Using the Wangdd22 1500W 30A DC Boost Converter on an ebike

wturber

1 MW
Joined
Aug 23, 2017
Messages
2,153
Location
Fountain Hills,AZ
Using the Wangdd22 1500W 30A DC Boost Converter on an ebike.

Background:
Why bother? Short answer - economy, versatility and fun to try.
Longer answer: This is my first e-bike and I'm trying to keep costs down. I bought a used donor mountain bike ($180) and a $225 rear wheel kit (48V. 1000W) on ebay (CNEBIKES.CO) . I live in a hilly suburb of Phoenix, AZ (Fountain Hills) and want to commute to my work about 15 miles away in nearby Scottsdale. The idea behind the e-bike is to "level" the hills that I'd encounter making a long commute practical. Given my expected commute distance, I figured 13 amps would be a minimum battery capacity.
I really liked the look of these 36v 10s 4.4ah battery packs currently on ebay. Five for $150!!. That's 22 amps and almost 800 watts, but the voltage is lower than my system's 48V. Maybe these would work with a boost converter? If not, I guess I could rewire them to 48V or 52V. Though I'd really rather not.

The Boost Converter:

I paid $40, but you can get it for nearly half that if you buy on ebay and are willing to wait a bit. I was impatient.

The 30A rating only applies to input voltages between 10V and 30V, so my real input current limit is 25A which works out to 1050 watts at 42V - not the nominal 1500 watts. Further, it has a 20A max output which works out to 1080 watts at 54v. So realistically, this 1500 watt boost converter will only deliver around 1000 watts in my application - which isn't going to be great for hard accelleration. But will it be enough to get up a 10% grade and help me climb out of Fountain Hills to Scottsdale on a commute?

Results:
So I finally got everything installed and the bike put together last night (though some of it is still temporary). So I took it for some shakedown cruising this morning. The short answer to my question is that yes, it will work for my intended purposes. The main caveat is that you may have to feather the throttle in demanding situations, and it really does help going up hills if you give a moderate amount of pedal assist.

The general booster behaviour is that it tries to always output whatever output voltage you set regardless of the input voltage. I set mine to 54V and the system's LCD reported 53.9V.

On accelleration, if you place too high of a demand via the throttle, the converter simply stops delivering significant amounts of current. It still powers the system LCD panel but the battery icon goes "empty" and the motor stops working. When the icon restores, you've got power to the wheel again.

The LCD doesn't show watts or amps, but it shows volts. Higher throttle demands cause the voltage to drop. It seems that when voltage drops below 50V or so, the risk of losing power to the wheel goes up. So monitoring the voltage is a good way to gauge the demands on the system and prevent any power cut-off.
So I rode the bike a total of 25 miles around Fountain Hills this morning. I recorded the last half of my testing ride using Sport Tracker Pro. Here's a link to that second ride. You can see that it does pretty much what I wanted. It "flattens" the hills a bit allowing me to maintain around 20mph going up many hills. I think the system will work better when I get a better 20mph gear. Right now, I'm over-spinning at 20mph in my tallest gear. The current gear feels comfortable at around 18mph. This jibes with my Sports Track Pro average moving speed of about 17mph. Spinning beyond 20mph is almost pointless. I need taller gearing.
http://www.sportstracklive.com/track/detail/wturber/Cycling/FH-to-Hidden-Hills/ebike-test/2346622

Power usage:
I don't have a separate watt meter, so I'm not sure if I have the boost converter configured to deliver its full power. But I think it is, or is at least close to it. Its an analog converter and adjustments are made with pots. You need measure with a meter while adjusting to know exactly what's happening. I've just spun the pot a lot in the clockwise direction. But I can still estimate total power usage based on the change in battery voltage.
I started with four 4.4ah packs wired in parallel showing 41.5V (about 94% capacity). I rode 25 miles with a net elevation gain of zero. My ending voltage was 35.9V (about 5% capacity). So I used about 89% of the 17.6 ah potentially available at an average voltage of 38.7V. That works out to 606 watts or about 24.25 watts/mile - not too far off from my benchmark expectation of 30 watts/mile for a commute.

Temperature:
It was around 90 degrees F when I was doing these rides and it was sunny. I measured the hub temp at the top of a long (1.5 miles 4.4% average grade sometimes 10% grade) climb up Palisades Blvd. It was 119.5 degrees F. Given that the hub is painted flat black, I think some of that temperature is just from being out in the sun. The booster itself never got hot enough so that it would turn on its cooling fan (somewhat disappointing - but maybe my placement for good airflow helps keep it cool). Spot measurements showed some booster components at around 115 degrees F, but most closer to the mid 100s. No wires or connections seemed hot or noticeably warm.

Conclusion
This isn't the optimal performance setup. It isn't for drag racing. But it works nicely for commuting. In fact, there is a nice unintended benefit from the booster cutting out under heavy load. It makes it hard for the rider to over-tax the batteries or cook the motor.

One small disadvantage to the booster is that system's LCD panel always thinks it has a full battery since it sees always sees 53.9V - regardless of the battery condition. You can set the booster to a low voltage supply limit. When reached, the booster stops boosting and simply passes direct, un-boosted battery current to the output. In that way the system can slap you upside the head when you've nearly fully depleted your battery source - at least keeping you from over-discharging your batteries. As it is, I think I need one or two (before and after the booster) watt meters so I can better monitor battery condition and figure out the efficiency of the booster. (They claim 92-97% efficient.)

FWIW, the motor and controller worked when directly connect to the 36V battery pack, my concern / question is whether the controller will cut out due to low voltage well before the batteries are near empty. That's something I'll test next. The 25 mph or so top speed that the 36V limits me to really isn't a big deal for my application.
 

Attachments

  • InstalledBooster.jpg
    InstalledBooster.jpg
    143.9 KB · Views: 13,243
  • SinglePack.jpg
    SinglePack.jpg
    95.4 KB · Views: 13,243
  • Whole_ebike.jpg
    Whole_ebike.jpg
    184 KB · Views: 13,243
Cool project! I also bought a few of those batteries! If you are ok with 20mph it will be cheaper and more effecient to buy a 36v controller. You can use any brushless controller with the motor you have, converting the voltage up creates heat which creates wasted enery that could be used to get you a few more miles instead.
 
skeetab5780 said:
Cool project! I also bought a few of those batteries! If you are ok with 20mph it will be cheaper and more effecient to buy a 36v controller. You can use any brushless controller with the motor you have, converting the voltage up creates heat which creates wasted enery that could be used to get you a few more miles instead.

Yes, that would certainly be more efficient. I considered doing that as well as possibly using a 72v controller (and wire the packs in series and parallel). But with this being my first ebike build, and with this particular kit coming with its own LCD display (that I wanted to use), I really didn't want to deal with a different controller and voltage from the outset. Having a controller with the wiring and voltage that is intended for the motor and LCD eliminates a lot of potential variables in just getting the bike going. With the booster, it was either going to supply enough current or it wasn't. The booster is a fairly simple and easy to understand/troubleshoot mod for a beginner like me.

I wasn't worried about a 5% or so loss in efficiency since I estimated that five of those packs would be plenty of power for my intended commute. Also, I came across a pretty good deal to buy 10 of those packs ($200). So losing 5% or so in efficiency wasn't a big concern. Today, I got 25 miles on just four packs in hilly terrain and the booster didn't get very warm at all. I couldn't even get the cooling fan to kick in.

The long-term and more elegant solution for being able to use a variety of power sources seems to be Grin Technologies Phaserunner controller. But for now, it just doesn't fit my budget oriented approach. Maybe if I get hooked I'll spring for the larger expenditure for the superior solution. For now, the booster seems like a fair compromise.
 
72 is a nice voltage for splitting back into 36v for charging so you don't need exotic chargers on hack together battery packs. Plus it wastes even less heat as less amps are flowing during use than running thru converters.
 
Voltron said:
72 is a nice voltage for splitting back into 36v for charging so you don't need exotic chargers on hack together battery packs. Plus it wastes even less heat as less amps are flowing during use than running thru converters.

Yeah - one nice thing about these packs is that each has its own BMS circuit board. Further, I can bypass the BMS circuit board and check a pack for balance and then balance it (if necessary) with a "hobby" charger. After about 10s, its gets pretty hard to find balance chargers and cables. So I'm really liking this the idea of having multiple 36v "modules."

72v crosses that line of voltages that are more likely to be deadly. Probably not a big deal if you take care to wire and maintain things appropriately, but it is another consideration. If I were going for high performance, I'd definitely consider it. That may be something to play with down the road.
 
Haha thats alot of chargers! Well if you want to keep your display you could always buy a 5s battery to put in series wih the 10s you have and your controller will like that voltage much better. Or just get a new controller and a cycle analyst which is expensive but awesome. Either way this hobby makes you waste your money all the time.
 
Voltron said:
Did somebody say multiple 36v modules? Lol :D

That's just the kind of thing I'd like to avoid. I just charge the packs in parallel. I just change the charge amperage based on the total number of packs on my pack "bus". So far, that seems to be working out OK.
 
Yes, this particular session was more about testing the cutoff voltage of my bin of chargers. Usually I'm in no hurry and use just two or three chargers at a time. I avoided the single bus thing as my pack is modular, so when switching to long range mode, I'm sometimes charging one pack up from storage voltage, and the others from some mid stage charge, so they can't all be paralleled yet because of the different starting voltages...
 
Just make a 13s or 14s 6p or more in parallel and buy a proper charger for the voltage. And ride your bike. Don't rip off the tabs...
What is the amp demand of the controller. 18amp 20amp maybe 22amp ? The important part. This will determine how many cells you need in parallel and you always need more just like when remodeling the house it's always going to cost more. More P
 
999zip999 said:
Just make a 13s or 14s 6p or more in parallel and buy a proper charger for the voltage. And ride your bike. Don't rip off the tabs...
What is the amp demand of the controller. 18amp 20amp maybe 22amp ? The important part. This will determine how many cells you need in parallel and you always need more just like when remodeling the house it's always going to cost more. More P

To what end? This system with booster appears to be working satisfactorily. The cost/benefit of re-configuring all those cells doesn't seem favorable to me - especially if I want to balance charge them and/or include a BMS of some type.

Not sure of the amp demand of the controller. Should be at least 20 amps given that this was sold as a 48v 1000w kit. I'll check to see if it is documented. I don't think it is. Horrible documentation on these kits.
 
wturber said:
Not sure of the amp demand of the controller. Should be at least 20 amps given that this was sold as a 48v 1000w kit. I'll check to see if it is documented. I don't think it is. Horrible documentation on these kits.

FWIW, the controller is 13 amps with a 26 amp maximum.
controller.jpg
 
Did more shakedown testing this weekend. The first was to check range. I set the LCD controller to a 20mph max - that's the legal speed limit for an ebike right now in Arizona. I only exceeded 20mph by pedaling faster or using downhill gravity assist. I did a round trip from Fountain Hills to the Scottsdale Airpark area where I work. I actually stopped and worked a bit at work and also stopped at one of my work partner's home to chat and show the bike. The idea was to test range and see if I could do a round trip on one charge (I brought two spare batter modules just in case). Here's a summary of that trip.

I used five of the 36v 4.4ah battery packs. I charged to 41.5 v. End voltage was 36.6.
Distance 33.25 miles
Average moving speed about 17.5 mph
Altitude change 745 feet
Steepest grade est. at around 18% (SportsTrackLive is reporting steeper grades like 24%, but I don't trust that report.)
Estimated wh/mi 22.25

This configuration handles long mild to medium inclines with no problem ( est. up to 7-8% grade). On steeper grades, you really want to help out some. It doesn't require a lot, but keeping the pedals moving really helps to keep the bike in motion at about 14mph up the steepest grades.

The second test was to see how much performance I could squeeze through the booster. My Tenergy Watt Meter/Power Analyzer arrived yesterday, so I installed it between the battery and booster. I then turned off the speed limiter on the LCD/controller. I did a shorter ride to explore commute options for exiting Fountain Hills and on this trip I tried to go nearly as fast I could, using power assisted on flats and moderate downhills and even moderate inclines to go well over 25mph. The biggest limitation here is that I don't have enough gear inches for pedaling. 42T x 14T is my tallest gear and while I can spin that up to 25mph or so, doing so isn't comfortable.

Watt hour/mile 38.4
Peak watts 1247.8 (drawn on the batteries, not necessarily delivered to the motor - should max at 1025)
Peak amps 34.74 (should max at 25)
Low Voltage dip - 34.84

I did some testing using the 36v battery directly connected to the controller and bypassing the booster. With the charge voltage starting at 41v, a heavy load caused a voltage dip that triggered the controller's low voltage cut out and the system shut down and did not recover. The Tenergy meter reported a dip to 40.28. I'll have to put the Tenergy on the other side of the booster to confirm this, but seeing the controller cut power leads me to think that with the booster in place, the controller cuts out when the booster dips below about 40.v. But with the booster, the voltage rebounds by about 14 volts, not just a fraction of a volt. So the controller sees the higher voltage and resets in 5-6 seconds. So its a kind of cascade. The booster tries to draw too much power and either shuts down or falls back in some way. When that happens, the controller sees the dip and cuts out. With no load, everything restores.

I understand that there is a resistor hack that you can make on the controller to change the low voltage cutoff. But it also appears that this motor and LCD/controller are the same items listed on BMSBattery.com. And according to the documentation there, the LCD/controller can be configured to change the cutoff and also run at 36V. I may try that out next weekend. If it can be reconfigured, then it may make more sense to eliminate the booster. Bur for now, the booster is doing a pretty good job of

So, the voltage booster will pull a little more than the 1025 watts. I'll test later on the other side of the booster and see what the booster delivers to the motor. Guess I shoulda bought two and put one on each side of the booster at the same time.

Anyway, after putting about 70 miles on the bike using this booster, while not perfect, it does seem to be a perfectly viable option for using a 36v battery on a 48v system. I could probably stop the system from every cutting out by physically limiting the how far the throttle can turn. That might be easier to do with a thumb throttle - which I think I'd prefer to this twist throttle. OTOH, if the LCD/controller will actually let me reconfigure to 36V with a lower cutoff, that may make more sense. I'd need to compare performance differences to find out. Maybe I can try that next weekend.
 
For reference, my LCD/controller is the S-LCD5 that can be purchased at BMSBattery.com. I guess that means that my controller is one of the S-series controllers or some custom variant of that. Anyway, the BMSBattery.com website has fairly complete documentation on the 72v S-LCD-3 which is largely the same as for the S-LCD5 and I was able to go through it and find some options that seemed relevant and potentially useful.

I changed two configuration items. I lowered the Low Voltage Cutoff by 2 volts. Would have like to go further. That makes the system much more usable if I connect the 36v battery directly. I will ultimately need to rely on the booster for LVC battery monitoring. I still need to configure that. I then imposed a max current draw limit of 80% of max. Doing that appears to have completely eliminated the power cutting off. I can run full throttle with no interruption. Peak watts drawn on the battery are still around 1250w and peak amperage is around 31.5a (still about 6a higher than the spec on the controller label.) But limiting the amp draw seems to have tamed the voltage dip and hence the controller cutting out.

So I'm really liking the results here. I can run the motor at 1000+ watts. The configuration prevents me from using much more and potentially putting reliability reducing stresses on the controller or battery. I don't have to rewire batteries and each pack has its own independent BMS. If a BMS fails (and doesn't short out the other packs in parallel), I just unplug that pack and go on. I can increase my range by running more 4.4 ah packs in parallel or by simply carrying extra packs and changing packs out as needed. I shouldn't have to charge at work if I don't want to. For those packs with the right connector, I can do a proper balance charge on an individual module basis (I need to find the right connector for the other BMS type so I can balance them as well) which should be helpful in heading off some battery issues. The packs are 10s 2p, so I'm able to balance as fine as two cells of I want to go to the trouble.

All in all, I'm pretty pleased. Time to shorten some wires, tidy things up, and find/build a proper battery case.

So right now, I see perhaps three drawbacks to using the booster. All of them small from where I sit.
1) Added complexity. It is one more item that might fail.
2) Valuable mounting space used (one water bottle mount right now).
3) Interferes with implementing regen since the regen can't charge back through the the booster. (Found out I can configure the LCD/controller for regen - assuming the motor controller supports it.)
 
It has been about three weeks and nearly 400 additional miles since my last post on this project bike. So here's an update.

I purchased a Tenergy power analyzer so I could better understand what the booster was doing. In short, the booster will draw up to about 1250 watts from the battery . That's over 30A at 41V, on the battery side. The booster will deliver up to about 1175W on the controller/motor side at the 54V value I have set the booster to. That's 21.75A. The controller is rated 13A with a 26A maximum. The 21.75A peak that the booster is delivering is about 80% of the controller's maximum. And, no surprise, I've had to set the controller to 80% max output in order to stop the system from cutting out at full throttle.

1175W and 21.75A has proven to be perfectly adequate for my commute that involves long hills with 6% grades and shorter hills with 10% or steeper grades. I'm averaging about 20-23mph when moving, reaching speeds on flats up to about 30mph. I have found a 50A boost converter that I may try out. Either that, or I might try using two of these in parallel (though people that know tons more about electronics than I do counsel that this may not work well). When I travel fast (closer to 25mph) I'm using about 15-16 wHr per mile average on my 16 mile commutes. When I'm staying around 20mph, I'm using more like 13-13.5 wHr per mile. I'd guess that my pedaling is adding about 5wHr per mile (assuming 100 watts/hr and traveling 16 miles each 50 minutes).

So I'm losing about 6% in efficiency through the booster at peak draw (1175w vs. 1250w). I'm also losing 20% of the maximum power that the controller could draw. But I'm gaining the ability to use the 10s 2p hoverboard batteries unaltered, connecting as many in parallel as I think is appropriate for a given trip.

I've made more configuration changes and have tidied things up a bit. I'll call this Phase 2. I still need to get to Phase 3.
I added some simple splash shields to the booster. I'll eventually have to do more to make it fairly water resistant. I see a faux water bottle case in its future. Water proofing isn't such a huge deal in the Phoenix area given how infrequently it rains, but it is still something I should address.

I've added head and taillights. The headlight is a circa 1990 NiceLite that was originally modified to use an 20W MR16 bulb. I've now changed that to a 50 watt equivalent (7.5 watt 12v actual) Soraa LED MR16 replacement. It sends out 470 lumens in a 10 degree spot beam. Taillights are bright LED strips on the seat stays and a circa 1990 LED taillight flasher (pretty novel and state-of-the-art back then.). The 12Vs are supplied by a 75 watt 12V DC converter. Since I had the 12V wired at the handlebars for the headlight, I went ahead and gutted some car cigarette lighter USB converters to make two USB power supplies. I'll use those to power my cell phone and a JVC action cam from time to time.

Aside from the lights, the most important upgrade was to get a larger chainring installed. The standard MTB chainring was too small for cruising at 20mph and greater - especially when combined with the 14T smallest cog on the new rear freewheel that came with the hub motor. I could have just purchased a 52T chainring for the mountain bike crankset, but that would have left a big jump going down to the middle chainring and a pretty useless third chainring left on the bike doing nothing. I had a brand new road crankset (Sakae Edge) in my parts box from around 1990 (note a pattern here?) and I had a 53T chainring I could put on it. Only problem was that this crankset uses a square taper spindle and the installed bottom bracket spline uses an ISIS spline. So I ordered a new square tapered crankset with a wide enough (127.5mm) spindle to maintain a decent chainline and ensure clearance of the chainring and crank arms with the chainstays. I had to move the derailer higher up the seat tube to accomodate the larger chainring. A new chainring and new freewheel calls for a new chain. Hmmm ... had one of those hanging around from around 1990 also. Cleaned it and soaked it in some paraffin and the gears are now much nicer for 20mph + . With the larger chainring, I find myself shifting again. I'm using two to three gears on long rides. The three smallest cogs on the freewheel.

I tidied up the wiring a bit and painted the controller and fabricated rack supports with flat black spray paint, added a frame pump, and Gorilla Glued a long close-fitting wooden dowel into the seat post as insurance against the rack causing a seat post failure. The main thing that I still have to do is work out a better system for holding the batteries. The old camera bag I'm using is fine for now, but I need to work out something that also allows/includes some cargo carrying ability. Carrying lots of stuff in a backpack is getting old.

Booster_splashGuards.jpg
View attachment 6
Cockpit.jpg
LED_Tail_SeatStay.jpg
Chainring.jpg
EntireBike.jpg
 
Looks like a wheelie machine!

should try getting the batteries lower and centered more

Nice progress tho!
 
skeetab5780 said:
Looks like a wheelie machine!

should try getting the batteries lower and centered more

Nice progress tho!

Yeah. Lots of people do make that point. But this is a commuter bike. I'm not taking it off-road on trails. I'm not hot-dogging anywhere. I'm trying to get from point A to point B comfortably and with reasonable e-bike speed. For this purpose, the handling is fine. Haven't popped a wheelie yet.

Frankly, my experience is that the main problem with the batteries located where they are is that they makes it harder to maneuver the bike by lifting the rear wheel and pivoting the bike around like I'm used to with a regular bike. The combined weight of the rear hub motor and batteries is substantial and just it isn't very easy to whip the bike around like I'm used to. Putting the batteries in the frame triangle would probably help that. I did, after all, pick this bike and frame in part for its ample front triangle space.

I agree with lowering the batteries but am of a mixed mind on moving them forward toward the center. Moving them forward would improve on handling in some ways and would reduce the amount of shock that the batteries must endure. But having them in the back (and especially lower in the back) probably improves braking by giving the rear wheel more traction to work with. It also keeps the bike looking more like a conventional bike. I do ride past legal speed limits, and I want to be as otherwise inconspicuous as I can be. Having stuff on the back is more typical of the commuter bikes I see in this area. I learned the lesson 30 years ago that a shiny white 1970 Trans Am making engine noise at 35 mph attracts far more attention from LEO than a brown oxidized hatchback Toyota doing 50 mph. I could speed with impunity with that ol' Toyota (given a long enough road :^) ). I want my e-bike to look like someone's "beater" commuter.

So, in all likelihood, I'll move the batteries into rear side panniers or something that looks like that and make sure that the batteries are well cushioned. Lower is good for handling and for weight transfer in braking. That also leaves room up top for regular cargo. That said, I might consider splitting the pack and putting some of the battery packs in the main triangle if I can keep the mass looking small in appearance. If I move the DC booster back in with the batteries, that will free up some space and let me keep a water bottle cage up front. I'll keep noodling on it.
 
OK - still haven't made a proper battery box for the LG packs.But I have made a number of minor tweaks and improvements.

1) Added a double leg kickstand. I love this thing.
doubleLegKickStand.jpgView attachment 4

2) I added a bell, a basket, and a second, STVZO approved, headlight. The headlight was only $15 and was a lot smaller than I had assumed. But it has its own battery and is surprisingly bright - though not as bright as my 50 watt equiv. LED Nice Lite hack. The beam does have the proper shape with a brighter top edge and abrupt top horizon. It is probably enough for 20 mph night cruising. But for the 25 mph speeds that I'll frequently hit, I'd want to mount two or three of these. One nice thing about the light is that it detaches easily. It would make a handy flashlight for evening roadside repairs and I like the idea of having a backup headlight.

My handlebar setup is now officially ridiculous. But I'm rather enjoying how silly it is. The camera is totally not necessary, but I have four of these, so what the heck. My bell has a flaming skull graphic. Nice and juvenile. Perfect. BTW, I generally have the phone display dark for safety reasons.
Handlebars.jpg

3) Power is supplied to phone, camera, and potentially to the the new light by a USB circuit scavenged from a car USB adapter and a car USB adapter in a 12v car "lighter" socket that is mounted in a 35mm film canister (for those of you who remember photographic film.)
12V_accessory.jpg


4)I installed the PAS sensor a few weeks ago, but it wasn't working. So I looked into it today and found that pushing the sensor onto the bottom brackets spindle cracked the center spindle. The crack allowed the rotor to expand and press against the housing like a drum brake. Hence no rotation. So I wired the crack closed and ground the gripping "teeth" down so there wouldn't be as much pressure on it. I also put a bit of epoxy on the inside of those gripping "teeth" to help ensure that they would stay stuck to the spindle and not slip. So if you are installing one of these, be careful about forcing it onto the spindle. You may want to grind down those gripping teeth a bit. The PAS seems to work fine at a "gear" setting of 3, but it still needs augmentation using the hand throttle. I wouldn't want to ride with this PAS only. Though it would be workable.
PAS_sensor.jpg

5) I've now had three flats. The first was a puncture from a staple. That prompted me to put some slime in the back tube. The Slime saved me on the second puncture which was a pretty deep cut that it sealed up quite effectively. But the Slime also hurt me in that it fouled the patch I later applied to that cut and was the cause of the third flat. Recommendation: If you path a Slimed tube, don't reinstall the tube right away. Let that patch cure with no pressure in the tube for a day or so.

The flats were annoying, so I upgraded the rear tire to the Panaracer Ribmo that Chalo recommends. Would have bought two, but they only had one in the 2.0" size. My front tire was a knobby, so I decided to replace it with the smoother tire that came with my kit. When I pulled the tire off the front rim I found that it had a protective anti-puncture strip around the tire. Hmmm.... so that's probably why I've had no front flats in nearly 1000 miles. I re-used that strip and now have road tires front and back. The bike seems to roll better and the Panaracer does seem to ride nicer - though I suppose that could be all in my head. If it flats, I'll probably put one of those liners on it.

I heard grinding on my back brakes and discovered an additional reason for filing dropouts a bit deeper when installing a rear hub motor. It is clear that the shoes are not fully engaging the rotor. That wouldn't be such a bad thing, but I think that may have led to an early disintegration of one pad and the scraping that I heard. OTOH, it may have been caused by the cotter pin that holds the pads in place not being properly installed. Time will tell.
discPads.jpg

The batteries and voltage converter have been working perfectly well. No hiccups. No problems. Plenty of juice to commute one 32 mile round trip plus errands at lunch.
 
I forgot to include this in the update. I enabled electric braking (regen) a few weeks back and found that it does not work and play well with the booster. It seems to cause a system shutdown. When braking for more than a few seconds, the LCD display will shut off. My guess is that with electric braking enabled, the motor sends enough voltage and current back toward the converter to cause it's protective circuits to shut it down.

So without doing some kind of modification, the converter interferes with using electrical motor braking.
 
wturber said:
So without doing some kind of modification, the converter interferes with using electrical motor braking.

I wonder if you couldn't use a pair of zener diodes and bypass link to feed the regen back to the batteries bypassing the converter; but I don't have enough electronics nounce to suggest a circuit.
 
Buk___ said:
wturber said:
So without doing some kind of modification, the converter interferes with using electrical motor braking.

I wonder if you couldn't use a pair of zener diodes and bypass link to feed the regen back to the batteries bypassing the converter; but I don't have enough electronics nounce to suggest a circuit.

I'd consider trying it if anybody can give specific circuit details.
 
It's complicated because of the different voltages and how things behave due to that.

Since the converter is a boost converter, you can't just use a diode to bypass it, because they'd let any higher voltage go back to the battery, meaning the output of the unit itself would also go back, and short the converter out (the voltage will drop to that of the battery, depending on internal battery resistance).

Also, even if you can send the regen around the converter somehow (set of relays to disconnect the converter from circuit, and directly connect the battery-to-controller), you have the issue that the regen voltage is much higher than the battery voltage. Or rather, that the battery voltage is lower....meaning current flow might be high, and unless battery resistance is high it will "short" the regen voltage down to the batteyr voltage, and unless the controller LVC is lower than that, it'll then disable the controller, which will also shut off regen.

Until the switch/relay/etc that bypasses the converter is shut off, the system is now powered down.


If it's an automatic system, such taht the converter is unbypasses as soon as the braking output stops, and is rebypasses as soon as the braking output restarts, then it's possible you'd get a chattering braking at the rate of the systems' startup/shutdown delays.

If the controller doens't re-start braking when powered on, but rather just sits in an error state because the brake lever was already pulled when it turned back on, then you'd have to release the lever before you can get the controller back on to either brake or throttle.

In all cases of non-normal operation, if you were depending on that braking, you're kinda screwed. ;)



Anyway, there are ways (like the relays or switches to bypass the converter during braking) that could work, depending on the behavior of the controller at LVC (and what LVC is for it), and if the battery has a BMS with HVC, battery resistance under the currents and voltages braking would generate, state of charge, etc.

But the only "easy" way to do it is to eliminate the converter and use a battery that will natively run that controller at the voltage the converter presently puts out.
 
Yep. It seems to be a fundamental limitation of using a boost converter.

While you are noodling this, how possible would it be to route the energy generated by the electric brake to a resistor or some other element where it is cast off as heat - like maybe a halogen brake light. :^)

You lose the regen benefit, of course, but that's generally kinda small anyway from all I've read. But you gain an additional brake and reduce brake pad wear.
 
amberwolf said:
Also, even if you can send the regen around the converter somehow (set of relays to disconnect the converter from circuit, and directly connect the battery-to-controller), you have the issue that the regen voltage is much higher than the battery voltage. Or rather, that the battery voltage is lower....meaning current flow might be high, and unless battery resistance is high it will "short" the regen voltage down to the battery voltage, and unless the controller LVC is lower than that, it'll then disable the controller, which will also shut off regen.

Until the switch/relay/etc that bypasses the converter is shut off, the system is now powered down.

Even though this isn't a problem I'm itching to solve, I re-read your post just so I could understand better. And upon re-reading it, I don't understand your comments above.

If I used relays somehow to send regen voltage to the batteries, that voltage shouldn't affect the controller because the controller is on the other side of the booster - right? The controller should still see the standard output voltage from the booster - especially if the batteries short the regen voltage down to something closer to battery voltage (around 40v - the booster output is around 54v). Or am I missing something?

Anyway, its an academic question for me. If I were to get serious about hub motor braking, I wouldn't care a lot about what happens to the regen voltage.

amberwolf said:
In all cases of non-normal operation, if you were depending on that braking, you're kinda screwed. ;)

I can't imagine also disabling the rear disc. This would only be additional/redundant braking.
 
wturber said:
If I used relays somehow to send regen voltage to the batteries, that voltage shouldn't affect the controller because the controller is on the other side of the booster - right? The controller should still see the standard output voltage from the booster - especially if the batteries short the regen voltage down to something closer to battery voltage (around 40v - the booster output is around 54v). Or am I missing something?
The regen voltage is created by the controller itself, and is output back thru the battery input wires. So the controller always "sees" the regen voltage.

In order to get the regen voltage to the battery, you have to disconnect the converter *completely* from the circuit, because the relay will be shorting across it from output to input, by connecting the battery input wires of the controller directly to the output wires of the battery.

If it does not also disconnect the converter, the converter could be damaged by that--either by the output voltage being higher than the input side can handle, or by the output side being dragged down and causing too high a current flow from the output (and back into the input). It depends on how it's designed.

But either way, it still will not be outputting it's normal higher voltage into the controller at that point.


Because that voltage is on the battery wires, then when you use a relay to directly connect the battery to the controller, then (depending on resistances between and inside the battery) the regen voltage is "shorted" down to the battery voltage (it will not go down all the way, because the battery voltage itself will rise; how much depends on it's state of charge and cell resistances/etc).


The problems arise under one of two conditions:

If teh controller's LVC is higher than the battery's voltage at that point, the controller will shutdown. If the controller's LVC is lower than the battery's voltage will ever be, there's no problem there.

If the battery's BMS HVC is lower than the battery's voltage at that point regen current is flowing into it, it will shutdown the output, disconnecting the controller from the battery. Probably it will reenable the output shortly after this, once the voltage drops below HVC. If the regen voltage is never higher than the battery's HVC voltage, it's not a problem.




These things aren't normally an issue with a controller and battery that are rated to work with each other. Like a 36v battery on 36v controller, or 48v b/c, etc.


I can't imagine also disabling the rear disc. This would only be additional/redundant braking.
Yeah, I just meant that if for whatever reason you were in a situation in which this braking was expected to work in one way, but something else happened instead, the delay in reaction time could be enough to cause a crash, impact, etc.

The braking issue isn't really the only one--the possiblity that the controller would shutdown, and then expecting to be able to power out of the situation failing (or being delayed), then the delay in reaction time before you could just crank on the pedals instead, could cause...whatever.

Anyway, not all that likely a circumstance--but I have found that being dependent on the power (and braking) of the motor system, when something doesn't work exactly like I'm used to, the reaction time delays for me have almost caused problems--if I weren't predictive and reactive and plan to not be too close to others I'd've definitely had a problem sometime.
 
Back
Top