Throttle interface for sensitive high power setups

[Stealth_Rider]

Jeremy,

I Installed the throttle interface on my Stealth Bomber with detail pictures shown here:
http://www.fareinc.com/65/Throttle_Interface/index.html

The install was plug and play with JST-SM connectors on the interface.

The Bomber battery, motor, controller and throttle are all standard as delivered. The throttle interface was set up with a JST-SM female connector so the existing throttle could plug into it and the other end was sent up with a JST-SM male connector so that it could plug into the controller. The Bomber comes with a frame mounted small screen CA but I've removed that and installed a large screen CA on the handlebars. As a result I was able to route the wires into the frame and leave the throttle interface, mounted in an ABS enclosure, on the top frame surface so I could make changes.

Stealth Bomber specs are here: http://www.stealthelectricbikesusa.com/bomber-spec-sheet.html

Motor: Crystalyate 5403, 4,500 watts
Controller: Crystalyte http://www.fareinc.com/65/Throttle_Interface/content/IMG_2784_large.html
Throttle: PTHE hall sensor
Battery: 72V, 18AH, 1.5 kWh
CA: Large screen V2.3
No Load speed=60.2 mph
Top speed with no pedaling about 48 mph

The install directions were very well written and with the bike rear wheel off the ground it was easy to set the MINV and MAXV per the detailed instructions provided by Jeremy,
http://www.fareinc.com/65/Documents/Throttle_interface.pdf

I am still experimenting with the delay and buffer setting but I now have them set so I can slowly turn the throttle with the bike stopped and then walk next to the bike as it starts to move very slowly. Before the throttle interface if you turned the throttle slowly with the bike stopped it was almost impossible to get the bike to move slowly and lifting the front wheel was much too easy. Now the bike is much better behaved and power can be applied in a more controlled manner.

Where do I measure the voltage for the Delay and Buffer settings so I can report them. I see you reported Delay=0.65V and Buffer=1.43V in your written instructions.
Here's a picture of the board if you can tell me where to measure them. Bigger picture of the board is here: http://www.fareinc.com/65/Pics/Throttle_Int.jpg

-Jim

 

[zombiess]

Voltage can be read from bottom of the board on the middle pin of the pot. Ground reference should be measured from the board at a throttle in or out pad.

 

[snellemin]

I'm going to wire it up now. All add one of those lipo cell readers later on to read all 4 voltages at the same time. I know the color wires are not in order at the input. So no yelling at me that I have it wired wrong. :lol:
The new unit is way bigger then my simple throttle delay I had made for me at work.

 

[zombiess]

I think some sort of adjustable exponential response might be cool as well.

The throttle response is exponential and it's varies based on how quickly you rotate the throttle. Best I can do for right now. Still trying to find that pesky delay.

I changed the way I rode over a short period of time of using the interface and it became more natural to me. Could just be me though. This is why it's beta testing. I have other plans for something much nicer related to this I'm going to be working on.

 

[snellemin]

Just went riding in out of the retention pond. I love this thing so far. Just running the stock setting with the GNG motor and Lyen 6 fet sensorless controller. I didn't use the three speed switch at all. I just set it to the highest setting and played with the throttle. Rode partial throttle and all the bumps didn't change my speed like it normally would of have. Looks like I will up the phase current and see how it goes. I got use to it pretty quickly.

Sounds like it is a different can of worms if I install the interface to my 24s magic pie setup.

 

[zombiess]

Snellenim, how does it compare to your previous delay setup?

Good feed back so far. I like hearing the good and bad. It's also nice to see what people find difficult to adjust or setup. Very glad all the users have very different setups and power levels. Thank you to everyone posting so far.

 

[teklektik]

snellemin-
So - you are running this into your CA and using CA Current Throttle?

 

[zombiess]

So - you are running this into your CA and using CA Current Throttle?

That is something I plan on trying myself if I get some time to install my ca v3 sitting in a box. I have 500 irfb4115 FETs to miller plateau match later today for controller builds.

 

[snellemin]

Naah, I'm not using the current throttle feature. Would be nice though. I'll get to that once I have a lot of time on my hands.

Zombies, your throttle feature is a step above the one I had previously. Mine was just a simple adjustable time delay for the Yescomusa controller. It dropped straight to "zero" when you release the throttle. It worked good, but not with the Lyen controllers. Time got in the way to add features to it. But you did the work and so far I like it with the GNG setup.

 

[Stealth_Rider]

Jeremy,

When you find the elusive delay bug and change the programming to eliminate it, can we pay you for another chip, have you send us the reprogrammed one, and swap out old for new? That way we can keep using the interface without interruption?

I am working on a tuning strategy for my riding style since both the Delay(i.e. throttle acceleration setting per the manual) and Buffer are variables.

My most recent changes were to leave the Delay setting as is. Call this D1 and with D1 set I could stand next to the stationary bike and "very" slowly twist the throttle and be able to walk next to the bike with the bike under complete control and moving forward slowly. Prior to installing the interface the first slow twist of the throttle, for just a degree or so, was acting like a on-off switch and the bike would just leap forward from a standing start. The interface is sooooo much better than before and this factor alone makes the throttle interface a huge improvement for me.

Leaving D1 set I then went on to experiment with the Buffer. To do this I took the bike out for a ride, screwdriver in hand.

I first increased the Buffer 3 turns CW to increase the throttle position averaging and evaluated (with the bike moving between 10-30 mph) if the throttle lag was excessive or if I like the desensitization of throttle rotation to bike response with the same rate of turn of the throttle. When I went too far CW I started backing off on the Buffer Pot until the throttle response. while cruising and twisting the throttle both slow and fast, was to my liking.

My next ride I'll leave the Buffer as is and tweak the Delay setting.

Now, to help my tuning strategy, you said in your previous post,

The throttle response is exponential and it's varies based on how quickly you rotate the throttle. Best I can do for right now. Still trying to find that pesky delay.

So, does this mean that the Delay setting would be most affected by the rate of throttle rotation and, if I like the Buffer setting, fine tuning the Delay should have minimal effect on the Buffer setting?

-Jim

 

[zombiess]

Stealth_rider, the delay and buffer settings do have some effect each other. If you feel there is to much delay in throttle response you can do two things to improve it, twist the throttle faster or turn the delay pot clockwise a little more, both have the same effect but it needs to be tuned to your riding style. When you get close start making delay pot adjustments in small amounts like quarter or half turns. So yes, if you like the buffer feel, start tweaking the delay pot.

The buffer has the most noticeable effect at really slow throttle movements, a high buffer setting causes a slow ramp up unless you quickly turn the throttle, only experimentation will help you get it feeling right. I'm finding it hard to find the word on how they interact with each other. I need to come up with a good example or analogy.

Has anyone else noticed the slight delay in throttle response when letting off the throttle?

When I find the bug I can offer a swap out. The pic processor and shipping is about $5 so not expensive. Shipping is $2 because the envelope is over 0.25" thick.

 

[Stealth_Rider]

It's difficult to find a road where I won't run out of space to test this:

Has anyone else noticed the slight delay in throttle response when letting off the throttle?

So I tried a no load test.

I put the Bomber on my test stand which elevates the rear wheel. I then twisted the throttle and went to my no load speed of about 60 mph. The CA was reading about 1000 watts and when I released the throttle quickly (just let it go) the CA watts reading immediately went to 200 and then to zero very shortly thereafter. Is this indicative of the delay?

Jeremy, do you have any other test I could run with the rear wheel elevated? How should I look for this condition?

-Jim

 

[ZOMGVTEK]

The delay is BY FAR the most noticeable on my bike when accelerating at full throttle then slamming off while the bike is still pulling. It's been extremely noticeable 100% of the time I do this.

Certainly would be nice to have another opinion on this delay.

 

[zombiess]

The delay is BY FAR the most noticeable on my bike when accelerating at full throttle then slamming off while the bike is still pulling. It's been extremely noticeable 100% of the time I do this.

Certainly would be nice to have another opinion on this delay.

I can feel it on my bike too as a really short delay in cutting throttle when letting off wot. Other than this bug I'm waiting on feed back on how people like the response after they get it adjusted.

Stealth, Not much useful testing can be done without riding. I encourage people to try changing their riding style to one more similar of a gas engine (its not going to be identical though). When I made this change it felt more natural and responded faster to my input. Before I use to always let off the throttle fully when cornering and slowing but now I just reduce the throttle.

Shame there isn't a 180 degree turn throttle, a simple filter on that would be a huge improvement vs thee that only turn about 80 degrees.

 

[Stealth_Rider]

Hmmmm. I'll try way you describe. Do I have to be above a certain speed or as long as it's accelerating will be OK? I'm thinking I'd slam off about 30 mph.

-Jim

The delay is BY FAR the most noticeable on my bike when accelerating at full throttle then slamming off while the bike is still pulling. It's been extremely noticeable 100% of the time I do this.

Certainly would be nice to have another opinion on this delay.

 

[ZOMGVTEK]

It's most noticeable to me I slam the throttle on from a stop, get up to 20, and snap it off. I continue to accelerate at full throttle (or nearly) for what feels like about .2 seconds. It's very repeatable and I consider it to be very noticeable. As long as you can easily feel the bike accelerating for another 1/2 second, snap off and see if you notice it or consider it to be a big deal. It still does it at top speed, but it's hard to notice since the bike isnt accelerating anymore.

I notice it about every time I snap off the throttle. Just to make sure I'm not crazy, I pulled the board and its without a doubt gone. The bike is way more predictable now and I never really had wheelie issues or low speed control issues before, unless I was doing something stupid. The throttle sucked, but did indeed work fine and I'm used to it. Leaning forward and getting the rear tire up to 60 MPH while moving at 5-10 MPH can be fairly dangerous with this delay, unless you roll off the throttle and let the tire spin down before you sit back down.

 

[snellemin]

Just came back from the garage to test the .2 second delay issue. I was just freewheeling the bike and mess around with the throttle. I did full and partial throttle scenario and didn't notice it at all. Is it because I'm using the GNG and it goes through several reduction stages vs other peeps with DD setups?

 

[zombiess]

Just came back from the garage to test the .2 second delay issue. I was just freewheeling the bike and mess around with the throttle. I did full and partial throttle scenario and didn't notice it at all. Is it because I'm using the GNG and it goes through several reduction stages vs other peeps with DD setups?

I don't know, I can feel it on my bike, but it's a tiny delay for me that I barely notice and I have to be WOT and accelerating to even feel it. How about the rest of the riding characteristics? Are they any better than stock? Have you spent much time dialing in the interface? It's nice to see that there is a wide variety of equipment being tested.

ZOMGVTEK, what happens if you just roll off the throttle and or ride part throttle instead of snapping off the throttle? I know it's a bug and I'll find a fix for it, but I'm interested in the other characteristics, mainly part throttle. Just so I understand, your bike can do a rolling burnout at 5-10mph but can't wheelie? My bike can't do a rolling burn out, but can power wheelie at 25mph with me leaned over the handle bars when I run around 12kW and I have a long swing arm. What controller are you running? I wonder if there could be some sort of interaction happening between the interface and the controller such as a capacitor staying charged. You could try snipping or unsoldering the 0.1uF cap that is parallel between the output and ground to see if it makes any difference in the delay.

 

[ZOMGVTEK]

If I roll off the throttle it's fine, but I'm not a big fan of the fact that the throttle 'sticks' on for a bit should I need to snap off. However I have the board configured, even with all filtering 'off', it feels like the throttle input gets lost for a brief period, and it sticks on when I snap off.

The bike won't lift the front in any reasonable riding position at the current power levels, it's on a few year old Lyen 12 FET. I do have a few spare 6 FET's laying around I could try, but anything bigger is a little charred. To get the rear to spin I need to nearly lock the front brake and lean forwards a lot. Then it will easily hit nearly freewheel speeds. I run really small overinflated road tires to make this possible. Traction is rarely a problem, and I would prefer if it was. The bike is something like 100lbs, possibly more, and its rear biased.

I will remove that cap and see if it changes. It might take a few days.

 

[snellemin]

If I roll off the throttle it's fine, but I'm not a big fan of the fact that the throttle 'sticks' on for a bit should I need to snap off.

How about if you use your brake disable function? That will cut off the controller instantly. I'm assuming you like to coast and don't use regen.

 

[ZOMGVTEK]

I do a lot of stupid things with the bike, all the time. The throttle sticking on, even for a relatively brief period, is a very bad thing. Sure, I could grab the switched brake lever, but that would turn regen on, which given my very high regen power, can be just as bad as unintended acceleration in some situations. The front won't lift under acceleration on level ground. Up steep inclines, or even if you hold the throttle WOT over something like tree roots and the wheel bounces off the root, accelerates up to 50, and then slams into the ground and gets great traction, the front end lifts. This is normally not much of an issue, I can snap off the throttle and the bike is usually only doing a little wheelie. If the throttle was still WOT for another .2 seconds, at 18kW, the bike is coming out from under you. The bike only has a little 12 FET on it now, but its still kinda a big deal. Same deal with rolling burnouts. The wheel spins up and gets all sticky, then if I sit back down as I let off the throttle like normal, the throttle sticks long enough to have the front end snap up, usually this happens at moderate speed.

Plus, I would like to be able to give someone the bike and not have them kill themselves from the throttle sticking. It's way more of an issue than I assumed it would be. It's really a big deal. I would imagine you would agree if you rode my bike like I do.

I'm assuming there has to be something unexpected going on, since I don't think zombiess would ship the board knowing the issue I am experiencing exists, and nobody else appears to think its a big deal.

 

[snellemin]

Ah I see. I'm using a bottom bracket drive and my phase current is pretty low, hence I don't see the issue. Looks like I'll see the issue if I swap the unit to my MagicPie bike.

 

[zombiess]

Zomgvtek, is there a possibility that you have the max throttle v set too high. That could possibly exacerbate the delay. Its strange that so far you are the only one reporting it as a major issue. I can feel it on mine when I'm looking for it but its really tiny and not a safety concern at all on my setup.

How are you figuring your controller is drawing 18kw with it being just a 12 fet? At 100v and no sag that would be 180a battery current which should vaporize a 12 fet. Even my 18fet at 125v 110a is only around 13 kw input and it won't handle much more due to heat. I also have it heavily modded.

I am guessing your controller is an eb212 board and I have one of those on my 9c 8x8 26in setup at 4kw and I don't notice the delay on it.

Are you using a hall throttle or a resistor throttle?

 

[ZOMGVTEK]

Zomgvtek, is there a possibility that you have the max throttle v set too high. That could possibly exacerbate the delay. Its strange that so far you are the only one reporting it as a major issue. I can feel it on mine when I'm looking for it but its really tiny and not a safety concern at all on my setup.

How are you figuring your controller is drawing 18kw with it being just a 12 fet? At 100v and no sag that would be 180a battery current which should vaporize a 12 fet. Even my 18fet at 125v 110a is only around 13 kw input and it won't handle much more due to heat. I also have it heavily modded.

I am guessing your controller is an eb212 board and I have one of those on my 9c 8x8 26in setup at 4kw and I don't notice the delay on it.

Are you using a hall throttle or a resistor throttle?

I recently blew up a 24 FET. Expecting it's failure, I just left the 12 on the bike, and thats what I'm running now. The MaxV is set about 1/8 - 1/4 turn above where I no longer see any impact on the freewheel speed.

I don't really pay attention to what the bike draws. I've been slowly increasing current until the controller pops, then backing down until it is reliable. It's at something like 100-120A, but the shunt value is only calibrated at 20A. The controller is a bit modified. I do believe it is an eb212.

I can play with the thing a bit more when I get some time. I've been very busy lately, and I probably shouldn't even be spending time on here.

 

[John in CR]

I don't really pay attention to what the bike draws. I've been slowly increasing current until the controller pops, then backing down until it is reliable. It's at something like 100-120A, but the shunt value is only calibrated at 20A. The controller is a bit modified. I do believe it is an eb212.

That's the surest route to unreliability I've ever heard. I've never seen any controller that was reliable after popping that first time.