Lots of load cell foot brakes on google for racing simulators. A load cell hand brake would be perfect for variable regen, so it's pressure not travel sensitive. I want this on my bike. Anything like this exist off the shelf before I make one?
Load cell brake lever
- Thread starter flat tire
- Start date
Didn't Grin Tech have something like that at one time?
I bought a 10kg strain gage and got my arduino outputting a constrained 8 bit pwm to my controller thru a 1kohm / 100nfd "diy dac" but it is on off and laggy. The strain gage itself works great and outputs very precise values. I will try again with a real DAC.
larsb
1 MW
Good idea!
You just need an analog pressure sensor and an adapter, something like this :
https://m.banggood.com/300PSI-Pressure-Transducer-Pressure-Sender-for-Oil-Fuel-Gas-Water-Air-p-1034362.html?gmcCountry=SE¤cy=SEK&cur_warehouse=CN&createTmp=1&utm_source=googleshopping&utm_medium=cpc_ods&utm_content=simon&utm_campaign=pla-mech-se-m-en&creative=295264625514&gclid=EAIaIQobChMIrqzYlKzx3QIVR4myCh2TLQAFEAQYASABEgLo_vD_BwE
http://www.mattlewisracing.co.uk/product.php/1496/29/gauge_adaptor___matt_lewis_racing
You just need an analog pressure sensor and an adapter, something like this :
https://m.banggood.com/300PSI-Pressure-Transducer-Pressure-Sender-for-Oil-Fuel-Gas-Water-Air-p-1034362.html?gmcCountry=SE¤cy=SEK&cur_warehouse=CN&createTmp=1&utm_source=googleshopping&utm_medium=cpc_ods&utm_content=simon&utm_campaign=pla-mech-se-m-en&creative=295264625514&gclid=EAIaIQobChMIrqzYlKzx3QIVR4myCh2TLQAFEAQYASABEgLo_vD_BwE
http://www.mattlewisracing.co.uk/product.php/1496/29/gauge_adaptor___matt_lewis_racing
larsb
1 MW
I figure you might want a signal directly proportional to the front braking force and this is what the pressure sensor gives. If you'd want it independent then you could mount a blind plug on the adapter end.
It is just a screw-on piece so that's what i'd do. How do you plan to mount the strain gage?
It is just a screw-on piece so that's what i'd do. How do you plan to mount the strain gage?
If it's just on-off, that's almost certainly a controller issue, where either it cant' respond in any other way, or isn't yet programmed to. Or it's variable-input range is much smaller than the range of the strain gauge / arduino / dac output.flat tire said:
You can verify this by using a potentiometer instead, wired with one end to 5v, the other to ground, and the center tap / wiper to the brake input.
The lagginess would also probably be the controller's own response time, unless the arduino program simply takes a huge amount of time to process the info. DAC response times (even resistor/capacitor versions) take so little time to process that you probably wouldn't notice the delay. But controllers often have long response times, sometimes as much as half a second or more, to any kind of input (even throttle or PAS, including shutting off once their input stops). Not very safe, but common.
You can verify that by using a simple switch to engage the brake, as it's response time is essentially instantaneous. (or use the potentiometer as noted above, as it would also be instantaneous).
amberwolf said:
It's a BAC-2000 and I have already programmed a Domino for variable regen and I can monitor the regen input voltage which works great for the pot throttle and not well for the arduino so that tells me there is a problem with what the arduino is outputting. There is no lag on the controller side. I can monitor the raw values on the arduino and what I see there makes it clear the gage itself (or its amplifier) is laggy. The arduino itself is extremely responsive and does not lag. Anyway that lag is kind of a huge problem even if I get a proper DAC. I'm trying to figure out if it's an intrinsic feature of strain gages or just the cheap one I got.
rowbiker
100 W
Even cheap chinese strain gauges have no lag when they are properly mounted and strain is applied. I have used strain gauges to supply a voltage signal in EV projects for several years with no "lag" issues. Note that in common parlance a "load cell" is simply one or more strain gauges already mounted to a host device.
I have, however, found that a *true* analog (not PWM) voltages in the 0-5 range is what 'most' controllers work best with. Note that this information is anecdotal, based on personal experience with about a dozen sensor/instrumentation amplifier/controller/motor combinations. I now get consistent results with a variety of strain gauges, a custom strain gauge amplifier feeding a true analog signal to the throttle input on a CAv3 and thence to a Phaserunner controller. Justin's (Grin) experience and development work on this was instrumental in my arriving at a successful outcome.
The result using the above combination of parts is a very responsive curve from standstill to speed to a near-stop with regenerative braking. I use a wide variety of non-geared BLDC motors, all of which work just fine for this. Having one or more firmware/software configuration layers in the mix can be very helpful for 'tuning' the throttle/regen curve for your particular hardware input devices.
For manual control, the still-missing piece here is a single handlebar-mount input device which supplies a 0-5VDC signal with a detent for 'neutral' (neither accelerating or decelerating) in the 0.85-1.1 volt range. The device could be designed as either a Hall or a pot device, as long as the spring-loaded 'at-rest' position kept it in the throttle dead zone, for safety reasons.
Do you have schematics and parts lists you'd be willing to share, to help people make their own?
Brake levers or other systems that do this aren't commercially available anywhere I've been able to find in recent times, so DIY is the only way people have to do this. If each person doesn't have to develop a system from scratch themselves, it would save them a lot of time and money and frustration.
Brake levers or other systems that do this aren't commercially available anywhere I've been able to find in recent times, so DIY is the only way people have to do this. If each person doesn't have to develop a system from scratch themselves, it would save them a lot of time and money and frustration.
rowbiker
100 W
amberwolf said:
Take a look at erowbike.com, which is the link in my signature. I've tried to document many of my efforts, including posting the eagle schematics for the strain gauge amplifier board. I'll happily entertain suggestions for improving that website via either edits or posting additional material, if available. I'm fully aware that I'm standing on the shoulders of giants here, such as Justin and teklektik, and yourself. Plus, as authors say, credit goes to all of you, but the blame for any errors remains with me!
Brake levers or other systems that do this aren't commercially available anywhere I've been able to find in recent times, so DIY is the only way people have to do this. If each person doesn't have to develop a system from scratch themselves, it would save them a lot of time and money and frustration.
While I have some ideas for the 'two-way' throttle input, I'm just really starting the prototyping process, so it's too early to make any meaningful suggestions. It would be nice, however, for us to come up with a unique name for such a device so people could search for it. I ran across the initialism "TPS" while searching for info here on ES, and discovered that "throttle position sensors" were a thing, mostly in the automotive world it seems. While what we're after isn't exactly that, such an apparently unique phrase makes searching much easier. Since you, amberwolf, are obviously an 'old timer' (based on your post count, not your age
), maybe you could give this some thought to this and come up with a device name we could use to corral our efforts.
rowbiker said:
I'll look around at the site and see if I can help. It's cold and raining too much today and probably tomorrow for me to do any of the things I'd planned to get done this "weekend", so I'm "stuck" on the computer for various reasons related to the weather (and how my body reacts to it).
My first suggestion would be to retitle the links on the homepage to be more descriptive of what one can *do* with the stuff there. I would not have realized, without already knowing it was going to be there, that the INA link would take me to somethign that could be used to control throttle, braking, etc.
(similarly for the e-Assist link)For a throttle that does this, "bidirectional" is the common terminology (like the Vectrix); normal rotation direction provides throttle output, reverse rotation (past the detented "stop" position) provides braking output.While I have some ideas for the 'two-way' throttle input, I'm just really starting the prototyping process, so it's too early to make any meaningful suggestions. It would be nice, however, for us to come up with a unique name for such a device so people could search for it. I ran across the initialism "TPS" while searching for info here on ES, and discovered that "throttle position sensors" were a thing, mostly in the automotive world it seems. While what we're after isn't exactly that, such an apparently unique phrase makes searching much easier. Since you, amberwolf, are obviously an 'old timer' (based on your post count, not your age
What would be really nice, and what I'm presently working on a primitive idea for using a cable-operated throttle and an ebrake lever with a switch controlling a relay, is a regular brake lever that will "take over" for the throttle during braking, to provide the 0.8v - 0.0v required to generate proportional braking response for controllers like the PHaserunner and Grinfineons, *and* would be adaptable to the other controllers that use a completely separate analog input to engage and control braking, using a full 0-5v input range (or whatever). This is so one does not have to use the Cycle Analyst to provide the signal, and thus you don't have to use the throttle to control braking level but instead can do it naturally with a brake lever. (either replacing or adding to the regular mechanical brakes controlled by that lever).
A load cell actually on the lever would be simpler than what I'm doing by far, as it would leave the lever capable of operating a standard mechanical brake, as well. (or I suppose could be used on a hydraulic lever type).
What I'm doing is taking these parts:
--basic mechanical lever, with built in ebrake (or brake light) switch
--cable operated throttle (COT) body with analog hall sensor
--potentiometer
--cable from lever to COT
--5v or 12v relay
--0.8v supply from a diode/resistor off the 5v or 12v supply
I have nothing drawn up yet, so this is all just out of my head where I"ve been working it out:
The lever is cabled to the COT, using an intervening pulley if necessary to give sufficient cable pull to rotate the COT pulley the entire range of motion.
The potentiometer replaces the hall/magnet in the COT, drilling into the shaft to bolt the cable pulley to it, and bonding or bolting the pot body to the COT shell.
THe pot is wired to the 0.8v and ground lines on it's ends, and the wiper goes to the NO contact of the relay.
The NC contact of the relay goes to the actual throttle signal output that would normally go straight to the controller.
The Common contact of the relay goes to the controller's throttle input.
The relay coil goes to ground, and one side of the ebrake switch.
The other side of the ebrake switch goes to whatever voltage is needed to drive the relay coil.
(if you're using brake lights, you can use a DPDT relay instead of SPST, and use the other side to drive the brake lights off the same lever; this is what I'll be doing).
Operation:
When the ebrake lever is pulled, the switch in it disengages the throttle from the controller and connects the controller to the COT output, which is 0.8v at this point, which would just barely (or not at all) engage the proportional braking of the controller.
As the lever is pulled harder, the COT is rotated more, which drives teh voltage closer to zero, increasing braking by the controller.
If the controller is tunable for that, then you don't need to be able to adjust the voltage range of the pot in the COT. If the controller isn't tunable, then you will want to use a pot instead of resistor on the 0.8v supply to the COT's pot, so you can adjust the range of voltage the COT outputs over it's travel from the lever, so that the controller gets the voltage range needed to give you zero or absolute minimum braking when lever is first engaged, all the way to max braking at the point you want it to happen in the lever's travel.
Since mine will eventually control both the proportional regen of the rear hubmotors *and* the mechanical disc brakes I'll have on the rear wheels by then, I'll want the COT output to reach zero before the lever travel reaches significant mechanical braking, so I'm already getting max electric braking before the mechanicals are necessary. If I need more braking I can pull harder and get that, too. That can be done electrically in the supply to the COT, or it can be done mechanically by using a dual-cable lever with independent adjustments, and adusting the COT cable pull to be faster than the mechanical cable pull.
Ideally I'd like to make a system that would work on most people's bikes, that doesnt' require this second cable at all, like a load-cell on the lever itself, that could be universal in nature. It'd need an adjustable or alterable "black box" that is very simple to build, or cheap to buy and easily found commercially, that does the switching of throttle signals and sends the adjsutable-range throttle (or braking) output signal.
The load cell would make it more complex, but more universal, than the COT/pot/relay setup. Probably wouldn't be any more expensive. The COT/pot/relay setup should cost around $30 in parts, max, from what I've found so far--less if like me you already have stuff laying around to scavenge and build it from.
rowbiker
100 W
Thanks for the suggestions, amberwolf.
I'd originally put the website together for my own use as a reference 'notebook', and then for several other team members working on these related projects. I agree with you about editing the titles and now have that on my todo list (great activity for those bad weather days).
After having done more reading on ES, it's clear that there's a lot of interest in variable regen. Many of us think of it as a wonderful way to provide smooth, reliable, and largely maintenance free braking. The recuperated energy is just the icing on the cake.
My own reason for wanting to control both acceleration and braking with a *single* device jumped into existence back in 2013 when I first drove a Tesla Model S (electric car). Aside from all the other positive impressions this gave me, the use of the simple "gas pedal" to make the car go and (nearly) stop was a revolutionary moment for me. This is the basis for my quest for a "bidirectional throttle", being a *single* input device that controls all aspects of the vehicle's speed. In the automotive world, this "single pedal driving" is actually a 'thing', and a feature for most who experience it in a well-engineered vehicle.
Like others here, I've experimented with different setups, including the momentary push-button switch to turn on electric braking (regen) and then using the throttle to vary the amount of regen/braking. While functional, it was not thoroughly intuitive for me (although I'm still using it on some our prototypes because it works). As some have noted, it makes it more complicated to put someone else on your ebike who wants to test your ride. For me, the single input device, be it a lever, knob, twist grip, whatever -- is the holy grail. Picking a controller (and optionally a CAv3) that supports this functionality is the price of admission. Let's continue to share our progress!
I'd originally put the website together for my own use as a reference 'notebook', and then for several other team members working on these related projects. I agree with you about editing the titles and now have that on my todo list (great activity for those bad weather days).
After having done more reading on ES, it's clear that there's a lot of interest in variable regen. Many of us think of it as a wonderful way to provide smooth, reliable, and largely maintenance free braking. The recuperated energy is just the icing on the cake.
My own reason for wanting to control both acceleration and braking with a *single* device jumped into existence back in 2013 when I first drove a Tesla Model S (electric car). Aside from all the other positive impressions this gave me, the use of the simple "gas pedal" to make the car go and (nearly) stop was a revolutionary moment for me. This is the basis for my quest for a "bidirectional throttle", being a *single* input device that controls all aspects of the vehicle's speed. In the automotive world, this "single pedal driving" is actually a 'thing', and a feature for most who experience it in a well-engineered vehicle.
Like others here, I've experimented with different setups, including the momentary push-button switch to turn on electric braking (regen) and then using the throttle to vary the amount of regen/braking. While functional, it was not thoroughly intuitive for me (although I'm still using it on some our prototypes because it works). As some have noted, it makes it more complicated to put someone else on your ebike who wants to test your ride. For me, the single input device, be it a lever, knob, twist grip, whatever -- is the holy grail. Picking a controller (and optionally a CAv3) that supports this functionality is the price of admission. Let's continue to share our progress!
ZeroEm
1 MW
I like that feature in my leaf, wished my regen was stronger but need to feather the foot pedal to coast. A bike with pas would have the same issues coasting.
Yes, this is why I wanted the variable / proportional regen for my rear hubmotors--also because I can get much more braking there than I can ever get in the front, because the front is so light vs the rear tha it'll skid the wheel long before I reach the potential of a good brake.rowbiker said:
The rear whcih doesn't have any mechanical brakes due to frame design (soon to change, hopefully) could almost certainly drag me to a stop really dang quick (instead of several few trike lengths minimum, depending on road conditions) if I could get enough braking force--but slamming that kind of force on instantly tends to break axles of common hubmotors, with this much mass to stop. :/
If I could get stronger braking I wouldn't even care about the "regen" itself; I'd be happy to have the controller actively use *more* power to get me btter braking if it would help. :lol: So far teh best I can get, with two full-on-instantly 35-45A controllers, is around a couple of trike lengthsj. Even that is enough to damage these cheap axles over time, hence the desire for controllable braking to reduce the isntantaneous stress on axles and dropouts, until someday I can build custom axles and stator supports that can handle anything I could throw at them.
And it would probably also work ok on a bike...but at present there's no system that can do it on ebikes; the closest there is is stuff that uses the brake lever to turn on braking, but then you still have to actually control it with the throttle, and that's too many things to do, to much reaction time, riding in traffic. Needs to be a single thing, and for most cyclists coming into ebikes, the easiest way is to put it in the brake lever completely. They already know what that does, it's ingrained in from ridng bikes for however many years.
Whatever is most instinctive, most ingrained, is what people revert to in a panic stop or high-stress situation. Even if they have learned something different since then, they'll still often not do that, and instead do whatever they first learned. :/
Anyone coming from a motorcycle background could probably do the throttle-only braking thing easily enough, becaues they already have a throttle control they're used to handling, but brake levers are also a thing there, too, so a brake-only one would also work (and still might be more instinctive).
But as long as it is a single action, it's still much better than the present dual-action solution.
Using the Cycle Analyst with Grin products, it's of course possible to make full-on braking a single action, but it's uncontrollable, just on or off, that way (you just set the initial brake level to 0v), so it's not the best way to do this; you might as well not have the abiltiy to do the variable braking at all since it can't be used this way. This is how I'm using mine at the moment, so I can still get full instant braking for the times a car suddenly changes lanes right in front of me, or some other thing that would result in me crashing into them or whatever, if it took more than an instant for me to react.
What I'd *really* like to have is a PAS that didn't just shut off throttle when backpedalled, but could use a sensor of some type to control braking force, like having an electric coaster brake. Presently there is no system I can find that does this. If I could have that *and* a variable braking brake lever, I'd be very happy. Since the IGH I have presently on the tirke in the frame doeas actually have a coaster bfrake on it, theoreticlly I could use a troque sensor BB "bakcwards" to do braking with. But I'd have to build electronics to convert the TS output to a 0.8-0.0v output, since the CA can't operate it this way. (Seem silly for it to only be able to use the TH-in input as a control voltage for it's ebrake output, but AFAICT that's how it works). And I'd hae to go singles-peed in the chainline itself, using only the gear range in the IGH, to be able to tension the chain in reverse this wya. Presently I have the option to manually shift to different rings on the cranks for better low-spped or high-speed operation, though I can't really do much rel pedalling anymore I'm trying to keep my legs excercised with it. If I had a PAS-brake control system that required it, I'd live without it, though.
Personaly, I don't want the throttle to engage braking if I back off of it, because my hands go numb randomly and I can't feel where throttle position is, and I start decreseng force on the throttle and that would cause me to brake, which could be very bad in traffic. This is one reaosn I'm trying to work out a really good PAS solution (nothing works right so far). But I would be ok with a zero-detent position of the throtltle, above which it makes me go, and below which it makes me stop.
Like Vectrix's throttle mechanism. In a car, I'd say a single action thorttle pedal is a good idea, because people afe are hopefully already used to backing off the pedal before they brake anyway, so if it does the braking at the same time they back off, then that even saves reaction time.
Bvut if those same peopel got on a bicycle / etcx with a hand throttle, they'd still have to relearn this action for muscle memory. So they could learn any single-action braking solution and it would stick.
Exactly, and it also takes more reaction time to use.
I could devinitely use such a device, as long as it operated as I described above, with the zero detent above/below mechanism.
So if you work one out, I could test it, too. I'll still be working on the brake lever / COT mechanism in the meantime, as th4e simplest thing I can build that might be reliable for me, and I'll keep this thread updated as I work stuff out.
Not if it required backpedaling like a coaster brake. (see above post).ZeroEm said:
Though at the moment I know of no bikes or systems that can use the PAS for braking anyway.
rowbiker
100 W
amberwolf said:
IIRC the "Copenhagen Wheel" (rear DD hub motor with integrated controller, battery, and torque sensors) does proportional braking/regen based on the rider's backpedaling. I've only done several test rides and haven't read much about the actual technology used, but I remember thinking that it worked remarkably like the old-fashioned coaster brakes. The difference was that you couldn't just stand on the pedals going backwards and possibly even lock up the rear wheel, but you just back-pedaled the cranks freely and had a kind of reverse PAS going on.
Grantmac
10 kW
I've mentioned it before but I think with hydraulic brakes that an analog pressure sensor could probably work nicely. It's been done a few times previously and from what I remember 0-500psi was about right if you wanted some hydraulic modulation on top of the regen.
i did finally setup a version of the cable-operated throttle, and made a how-to thread here
https://endless-sphere.com/forums/viewtopic.php?f=2&t=105460
https://endless-sphere.com/forums/viewtopic.php?f=2&t=105460
Similar threads
