comment Share on other sites More sharing options...
Tawpie Posted August 10, 2022 Share Posted August 10, 2022 (edited) You raise valid points @mrelwood for certain... and that's the challenge with 'data'. What can you legitimately infer from it? Because in reality the only thing data provides is a small number of specific of observations under a given set of test conditions, it is an astonishingly narrow snapshot. And we all know what an odd cat or two can do to observations. Particularly if they're Austrian cats. Still, I do think we should continue to seek quantitative measures even if they don't translate into effects that can be measured by a user while the system is in use. Remember street rods? Brake horsepower was (and still is—sorta) a common measure of an automobile's "power" and was used to gauge acceleration, straight line speed, (lack of fuel economy)… mostly good for bragging rights but it gave you a possible clue that this motor/transmission/rear end/wheel/tire/track combination might be faster in the quarter mile than that combination. Or not. Introduce things like traction control, turbo, fuel injection, variable timing, custom ROMs, all wheel drive, launch mode, that huge spoiler and trombone muffler etc. etc. and brake horsepower is reduced to "necessary, but definitely insufficient" as a predictor of whether or not you should race for pinks. If I had a lot more money than I do, I'd put dual motor Tesla electrics into my '86 Trooper... give the better-than-you yuppie in the beemer a little moment. (for about a hundred yards) We're getting to the same spot with wheels... much to the dismay of marketing they're difficult or impossible to reduce to a simple number. My 16X has more Hz than my MTen, and it goes farther and faster and corners better at speed. Has a bigger wheel, more volts, amps, p and mosfets (actually, it might not have more mosfets, it's really an XS), and sports twice the pulling power at stall. But for some reason, the MTen from 0-5 mph feels way more capable than my 16X even though it doesn't even have a tube. In the 0-5 mph measure, my S22 is going to be a real dog. At any rate I do agree with you 100% @mrelwood, be cautious about the conclusions you draw from this information. Still fun though!  (BTW, I reserve the right to remain stupidly smug because my lowly slow old ugly 16X is more beastly in this particular measure than all of the "performance" wheels tested to date. I do fully expect Master will crush everything though, including itself if it tips over on its nose) Edited August 11, 2022 by Tawpie 2 Quote Link to comment Share on other sites More sharing options...
dycus Posted August 11, 2022 Share Posted August 11, 2022 3 hours ago, mrelwood said: If someone would bolt down the rims of a car and measure the maximum force that the car can give out, I don't think anyone would extrapolate it into acceleration values either. No, because cars don't have full torque at 0 wheel speed, whereas EUCs do. But I definitely agree that this test has questionable usefulness because it doesn't really model or test any real riding conditions. The only thing it tests is... how hard a wheel can pull a rope attached to its axle, at 0rpm. I will still stand by my statement that the acceleration values I provided, while not accurate in terms of absolute numbers, should be roughly representative of the maximum relative acceleration between wheels. This is only an instantaneous peak, once you go above 0mph then things will change. This test should show the maximum torque a wheel can put out. Once the motor starts turning, torque only goes down. 1 Quote Link to comment Share on other sites More sharing options...
Popular Post Tawpie Posted August 11, 2022 Popular Post Share Posted August 11, 2022 I would slice the pie finer and say "maximum thrust the wheel and its firmware will produce at 0 RPM". 'Cuz software rules the world. 1 3 Quote Link to comment Share on other sites More sharing options...
dycus Posted August 11, 2022 Share Posted August 11, 2022 1 hour ago, Tawpie said: I would slice the pie finer and say "maximum thrust the wheel and its firmware will produce at 0 RPM". 'Cuz software rules the world. Yes, sorry, that's what I meant when I said "wheel". The whole EUC, not just the motor. Quote Link to comment Share on other sites More sharing options...
RagingGrandpa Posted August 11, 2022 Author Share Posted August 11, 2022 (edited) On 8/10/2022 at 3:06 PM, supercurio said: I just measured V10F, 16X, V12 HS (two firmware versions) and S22 (two firmware versions)! And no smoke came out I hope??  Hey @supercurio we're waiting in suspense!  Edited August 25, 2022 by RagingGrandpa 2 Quote Link to comment Share on other sites More sharing options...
RagingGrandpa Posted September 16, 2022 Author Share Posted September 16, 2022 On 9/15/2022 at 12:17 PM, The Brahan Seer said: My impression from what I've learned, experienced and picked up from reviews in that non suspension wheels always exhibit more torque than their equivalent suspension counterparts Is this true? Well, so far, yes. And Master 134V should be equal, once it finally proves itself... On 5/10/2022 at 12:52 PM, RagingGrandpa said: Quote Link to comment Share on other sites More sharing options...
RagingGrandpa Posted September 20, 2022 Author Share Posted September 20, 2022 (edited) 3 hours ago, Bizra6ot said: what is the limit that is reachedpower of the magnets that reaches its limit then skip? More precisely: the magnetic field is not strong enough to overcome the force of the rider. The magnetic field has two parts of course: the permanent magnets on the rotor, and the electromagnet on the stator. Electrical current determines the field strength of this second part, so more current gives more force. When the limit is reached, we mean the field strength has reached its maximum, because the electrical current has reached its maximum. And as we saw, the maximum current is limited by the controller's firmware, which is different in each EUC model, even if models share the same motor.  3 hours ago, Bizra6ot said: why for eg on your pull test you can always reach the limit of the board In the pull force test, the tire (and therefore motor magnets) is fixed in position by a rope tied to an anchor. And then I apply force on the stator (EUC body and motor coils) with a long lever, which makes it easy to apply a turning force higher than the motor can create. The controller senses the motion of the EUC body, and applies motor current to resist it, up until its maximum current is reached. And then the EUC body (and pedals) dip, in the presence of this very high force. "Skip magnets" is a strange phrase I don't use... but yes, the motor begins to rotate, when I overpower it with the long levers. Hope this helps  Edited September 20, 2022 by RagingGrandpa 3 Quote Link to comment Share on other sites More sharing options...
Chriull Posted September 20, 2022 Share Posted September 20, 2022 27 minutes ago, RagingGrandpa said: "Skip magnets" is a strange word I don't use... but yes, the motor begins to rotate, when I overpower it with the long levers. Imho "skipping magnets" could be used if stall current would be reached and one forces the motor from one magnet to the next- but as you described firmware cuts off before. Out of curiosity - this firmware limiting the current is feeling still as a smooth counterforce or one can "feel it"? By hard breaking or accelerating at lower speeds one hears the typical rattling which i'd call "skipping magnets" - but no idea how much of this is created by the firmware limiting not beeing "smooth"... 1 Quote Link to comment Share on other sites More sharing options...
Bizra6ot Posted September 20, 2022 Share Posted September 20, 2022 31 minutes ago, Chriull said: By hard breaking or accelerating at lower speeds one hears the typical rattling which i'd call "skipping magnets" - but no idea how much of this is created by the firmware limiting not beeing "smooth"... Yep that's why I asked @RagingGrandpa cause from standstill I was able to reach multiple times 210 to 230A without motor cought(or rattling, what i called skipping magnets too)/causing pedal dip, then on an attempt it only reached 195 to 199A but this time it cough/pedal dip so i was wondering why it was not related Quote Link to comment Share on other sites More sharing options...
Popular Post RagingGrandpa Posted September 20, 2022 Author Popular Post Share Posted September 20, 2022 (edited) 5 hours ago, Chriull said: is feeling still as a smooth counterforce or one can "feel it"? Once I overpower the motor with the handles, the handles move and the pedals dip while maintaining a relatively constant pull feeling. It feels like pulling against a strong spring. The motor torque is high and relatively stable (despite being insufficient to keep the pedals level). I realize the torque must experience pulsations as poles move into and out of their highest-flux positions, but this happens so quickly during the 'pull' that I don't notice it. I think the video shows it reasonably well.  5 hours ago, Bizra6ot said: I was able to reach multiple times 210 to 230A without pedal dip then on an attempt it only reached 195 to 199A I'd attribute that to random undersampling: The bluetooth-based reporting of phase current only updates at 5Hz or slower, while the current measurement taken inside the controller is sampled much faster (100Hz+). Anyway, your observation sounds normal to me.  Edited September 20, 2022 by RagingGrandpa 5 Quote Link to comment Share on other sites More sharing options...
Freestyler Posted September 21, 2022 Share Posted September 21, 2022 I can prep you a firmware with a much lower limit (let's say 100a) if you want to safely observe the behavior with your rig @RagingGrandpa! 1 Quote Link to comment Share on other sites More sharing options...
Popular Post Freestyler Posted September 22, 2022 Popular Post Share Posted September 22, 2022 Some updates from the latest firmwares: ex20s c38 from 220 -> 250 master v2 from 220 -> 240 ex30 250a 4 Quote Link to comment Share on other sites More sharing options...
Popular Post Freestyler Posted October 1, 2022 Popular Post Share Posted October 1, 2022 (edited) Mten4: 180a Master V3 is back to 220a. Master Pro & Master X: 240a T4: 250a The Master v2 & T4 report back a "Current shunt" that seems to be read directly from an ADC (analog to digital) register. Someone should check the numbers to see if we finally got battery current instead of phase current. It's in the main protocol in the place of phase current (so it's shown by all apps already) and also in extended packets.  I've lowered my mcm5 to 50a for some tests, but I found the overtorque growl a bit scary (not sure if it is safe or not) so I didn't test more. I then increased it to 180a (from 150a), to match the mten4. All the pedal modes seem a little harder, but I also changed the calibration (that I had for many many months) to bring the pedals back a bit, so take it with a grain of salt. I can revert back after few days and see if I notice a difference in feel. Edited October 1, 2022 by Freestyler 6 Quote Link to comment Share on other sites More sharing options...
RagingGrandpa Posted October 3, 2022 Author Share Posted October 3, 2022 On 10/1/2022 at 5:54 PM, Freestyler said: check the numbers to see if we finally got battery current instead of phase current I think you could distinguish it by testing:Â + During max-speed freespin, battery current is nearly equal phase current. + During stall pulling, battery current is some 4x lower than phase current. I think one of the parameters must be phase current (because that's what is needed for FET protection limits). So you could check other parameters against the phase current, for this relationship. (2020 MSX example below- blue battery current is estimated by EUC World, but I think the relationship is realistic)Â 2 Quote Link to comment Share on other sites More sharing options...
Vanyayak Posted October 7, 2022 Share Posted October 7, 2022 On 10/2/2022 at 12:54 AM, Freestyler said: Mten4: 180a  I then increased it to 180a (from 150a), to match the mten4. Hey, so u telling, that upper phase current limit in mten4 bigger than in mcm5? 1 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.