Worth the wait, Hydra is here.

[bexamous]

Got 3 velcro straps to hold battery in place. Picture below was testing straps out.. work well. I've now gone on two rides with 2 batteries. One ride I took two dead batteries rather than waiting on charger. 2nd ride took a 2nd battery just-in-case, probably could have gotten home without it.. well certainly woudl have gotten home but maybe I'd have a unpowered bike for last couple miles. Few things:

- 3 velcro straps work well. 3 is nice for some redundancy. Also makes it easy to tighten the straps, two can hold battery so can always undo the 3rd to pull it tighter. Can then kinda seqentially though thrm them pulling htem tight.
- Mouse pad + strap setup works great.. there is zero movement possible in battery, no way to wear on paint with vibration or whatnot.
- I can't tell when I'm carrying 2nd battery.. bike starts off heavy and with 2nd battery its still heavy. I don't notice its heavier. I'm kinda surprised I can't notice an extra 10 pounds.
- Its nice having 2nd battery. With single battery being conservative can get 40-45 miles, but not being conservative more like 30-35 miles. So if I set off with single battery and not sure if I'm going to go 30 or 40 miles.. I kinda need to be conservative just in case. With two batteries was nice to be wasteful.
- I think setup looks 'fine'. It doesn't look like I have a wierd battery setup.

 

[bexamous]

That reminds me I'd like to try and waterproof the battery cover a bit better! Is your battery the 48V or 52V? I'm thinking I should eventually get a spare to keep around.

The bike is so open. Can see where cables enter frame behind motor, its kinda hidden by swingarm but there is just a big opening there. If I lived somewhere where it rained more than a couple weeks a year I'm sure I'd care more. But my plan is just avoid water, lol.

Also switched to 48v, got two 48v 17.5ah packs now.

 

[loamoaf]

I think that long, unbroken space in the triangle from having the horizontal shock is really nice for getting a second pack in there. If you were able to invert the shock then I don't see any clearance issues with a model that has an external bladder either.

Do you know what cells your pack is using? That can give you a clearer picture of what kind of energy capacity you can expect at specific continuous power draws. I set myself up with what I needed to build 2x 52V54A18Ah packs but they do protrude from the frame not quite 1cm, and both are identical made to fit in the downtube, so no triangle stuff for me. Looking at the cell datasheet I should be able to get around 730Wh from a pack at 1000W constant output between my high/low voltage cutoffs.

I don't think I could fit another 18Ah pack in the triangle without some serious effort into pack design and figuring a way to secure it to the frame without risking any vibrations being primarily handled by the cage bosses, but maybe a 12Ah pack would be more easy to make secure. I think on these longer trips where I'm asking more of the motor will involve me towing a trailer where I could just store the second frame battery. I've taken a 52V17.5Ah pack 45 miles with 40% charge at rest left over and that's honestly a fairly exhausting ride for me (2 hrs 45mins) - and this was on paved trails cruising with a 1500W hub setup. Of course hanging out and catching your breath & a snack always helps but I imagine 30 miles on trails has got to be equally exhausting if not more if your setup has the motor assist decently tempered.

I think I'm going to order a custom frame bag from Rockgeist the first of the month when orders open back up. They've got a lot of sweet options and I'm shooting for the cage boss attachment on the bottom and a lace-up along the entire length of the top tube. Would like to be able to get away with just having a pack small enough to hold a hydration bladder and keep my wallet, keys, tools, and snacks in the framebag. I'm always seeking cutting edge solutions for sweaty guy problems

You might be able to get some small, lightweight ratchet straps to avoid having to readjust the velcro straps on your setup. If the rubber side of the mousepad is face the frame then it's not sliding anywhere

 

[bexamous]

Oh and after

If you were able to invert the shock then I don't see any clearance issues with a model that has an external bladder either.

The bottom shock mount makes it a little difficlut, eg Mara shock mounted inverted doesn't work.. see picture below. The arrow points to little bit that hits. Its so close. Could probably modify the rear mount to make room.

Do you know what cells your pack is using?

Samsung 35E. Only opened one of the packs, assume the other is 35E too, its supposed to be.

I've taken a 52V17.5Ah pack 45 miles with 40% charge at rest left over and that's honestly a fairly exhausting ride for me (2 hrs 45mins) - and this was on paved trails cruising with a 1500W hub setup.

Yeah in general I just need a single battery. Look at past 20 rides I usually use 270-570wh/ride.

I'm always seeking cutting edge solutions for sweaty guy problems

I've been pretty happy with CamelBak Skyline LR 10, its like a cross between a backpack and a fanny pack. Backpack hydration setups are too hot, and the fannypack hydration setups are too uncomfortable, needs to be too tight to stay in place. Skyline LR is like a mix of the two, its a fanny pack but still held up with shoulders so don't need it tight around waist. I dunno I've now got a few hydration things I'm never going to use again after getting this Skyline LR.

 

[kwseattle]

I've been pretty happy with CamelBak Skyline LR 10, its like a cross between a backpack and a fanny pack. Backpack hydration setups are too hot, and the fannypack hydration setups are too uncomfortable, needs to be too tight to stay in place. Skyline LR is like a mix of the two, its a fanny pack but still held up with shoulders so don't need it tight around waist. I dunno I've now got a few hydration things I'm never going to use again after getting this Skyline LR.

Check out the Osprey Synchro pack (they make a 5 liter and 12 liter in a few different colors). I have the 12L. It has a trampoline-like suspension that holds the pack entirely away from your back, it's the best backpack I've ever had. Zero back sweat and it holds a great amount without being too heavy, plus it's specifically designed for MTB so it has a helmet carrier, tool pouches in the front pocket, and an integrated rain cover. Random pic from the internet below:

 

[loamoaf]

On my frame too looks like I'm unable to invert atleast the DVO shock - the rebound adjust and bladder probably won't even allow it to mount let alone travel. I know they (Dengfu) experimented with trunion mounts on the E06 at some point but it wasn't working out - I think this was before they moved to a single solid linkage between the shock and suspension arm and may work with the solid piece they now use. I've seen at least one person here with a Hydra that has the trunion version of the shock I have and the benefit of that type of mount is being able to squeeze a larger shock into a smaller space, but can't be inverted since the trunion mount is directly on the shock body.

35E's are pretty good especially below 5A each. I want to keep an eye on my rides and how much energy I use and see if I can get away with a 3p pack that's a little lighter and fits flush in the frame, but I think ultimately like some others have found if I want to strive towards a lighter bike then something with a smaller motor than the M620 may be worth the cost. I didn't catch it - do you have both batteries wired up through a splitter or just swap them out? If you can have both hooked up at once then you could get an even greater range since asking less current from each pack. Looking at the 35E bench tests round 10-15% at 1000W for 2 5p packs. Like you said most of your rides are short so really maybe not as worth the effort.

@kwseattle I've got an Osprey 20L for daytrips and love it - doesn't have the trampoline backing but it does have the mesh with a foam insert. I'll have to check those out - I'm assuming either or will fit a 2L bladder which is what I have. I was using warm water from the electric kettle when using a bottle bright tab in my bladder and didn't realize the kettle set itself from 104 to 212 and the few seconds the water was in there it caused some separation of the bladder and the rigid backing... I love the convenience of the kettle but it's getting 1 star for that one...

 

[kwseattle]

@kwseattle I've got an Osprey 20L for daytrips and love it - doesn't have the trampoline backing but it does have the mesh with a foam insert. I'll have to check those out - I'm assuming either or will fit a 2L bladder which is what I have. I was using warm water from the electric kettle when using a bottle bright tab in my bladder and didn't realize the kettle set itself from 104 to 212 and the few seconds the water was in there it caused some separation of the bladder and the rigid backing... I love the convenience of the kettle but it's getting 1 star for that one...

My bag came with a new reservoir so you won't have to worry about what size it fits! I think it's a 2L. I tend not to use it much though because it's a huge pain to clean them lol. Bottle Bright is great though.

 

[bexamous]

I didn't catch it - do you have both batteries wired up through a splitter or just swap them out? If you can have both hooked up at once then you could get an even greater range since asking less current from each pack.

Currently just swaping them. But I have parts to wire them up in parallel w/ideal diodes protecting both packs. Just not felt very motivated recently, sooner or later will get around to it though.

But yeah I want to see if I can see smaller voltage drop on climbs. Or maybe even see if I can measure a difference in effiicnecy. I track all my rides miles, ft-elevation, and wh charger puts back in pack. Theoretically seems like I should be able to measure the difference with enough data points.

 

[loamoaf]

@kwseattle the best way I found to clean the hydrapak is warm water and then toss a clean hand kitchen towel on the paper towel holder and slip the open bladder over that. The hardest part is remembering to do it after get home instead of slipping into a food coma after a big post-activity meal. I can't stand off tasting water so bottle bright definitely comes in handy.

@bexamous how much voltage sag do you typically see on your rides with the single battery? On my UPP 35E pack I think once the battery got down to half way it would sag quite a bit, but really only above 500W or so.

 

[bexamous]

how much voltage sag do you typically see on your rides with the single battery?

I don't really have good idea. I'm rarely using like a constant amount of power >500watts, its just so varied, and display updates so slowly. I've kinda learned not to look at display at those times, haha, only when motor has been off for a few seconds to get a better idea of pack's actual voltage. This is one benefit of using VESC controller over Innotrace.. vastly surperior data logging. I'd totally like data just for fun.

 

[mmcarm]

Despite all the harsh rhetoric going around got a tough ride in recently and was impressed with what the Hydra with Innotrace Archon controller can do. Almost 6k vertical feet and almost 20mi with a little over 30% battery left.

Below are the tuning params I have set for assist 1-3. Roughly 45% level 1, 45% level 2 and 10% level 3. Am thinking to readjust the assist parameters to 4 even steps to highest level being strong wheel lift on strong pedal. Turning 59 this year, I'm not a thoroughbred so motor is giving some efficient boost.

{"assist": 1, "torqueWatt": "40", "trottleWatt": "41", "sense": "4", "maxSpeedTorque": "100", "maxSpeedTrottle": "100", "motorTorque": "10" },
{ "assist": 2, "torqueWatt": "40", "trottleWatt": "51", "sense": "5", "maxSpeedTorque": "100", "maxSpeedTrottle": "100", "motorTorque": "10" },
{ "assist": 3, "torqueWatt": "41", "trottleWatt": "60", "sense": "6", "maxSpeedTorque": "100", "maxSpeedTrottle": "100", "motorTorque": "10" },

Wish Rico would make the programming more flexible including customer battery voltage config change and not locking to a specific PC but appreciate the work that went into the bike's handling, power and efficiency. Also really wish they could deliver similar high levels of ability with an 8-9lb motor like the M600 instead of the 15lb brick of an Ultra. Lifting the 60+lb Hydra in/out and over things is a serious PITA.

 

[bexamous]

Nice.. I had taken data from like 30 rides, just miles, ft-elevation, and watthours charger put back in when I got home. Best fit is: watthours ~= 6.828 * <miles> + 0.094489 * <ft-elevation>

That ride for me would be near 691wh+-5% probably. Theoretically that'd be 81.2% of if pack is really 850wh, a more realistic like 780wh.. 88.6%.

Elevation is what really kills my battery. I weight 220lbs which doesn't help, lol. And over time I've started to use more and more motor on climbs. Used to be more happy be in low level and use 250watts and be in lowest gear and just slowly go up climb.. but more recently I've tended to quickly go to 750-1000watts and I'll still pedal and keep my heart rate up but I'll be going at 10-12mph on like a 15degree cilmb. If I do 200 watts or something pedaling when limiting moror to 250 watts I'm doing like 45% of the work up a hill.. but when using 1000 watts get up hill so fast I've done ust 16% of the work.

Two things I guess I wish Innotrace was better about, with how bike rides...

- You can't really have a very dynamic setting. I forget what Bosch calls it, EMTB mode or something. Innotrace works fine if you like define a few levels and switch between them as conditions change. I wish it was a bit more flexible to be able to stay in a single level and not have to switch as much. Probably more than a few ways to accomplish this, not sure how EMTB mode works. But eg as example what if while rpm was constant or increasing power would be limited to 500watts or something, but if pedal pressure was high and rpm was decreasing.. eg someone is starting to climb a hill.. they're pedaling hard but bike is not accelerating and instead is slowing down.. if it could go to 1000watts to keep a minimium acceleation. That's type of code I'd be messing with. I think like Shimano or someone has an accelerometer and when bike is angled it knows its going up a hill and uses more power.. that kinda requires hardware but that is kinda their method of doing this. My workaround for this is basically just getting into habit of changing levels more frequently, approch climb.. gotta both change gears and maybe bump power level one or two maybe.

- I use speed limit a lot. Often on flat grond on level1 limit is like 15mph, and if I can pedal at 16-17mph then motor stays off. Or at level4 I have limit at like 22mph and if I cannot maintain 22mph motor might give a constant 100 watts or something to hlep me keep 22mph. BUT the annoying thing... what if speed limit is 15mph and I can just barely maintain 15mph on my own.. what happens is motor is constantly turning on and off. It seems like every second it might be switching. The issue I have is it sounds aweful. THe power is seemless, from my point of view I can just exactly maintain 15mph no matter what. But entire time I'm hearing motor going between off and on and its annoying as heck. My workaround for this is basically always going up or down a level (aka raising or lowering speed limit for me) to make sure my pedaling power doesn't exactly match up with current level's speed limit.

Its fine though. Like bike as-is is so great. I'm just nit picking a bit cause I like playing with stuff. Used to be more into quadcopters and stuff was so awesome data logging and being able to easily edit source to flight controllers.. could get stuff to fly exaclty how you wanted. I've not looked at VESC source but it should essentailly allow anything if you're willing to mess with src. If stock market goes back up (lol) I'd be tempted to get a Luna X2 just to mess around. I've been trying to justify the need for one. Maybe like put some slick tires on it and I dunno. Its kinda hard to turn it into something else, haha.

Oh for settings I basically raise up torque, motortorque and speed limit in each level. So kinda get a little more of everything going up a level and a little less of everything going down a level. I use generally non-sport for like paved paths and stuff.. speed limits are mostly a little higher. And then sport levels bit more power but then with lower speed limit. I dunno just kinda ended up with this after trial and error.

level torque  limit motortorque
1     15%/7a  21kph level1
2     20%/10a 23kph level3
3     31%/15a 26kph level5
4     31%/15a 30kph level7
5     31%/15a 34kph level10
S1    15%/7a  18kph level1
S2    25%/12a 20kph level2
S3    31%/15a 22kph level3
S4    40%/20a 24kph level5
S5    50%/25a 26kph level7

 

[loamoaf]

That noise from the 'throttling' seems like it'd be annoying - would wish for more of a 'softer' throttling that slowly pulls the motor back as you approach the speed limit, so that by the time you hit whatever speed, it's almost all you pedaling.

Should I place an order for a VESC 6 75V since I have a backup M620 to play with? It does have a gyro & accelerometer on it. 70x75x18mm won't fit in the motor compartment I don't think, though should fit inside the frame. I'm totally unfamiliar with the VESC tool but from some quick research it seems like you can create throttle curves - I guess the big mystery is figuring out torque sensor output range. If you could create a torque input to throttle output curve. Does the Luna VESC allow for end-user tuning?

 

[mmcarm]

Bex your experience mirrors mine, especially on weekday rides great to be able to do 10-12 scenic miles in an hour with some strong assist. I'm 168 (wannabe 160) so power consumption difference may be weight. BTW I'm in Carmel which I think is near you, would be good to plan a ride somewhere in the Bay Area.

I'm doing the same thing to keep speed when transitioning from level/3-7% slopes to 10-15% slopes to keep a fun pace requires dropping gear and raising assist and it would be nice if the assist was automated. Seems like an profile that gradually raises assist when seeing pedal pressure increase and speed decrease (after 1-2 secs speed decrease ramp up for 1-2 secs to attempt to maintain some % of average speed during pedaling last 15 mins and reset/ramp down after).

It's probably going to require a more open platform to really dial these things in. We need selectable and tunable behaviors.

For pie in the sky after seeing the amazing power to weight and power curve of the Tesla Plaid carbon wrapped 20k rpm motors what I'd really love is a 2-3lb motor with reduction gear that drives a cog on the left side of the rear wheel that can push 1200 watts for 15 min or so. The complex and heavy geared merge of motor drive and pedal crankshaft in the Ultra and other motors doesn't seem very reliable with greasing and bearing replacements common. Folks have experimented with astroflight.com motors but hobby work with loud gearing has been a problem. Would love something like that mounted on a Specialized Enduro geometry style frame for a 46lb ride. The LMX 56 is very close https://lmxbikes.com/index.php/en/lmx-56-2/ but they freewheel the drive cog on the wheel when going downhill and regen on that bike would be perfect, works so well on the Sur ron for preserving brake pads and keeping reserve braking ability on long descents.

Appreciate all the experience and best practices shared here.

 

[rtp]

That noise from the 'throttling' seems like it'd be annoying - would wish for more of a 'softer' throttling that slowly pulls the motor back as you approach the speed limit, so that by the time you hit whatever speed, it's almost all you pedaling.

Should I place an order for a VESC 6 75V since I have a backup M620 to play with? It does have a gyro & accelerometer on it. 70x75x18mm won't fit in the motor compartment I don't think, though should fit inside the frame. I'm totally unfamiliar with the VESC tool but from some quick research it seems like you can create throttle curves - I guess the big mystery is figuring out torque sensor output range. If you could create a torque input to throttle output curve. Does the Luna VESC allow for end-user tuning?

Yes, to a point. Technically you can still use 'full VESC' but I'm not sure all Luna changes have been merged back to mainline yet, some of which may be needed for e.g. for extended/new config params, etc. (I've used desktop VESC to re-flash firmware but then laid off it until all changes are merged etc.)

The mobile app lets you control 'basic' params, more or less overlapping with a handful of differences vs the Archon. User manual is available here: https://docs.google.com/document/d/1ds5D7xXNwUojAMevT7k0E4jYHtkbr01vpV6vjiHXamQ/view?format=pdf

 

[loamoaf]

@rtp good info. So it sounds like whatever Luna has done to get VESC working with at least the M600 motor is getting merged into the full VESC tool in a future firmware release, in which case input for torque and cadence sensors should be handled, which is good news. The Trampa VESC I'm looking at isn't slated for shipment until September, shortly after which it will be getting very rainy here and I'll have plenty of time inside to experiment, though not a lot of great weather outside to field test.

What's crazy is some the VESC models I see have 2x or 3x battery inputs and 2, 3, 4, or even 6 motor outputs (pretty sure that one is meant for drones). I kind of throw shade at dual motor setups as clumsy and unrefined (and unnecessary) but these probably give a much more refined 'eAWD' experience since the one controller is handling all the motors, and knows all the numbers they are dealing with. What a dream it would be to have somehing like a Utah Trikes quadricycle with 4 hubs being driven by the same controller. Call me crazy but you could probably pull off some interesting stuff like torque vectoring. I'm sure there's also people out there who've done a lot with the VESC Tool, I had only learned of it a year or so ago and there seems to be a huge amount of content out there about using it.

 

[bexamous]

Well rode up this climb, stop at top for like 5 seconds, and when I start to go again get this jerking for a second or before I stop. I look and rear through axle had backed its way out.. and when like I stopped and I guess no pressure was on it it was able to slide another few mm and actually fall out of the nut side. I guess this is like reason #1000 to check all bolts and stuff more regularly, heh.

But I fix it, all looks good. I go to start to ride and display says like communication error 30(?), something like that. And the voltage on display is cycling like 48v 24v 9v 48v 24v 9v.. something like that. I turn things off, turn on again and same thing. Tried turing off display and turning off battery.. turn them on and same thing. Then sorta just accepted stuff was broke and planned to ride home without motor.. set off and like 10 min later it turns back on and was fine all the way home.

The two things happened so close together seems like they gotta be related. Maybe some of the jerking feel was motor not being able to turn cause rear wheel was locked up and Innotrace did some failsafe thing? -- I dunno.

 

[WaxedWookie]

@rtp good info. So it sounds like whatever Luna has done to get VESC working with at least the M600 motor is getting merged into the full VESC tool in a future firmware release, in which case input for torque and cadence sensors should be handled, which is good news. The Trampa VESC I'm looking at isn't slated for shipment until September, shortly after which it will be getting very rainy here and I'll have plenty of time inside to experiment, though not a lot of great weather outside to field test.

What's crazy is some the VESC models I see have 2x or 3x battery inputs and 2, 3, 4, or even 6 motor outputs (pretty sure that one is meant for drones). I kind of throw shade at dual motor setups as clumsy and unrefined (and unnecessary) but these probably give a much more refined 'eAWD' experience since the one controller is handling all the motors, and knows all the numbers they are dealing with. What a dream it would be to have somehing like a Utah Trikes quadricycle with 4 hubs being driven by the same controller. Call me crazy but you could probably pull off some interesting stuff like torque vectoring. I'm sure there's also people out there who've done a lot with the VESC Tool, I had only learned of it a year or so ago and there seems to be a huge amount of content out there about using it.

I've tinkered with VESCs bit with electric skateboarding. They're super-customisable, and while I haven't seen it done (the multi-drive systems I've built were configured in master/slave mode to avoid accidental vectoring), I'm sure it would be pretty straightforward - I think the biggest challenge would actually be getting input on where you're trying to steer.

The transition to VESCs should be really good for ebikes - massive customisability, and open source to minimise supply constraints.

 

[rtp]

@rtp good info. So it sounds like whatever Luna has done to get VESC working with at least the M600 motor is getting merged into the full VESC tool in a future firmware release, in which case input for torque and cadence sensors should be handled, which is good news. The Trampa VESC I'm looking at isn't slated for shipment until September, shortly after which it will be getting very rainy here and I'll have plenty of time inside to experiment, though not a lot of great weather outside to field test.

What's crazy is some the VESC models I see have 2x or 3x battery inputs and 2, 3, 4, or even 6 motor outputs (pretty sure that one is meant for drones). I kind of throw shade at dual motor setups as clumsy and unrefined (and unnecessary) but these probably give a much more refined 'eAWD' experience since the one controller is handling all the motors, and knows all the numbers they are dealing with. What a dream it would be to have somehing like a Utah Trikes quadricycle with 4 hubs being driven by the same controller. Call me crazy but you could probably pull off some interesting stuff like torque vectoring. I'm sure there's also people out there who've done a lot with the VESC Tool, I had only learned of it a year or so ago and there seems to be a huge amount of content out there about using it.

The LudiV2 is a bespoke controller for Luna, not an off-the-shelf/pre-existing VESC controller. In reality, doesn't matter all that much, other than of course the layout of the board (i.e. the LudiV2 replaces the Bafang internal controller and fits inside the M600 motor case). The HW/sw developer of that controller, AFAIK, used a prior iteration of it for a MUCH higher output motor - I think there's a video buried somewhere on endless-sphere. One of the theoretical wins, outside of being open source, would be removing the 'exclusives' and single source of <better> controllers, but of course - that's assuming a reliable HW producer comes into existence. Grin seems to be pretty much out of stock on most of their controllers until 2023 (IIRC, maybe it's late 2022?), but at least having > 1 option out there can only be good for consumers, and potentially also push things forward collectively, e.g. if for example, 'Bosch eMTB mode' type clone becomes available on VESC, it might kick Innotrace into upping their game and making additional improvements, and vice versa.

I think you're right - it would probably be neat for an 'SUV'/hunting/'Jeep-like' bike to have single controller for multiple motors. I'm not so sure I want one as in to buy/build one, but would be cool to play around on for a bit without doubt.

 

[loamoaf]

@bexamous definitely a good reason to get a torque wrench if you don't have one and some blue permatex threadlocker gel. I know loctite says use less torque with threadlocker but screw that. Dab of the gel, twist those puppies in to whatever torque spec it says, let it cure. When you look at brake caliper or rotor bolts they have a dry gel applied to them, and they definitely don't wiggle their way out. What's a little worrying in your case is the hub shouldn't even be turning about the axle for it to loosen - the hub should be rotating on the bearings inside it and the part that the axle fits through should be static (as it's clamped between the dropouts).

It's very likely the error was some protection tripping either in the controller or the battery (which causes an error with the controller). In my own experience dealing with voltage errors or overcurrent protection I had to have the power of for a few minutes for it to 'clear' the error after powering on again. I would assume if it's showing a communication error then the battery had a protection trip and that cause the controller to be goofy for a second - error between the display and controller. Maybe not, I'm just guessing really.

If Innotrace wants to get ahead of VESC controllers becoming more common, having a bluetooth dongle and an app and battery communication would definitely give it an edge. Being able to set your profiles from the app wirelessly would be pretty amazing, and then seeing any error info details or having one place to change BMS settings or view battery status would pretty much put it on par with any Bosch or Brose or Shimano stuff in terms of convenience and usability. I can see why Bafang is moving their stuff to CAN and pairing it with their batteries - if they aim to have something similar and locking down their stuff to an entire system being a part of that.