2WD trike pulling left when accelerating

[cboy]

However, from this diagram and your description it seems the controller is powering the relays - which is not good.

Just to clarify, the coil side of each relay is powered by the controller. And during these most recent test runs, NONE of the switches were turned on except for the forward/reverse switch which does not power a relay. The forward/reverse switch is shown in Kelly wiring diagrams as being powered from the controller. So the 12V draw on the controller was limited to only the forward/reverse switch...which is by Kelly design.

 

[teklektik]

And during these most recent test runs, NONE of the switches were turned on....

Yep - you were very clear about the relays being inactive which is why I deferred pressing the matter of how they were powered and why this relay thing wasn't mentioned as an explanation for the confusing results.

You indicated you wanted to return to the aux 12V stuff shortly and it seemed that would be the time to ask again - and here we are just as you said...

The forward/reverse switch is shown in Kelly wiring diagrams as being powered from the controller. So the 12V draw on the controller was limited to only the forward/reverse switch...which is by Kelly design.

Right, but this switch draws less than 5ma - it has no coil. Interestingly, the controller has a low voltage error for the 5V supply but not for the 12V supply. They call out a limit of 40ma draw on the 5V OUT pin but have no spec for the 12V supply.

This issue of drawing extra power from a controller comes up from time to time in the context of the 5V supply on ebike controllers. There is usually enough extra capacity to run maybe 50ma or so of extra stuff, but nothing in the 100's of ma. So - not an uncommon idea, but it just doesn't quite work out - you are not alone on that one.

 

[Alan B]

Finally, the smoking gun, Excellent!

 

[cboy]

I'm certain the problem is that that the relays are dragging down the passenger (right) controller power supply which was only sized to power the controller, not other stuff.

I guess I was a little slow on the uptake. Do I understand correctly that you are now suggesting that the problem does not lie on the grounding side of the circuit at all but rather on the power side of the circuit...that there is too much stuff running off the 12V supply pin provided on the Kelly controller and this is drawing away or diverting power from actually spinning the motor?

If that is the case then I'm still confused.

1) The forward/reverse switch including the LED indicator light (it's blue so the power requirements should be added in) draws a very small amount of current. And that is the only switch in the system that is on and drawing current. Are you saying that switch alone is enough draw on the controller to create such erratic behavior?

2) The draw from the switches and relays has been there since the beginning. And always it has been powered by the passenger side controller. So why, early on, did the bike pull to the driver's side when it was the passenger side being robbed of power?

3) The kelly wiring diagram shows the 12 volt supply powering, at the same time, the forward/reverse switch, a "foot switch", a brake switch and a two-speed function switch. I am powering only one of those, the forward/reverse switch. So it would appear to me the controller is capable, at least in Kelly's eyes, of powering much more than I am powering without any negative effect on controller performance. Perhaps, if I had all of my switches and relays turned on the draw might be equal to or more than in the Kelly drawing, but I have none of my switches or relays turned on...with the exception of the forward/reverse switch.

Perhaps I'm still missing something here but it seems to me the "power side" theory doesn't fully explain what is going on.

 

[cboy]

Having thought about this overnight, possibly, early on, there were multiple problems going on. One sort of masking or overcoming another. When the "torque" issue was improved the "current stealing" by the switches/relays showed its face. And that might explain why the problem shifted from pulling left to pulling right. Anyhow, to see if we are on the right trail I'm rewiring things today to get everything 12V operational but off the controllers.

 

[teklektik]

Okay - this is long, but you asked some good questions, so...

Do I understand correctly that you are now suggesting ... that there is too much stuff running off the 12V supply pin provided on the Kelly controller and this is drawing away or diverting power from actually spinning the motor?

Yes. The controller has two parts: the electronics part with the microprocessor - basically all the 'smart' stuff connected to those RTN/5V/12V connections, and the power side that switches the motor that is connected to the giant B+/B- terminals. This is like the brain and the muscles. The problem is not so much drawing power away from the motors because of hooking up a few relays, it's rather that hooking up the relays uses up the tiny bit of power to make the electronics work and the controller goes brain-dead so it can't properly control the motors. Think head-lock to cut off a little blood to the brain so the guy passes out - his muscles still have lots of blood and 'could' work, just no one is home to command them.

Here's some background thoughts:

There are some issues that you have experienced that have baffled posters and some other issues for which explanations seem plausible, but in the end there remains behaviors that no one has been able to explain. So, we plow on...

Your project, data collections, and wiring has gone through a number of iterations. At each stage the idea has been to try to understand the present symptom and offer a remedy for that (new) difficulty while looking back with the hope that the remedy could explain earlier symptoms. This means that some unexplained behavior has been taken as an unknowable divine mystery and efforts redirected to the new issue without a full understanding of everything. This is an unsatisfying way to proceed, but getting the thing working is more important to you than turning it into a research project for the simple sake of understanding.

As things have moved along, I think perhaps it wasn't clear that trying to understand past mysteries was no longer an overriding focus. Just trying to eat the elephant a bite at a time...

So ---

Do I understand correctly that you are now suggesting that the problem does not lie on the grounding side of the circuit at all but rather on the power side of the circuit...?

>> The grounding side is a mystery. That was the whole business with the ground cables and it makes no sense. So, moving on...

>> The power side (from the motor perspective) is a bit of a mystery as well. I have alluded in earlier posts to the idea that with the present phase current limit of 380A it doesn't seem the controller is actually operating with 'torque throttle' on these slamming getaways. It seems that things are going straight because the motors actually are quite close in performance and not because of the Kelly torque throttle feature. I believe that to get that working correctly requires backing down the 'Current Percent' as mentioned earlier. That said, this is based on supposition and you have been able to get satisfyingly straight operation, so that seemed more important than pressing experiments to satisfy my curiosity. I'm happy it is working for you, but it seems to be working by accident, not because of controller regulation - so - confusing.

>> The power side (from the controller electronics perspective) seems a questionable approach, but easily remedied. I think this may be a participant in the puzzle, but not an explanation for all the behavior that has been seen. More on this below.

1) The forward/reverse switch including the LED indicator light (it's blue so the power requirements should be added in) draws a very small amount of current. And that is the only switch in the system that is on and drawing current. Are you saying that switch alone is enough draw on the controller to create such erratic behavior?

No. At this point, all the previous L/R torque imbalances (pulling this way or that) are not really the focus - no one has been able to explain them. Perhaps the way the relays have been wired could have participated but we just don't have the details about what was ON or OFF, or whatever. The recent remarks about the relay power are based only on the present experiment that shows loss of torque on the passenger side. It appeared that we can explain ths latest symptom (but maybe not - see below). Earlier stuff is still in question...

2) The draw from the switches and relays has been there since the beginning. And always it has been powered by the passenger side controller. So why, early on, did the bike pull to the driver's side when it was the passenger side being robbed of power?

As mentioned above, I would very much like to understand the earlier mysteries, but only have a working theory that can explain the present right-pull symptom now that the wiring has been reduced to the bare minimum. The evidence is very strong that this is a problem, and a remedy for this particular symptom is pretty clear. Here the idea is to start simple and work back up to the full wiring instead of trying to diagnose the whole thing at once. I believe that in the end there will have been more than one problem, that this is one of them, and that alone it cannot explain everything.

3) The kelly wiring diagram shows the 12 volt supply powering, at the same time, the forward/reverse switch, a "foot switch", a brake switch and a two-speed function switch. I am powering only one of those, the forward/reverse switch. So it would appear to me the controller is capable, at least in Kelly's eyes, of powering much more than I am powering without any negative effect on controller performance.

Switches do not draw power and as you point out, Kelly expects them to be there. However, LEDS, relays, throttles, etc do draw power. Here it seems clear that the relays are power hogs and could reasonably overtax the controller power supply. This is speculation based on experience and because Kelly provides no specification or exemplar wiring illustrating the practice. To clarify this point, I encourage you to contact them about drawing an extra 1/2A to power ancillary equipment. In the absence of more information I can only offer that this appears very questionable and remedying it with an unquestionably 'safer' design seems recommended.

As I said above, I was concerned about this earlier and asked for details. Even if the relays do not appear to be participating in the present right-pull behavior, I remain concerned about the design and strongly recommend that you query Kelly on the matter. In the end, it's just one guy's opinion and it would be good for Kelly to offer up contradicting reassurances...

3) ... Perhaps, if I had all of my switches and relays turned on the draw might be equal to or more than in the Kelly drawing, but I have none of my switches or relays turned on...with the exception of the forward/reverse switch.

Whoa. I believed in the post above where you described connecting up the relays and experiencing occasional right-pull that the relays were actually being switched ON. Here I could see that maybe one relay was okay but 3 or 4 would precipitate the problem. If you are saying that simply connecting the relay grounds causes this sporadic behavior with all the relays OFF, then there is yet another odd behavior to explain.... The details of these most recent experiments need clarification.

Perhaps I'm still missing something here but it seems to me the "power side" theory doesn't fully explain what is going on.

Nope. Never intended that it did. Again, I think there may well be more than one bug on the loose and this is just a matter of squashing each as they are revealed.

• I cannot tell you the number of times that I have dealt with engineering problems where not only were there more than one bug, but that the errors canceled each other out in whole or in part. Fixing one just made things worse until other issues were corrected. Argh.

EDIT - oh crap - you cross-posted in between. So - I think you have worked out much of what I have prattled on about here...

 

[cboy]

Okay - this is long, but you asked some good questions, so...

You are a very good teacher of things electronic...so I don't mind "long". Helps me to understand and diagnose not only this issue but future issues and future projects.

The problem is not so much drawing power away from the motors because of hooking up a few relays, it's rather that hooking up the relays uses up the tiny bit of power to make the electronics work and the controller goes brain-dead so it can't properly control the motors.

Makes huge sense. It explains why just a bit of misplaced wiring can have a massive impact on performance.

Switches do not draw power and as you point out, Kelly expects them to be there. However, LEDS, relays, throttles, etc do draw power. Here it seems clear that the relays are power hogs and could reasonably overtax the controller power supply.

I see this point now. When I think "switches" in any vehicle, car or bike, I think relays. In virtually everything I build I never run full power through a switch...the full power is always run through a relay. So, for example, when Kelly shows a "brake switch" they are probably just talking about the brake cut-off switch to eliminate power to the controller. And I'm thinking "brake switch" to run the tail lights. So yes, a large relative difference in power draw.

In the absence of more information I can only offer that this appears very questionable and remedying it with an unquestionably 'safer' design seems recommended.

This is a time consuming but relatively easy change to make. In fact it was already on my agenda today to rewire all the switches and relays over to the auxiliary battery and/or the dc/dc converter. I'm still working out in my head a couple minor details for doing that but its where I'm headed.

Whoa. I believed in the post above where you described connecting up the relays and experiencing occasional right-pull that the relays were actually being switched ON. Here I could see that maybe one relay was okay but 3 or 4 would precipitate the problem. If you are saying that simply connecting the relay grounds causes this sporadic behavior with all the relays OFF, then there is yet another odd behavior to explain.... The details of these most recent experiments need clarification.

The switches and relays are wired up to the controller 12V supply pin but only one switch is turned on, and that is the forward/reverse switch which is needed to engage the motors. And as noted, this switch does not operate a relay but it does turn on an LED indicator light. While the LED and connection back to the controller would be a very small current draw, it would be something rather than nothing. So I see it as potential interference with the inner workings of the controller. I think we'll know more once I get everything rewired and off the controller supply pin.

I cannot tell you the number of times that I have dealt with engineering problems where not only were there more than one bug, but that the errors canceled each other out in whole or in part. Fixing one just made things worse until other issues were corrected.

Yes, a bit of cross posting there but reaching the same verdict. There may have been multiple gremlins trying to sabotage our "expert" troubleshooting. So my next step is another rewiring marathon to get the entire 12V system working while keeping everything possible off the controller 12v supply pin.

 

[teklektik]

...my next step is another rewiring marathon to get the entire 12V system working while keeping everything possible off the controller 12v supply pin.

is good.

 

[cboy]

Another interesting tidbit to add to pile of questions. I have everything now off the 12V controller supply pin and all the switches and relay coils are now powered with the 12V aux. battery, including the forward/reverse switch. But when I tried to test this configuration the motors wouldn't spin. The controllers come on fine with green lights and I had 12 volts to the forward and reverse pins at the controllers (so the switch was working) and correct voltage on the throttle pins at the controllers. But there was no current being supplied to the motors. So I kept everything else the same but unhooked the forward/reverse switch from battery power and put it back on the controller 12V supply pin and ran the ground for the switch's LED to the shunt. And the motors both ran. So I'm thinking the controller is somehow programmed to operate only on its own 12V supply (which is actually more like 10.4 volts on the multimeter) for a forward or reverse signal. This means that I can't get absolutely everything off the controller 12V supply. But if I disconnect the ground for the switch, at least the LED won't be drawing any power. I haven't tested on the road yet but my guess is, if the LED isn't powered, that the bike will run straight. I just won't be able to have the indicator lights to show which "gear" I am in before taking off.

 

[cboy]

The 12 volt system is now functioning with everything but the forward/reverse switch operating on battery power. I also did not ground the LED on that switch for fear the current draw might mess up the controller again. Did some road testing with pretty good results except for the very initial hit of the throttle when it pulled right a bit. But after that it ran straight under acceleration every time I punched it. I feel it is totally driveable with this configuration but I'd still like to do two things:

1) Somehow get the forward/reverse indicator LED to operate without messing up the controller
2) Get the dc/dc operating so I don't have to keep charging the aux battery...or get rid of the aux battery altogether and just run the converter. This will require grounding the dc/dc back into the battery pack in order for the CA to keep track of total amp usage. And who knows, this might fowl up the works again...so I may take a day or two before I start tackling those two items.

 

[cboy]

... I have alluded in earlier posts to the idea that with the present phase current limit of 380A it doesn't seem the controller is actually operating with 'torque throttle' on these slamming getaways. It seems that things are going straight because the motors actually are quite close in performance and not because of the Kelly torque throttle feature. I believe that to get that working correctly requires backing down the 'Current Percent' as mentioned earlier.

I also want to revisit this part of your thinking and maybe tinker a bit on this front now that we have a decent baseline configuration which seems to be working. Would I be correct from reading this that I might actually be able to improve performance by limiting the percentage of phase current? As I understand the theory here, the upper end of the 380 phase current limit is really quite far out of reach for my set up and my battery pack. This is born out by the fact that even under full acceleration I've never seen my CA register much more than about 100 amps being delivered. For lack of a better description, there appears to be a lot of slop in my settings. And if I limited the percentage of phase current to say 50% or less on both controllers it may well make the controllers much more responsive to the throttle...possibly even delivering more torque during acceleration than with the phase current at 100%.

This may be an over simplification, but limiting the phase current might allow the torque curve to more accurately match the throttle curve. Obviously, a lot less current being delivered at the top end but potentially more current being delivered through the lower and middle ranges of the throttle curve.

It also seems this would be a very easy series of tests to run. I just start limiting the % of phase current in both controllers and we see what happens.

 

[teklektik]

So I kept everything else the same but unhooked the forward/reverse switch from battery power and put it back on the controller 12V supply pin and ran the ground for the switch's LED to the shunt. And the motors both ran. So I'm thinking the controller is somehow programmed to operate only on its own 12V supply ...

This means that I can't get absolutely everything off the controller 12V supply. But if I disconnect the ground for the switch, at least the LED won't be drawing any power.

Don't worry about the FWD/REV switch. As you pointed out previously, this was there by Kelly design and actually draws no/little extra power per se. I also wouldn't worry about the LEDs, again very little power - at least compared to the big offenders - the relay coils.

The reason that the FWD/REV switch didnt' work when run off the Aux 12V supply was not the source of the 12V, it was the absence of common Gnd to give the controllers a reference for the voltage - it basically looked like 0V.

• FWIW:
  The FWD/REV switch and/or regen throttle can be arranged to operate regardless of whether either controller is powered off by hooking them up with a couple of diodes to provide 'shared' 5V or 12V sources. In reality, only one controller will be providing power since the voltage sources from the different controllers will not be exactly equal, but the other controller will take over if needed. This is of no consequence at all in normal operation, but if you would like to better support redundancy to ensure a painless 'limp-home mode' on only one controller, then this is an easy addition.
  
  The resulting voltage is slightly depressed for throttles but this shift can be configured away for controllers with an adjustable throttle range. There should be no impact at all for simple switches like the Kelly FWD/REV, brake, etc. Common 1N4001 diodes or similar are fine and will introduce about a 0.6V drop. If you wish, you can halve that voltage drop by using a couple low power Schottky diodes instead. Either work - depends on what parts you have on the bench - Schottkys are arguably a better choice for throttle applications.
  
  Mounting tiny diodes in a hefty vehicle harness can be a PITA largely because of the fragility of the parts. I recommend adding a 3 pin female connector (e.g. JST3) and just jumpering the diodes across the pins of the mating male connector. Cover in some heatshrink and you have a nice robust and replaceable part.
  
  
  

 

[teklektik]

...I'd still like to do two things:

1) Somehow get the forward/reverse indicator LED to operate without messing up the controller
2) Get the dc/dc operating so I don't have to keep charging the aux battery...or get rid of the aux battery altogether and just run the converter. This will require grounding the dc/dc back into the battery pack in order for the CA to keep track of total amp usage. And who knows, this might fowl up the works again...so I may take a day or two before I start tackling those two items.

1) As mentioned above, hooking the LED to the controller should have no effect.
2) The idea of a common ground is something that is ideally avoided as a best practice, but sometimes becomes an economic necessity. This is really a no-no for high voltage vehicles where having a ground hooked to controls, etc with a 300V traction battery is just silly. In your case - no problem - but a live frame ground is still to be avoided.

On the downside: disconnecting the common ground seemed to correct some very strange behavior and a frame leak or other issue appeared in play.
On the upside: so many odd things were seen afterwards that it's unclear that hooking up this ground will actually have a adverse effect... (?)

 

[teklektik]

This may be an over simplification, but limiting the phase current might allow the torque curve to more accurately match the throttle curve. Obviously, a lot less current being delivered at the top end but potentially more current being delivered through the lower and middle ranges of the throttle curve.

It also seems this would be a very easy series of tests to run. I just start limiting the % of phase current in both controllers and we see what happens.

I think you are spot on here.

If you haven't worked through this already, I would start by revisiting your CA and controller throttle OUT and TPS settings to get these matched. For instance, the Kelly TPS Lo/Hi percentages are set to 20% and 80% respectively. This translates to 5V*20% and 5V*80% or 1V and 4V respectively which is pretty much 'hall normal'. But, you're driving the controller with a CA not a hall throttle so the exact range isn't that important - as long as the 'TPS Lo/Hi Err' settings are in the lo hi dead zones to detect throttle failures. You presently have:

TPS Hi Err = 95
TPS Daed High = 80
TPS Dead Low = 20
TPS Lo Err = 0

There's no strong reason to fiddle with this just now...

This suggests the CA MinOut/MaxOut should be something like 0.95V and 4.05V assuming a 0.05V deadzone (you can fiddle that margin). Ideally you want to look at the Throttle voltages on the controller monitor screen since it's likely that the controller and CA see voltages a bit differently. Anyhow, getting the throttle tuned up will ensure that your subsequent controller adjustments will be playing over the full controller throttle range with the maximum matched Kelly/CA throttle range.

Kelly is not too helpful with "TPS Fwd/Rev MAP" settings except to note they are used to control acceleration. My inclination here is to leave them alone for now but perhaps at some point adjust them for strong acceleration and control the actual behavior with the more easily adjustable CA UpRate and FastRate settings. Deferred experimentation....

Again, if you haven't done so already, you want to set up limit the max battery current for your rig. An idea here would be to leave the Kellys set to the max 'Battery Limit' of 100% and then configure the CA3 for the max safe battery current. Kelly actually has a kind of messy scheme for setting battery current - it's set as a percentage of phase current and so must be adjusted whenever phase current is adjusted. This is similar to the Infineon adjustment scheme, but backwards.

The max phase amps for the controller cannot be adjusted and for your controllers is:
PhaseCurrMax = 380A

The usable phase current is:
PhaseCurrMax * CurrentPercent = 380A * CurrentPercent

The battery current is:
PhaseCurrMax * CurrentPercent * BatteryLimit = 380 * CurrentPercent * BatteryLimit

So - to set you motor (phase) currents to 180A, you would set
CurrentPercent = 180A/380A = 47%

For a battery limit of maybe 120A, you would leave BatteryLimit at 100% and set the CA PLim->MaxCurrent to 120A. This would suppress Kelly limiting in favor of the more readily adjustable CA limiting.

Anyhow, just some thoughts/guesses on initial setup...

 

[cboy]

Anyhow, just some thoughts/guesses on initial setup...

Round #1 test

In each controller program I made these changes:

Changed battery limit to 100% (previously was 50%)
Change Phase Current limit to 47% (previously was 100%)

On CA3

Changed P-lim to 150 amps (previously was 300 amps)

Road testing did seem to have a little more pep in the throttle response and the bike seemed to "wind up" to speed more quickly. Also the trike ran straight and true under both light and heavy acceleration...and that's always a plus.

 

[teklektik]

...the trike ran straight and true under both light and heavy acceleration...

Woohoo!

If the throttle ranges are matched and you battery current limit is maxed out, then there aren't any further adjustments that will directly improve performance without impinging somewhat on the torque control. For instance, increasing max phase current will detract from the controllers' ability to control torque at speed which could re-introduce torque-steer issues. There might be some room to fiddle there, but not huge amounts - perhaps a research effort for another day. There remains the question of the Kelly acceleration parameters which sound as if they will have some effect on throttle responsiveness (as with the CA UpRate/FastRate), so there might also be room for some improvement in that quarter at the cost of a twitchier throttle - personal taste. Otherwise, it sounds like you have a baseline tune running pretty well.

 

[cboy]

There remains the question of the Kelly acceleration parameters which sound as if they will have some effect on throttle responsiveness (as with the CA UpRate/FastRate), so there might also be room for some improvement in that quarter at the cost of a twitchier throttle

I think those Kelly and CA3 settings will be next on the list to see if I can get a little more acceleration at get-a-ways. It's the old drag racer's lament. There's never enough torque until you break the tires...if you break the tires, install wider and stickier ones...there's never enough torque until you break the tires....

 

[wturber]

There remains the question of the Kelly acceleration parameters which sound as if they will have some effect on throttle responsiveness (as with the CA UpRate/FastRate), so there might also be room for some improvement in that quarter at the cost of a twitchier throttle

I think those Kelly and CA3 settings will be next on the list to see if I can get a little more acceleration at get-a-ways. It's the old drag racer's lament. There's never enough torque until you break the tires...if you break the tires, install wider and stickier ones...there's never enough torque until you break the tires....

My favorite from hot rodding has always been, "If a little bit is good, then too much should be just right."

 

[amberwolf]

I like to get torque up to to breaking axles, then upgrade the axles until the frame breaks, then upgrade the frame till the axles break, and so on.

 

[Alan B]

I dialed the torque back until the front wheel stopped coming off the ground. More torque just makes waste heat and creates handling problems. The front gets light and when I pedal under full torque the front bounces a bit. Works for me.

 

[cboy]

I tinkered a bit with the CA3 Up Rate, Fast Rate and Fast Threshold. I set them at the high end of the "typical" or "recommended" settings... Up Rate=3, Fast Rate = 8 and Fast Thrsh =2. This did seem to improve acceleration...although not radically. And no negative effects like twitchy throttle at this point. Are there any risks if I keep adjusting upwards?

 

[teklektik]

You typically want the Fast Threshold to be slightly higher than the unloaded WOT speed. I'm thinking your motors draw more than 2A unloaded

To get the CA out of the picture entirely, just crank all the rates up to 99V/sec. Then you will be looking at unadorned controller performance. These settings aren't going to make a huge difference after a second or two, but if you're looking for chirp and a jolt, then increasing the rate of power application will help.

It's unclear to me what the Kelly adjustments mean - an email to Fany might be clarifying. Anyhow, those hold some promise as well. Assuming they function similar to the CA3 settings, I would crank one or the other up all the way so it is not participating, then restrict tuning to the other functioning set of adjustments. This 'should' minimize CA-controller interactions and make the tuning results easier to analyze.

• With a Phaserunner + CA3 the ramping functionality is very similar and the idea is usually to make the PR ramp super fast and tune via the CA - but the situation with the Kelly may differ.

FWIW - there are undocumented Rate Flags that appear on the Diagnostics Screen in the space after the 'Out'. Little letters appear (U,D,F,P) to show you if the Up, Down, Fast, or PAS rates are in play. This is really pretty revealing when tuning - it shows you exactly what is controlling CA behavior. It frankly surprised the heck out of me and led to an immediate round of retuning...

 

[cboy]

FWIW - there are undocumented Rate Flags that appear on the Diagnostics Screen in the space after the 'Out'. Little letters appear (U,D,F,P) to show you if the Up, Down, Fast, or PAS rates are in play. This is really pretty revealing when tuning - it shows you exactly what is controlling CA behavior.

All very helpful advice. I think I'll start by taking the CA3 out of the picture (setting 99v/s) and then work backwards from there with the controllers and then put the CA3 back into play if necessary.

 

[cboy]

Short update on tinkering. I took the CA3 out of the picture by setting Up-Rate and Fast-Rate both at 99Volts/second. I did a short test run with the Kelly TPS Forward MAP at the default setting of 30. This produced improved smooth and solid acceleration but still nothing earth shattering. I then reset the TPS Forward MAP (which Kelly says is for acceleration) to 75 (the range is 0 to 100). This had little if any noticeable impact on acceleration performance. The only other acceleration related setting that I can see is something called "Accel Time" which the manual says is the time (in seconds) for TPS torque to go from zero to Max. The range for this setting is 1-250 and as best I understand the manual, the settings are in tenths of a second. For example, the setting of 5 equals 0.5 seconds to go from zero to max torque. The setting is already at 5 (or .5 seconds) so not much room left for increasing the torque application time but all sorts of room to DECREASE the time it would take to go from zero to max torque. So I've left this setting a 5.

My takeaway from this bit of tinkering is that I may just be at or near the best acceleration performance I'm going to see given the weight of the vehicle and the torque I have available in terms of battery pack/controller. It is quite satisfactory at this juncture but I'm not going to be winning any drag races. I'm going to do some longer range testing to see if the TPS Forward MAP settings are having any impact on efficiency. If they do, my final adjustments will be to maximize my "range" while maintaining decent acceleration.

 

[Alan B]

With six each 62 pound lead acid batteries the acceleration is dominated primarily by weight, and secondarily by voltage sag.

125 amp hours times 12 volts times 6 batteries times 50 percent is 4.5 kwh useable capacity. Using more than 50% capacity will destroy lead acid batteries very quickly.
Weight is 6 times 62 pounds equalling 372 pounds, plus support bracketry for this battery.

The Multistar 6S bricks I used in the Borg build weigh 1.11kg each and produce 22.2V at 8000 mAH. Useable capacity is about 90% of that, or 158 watt hours per brick. A similar pack (to the above lead acid battery pack) would take about 28.5 bricks for a total weight of 31.6 kg which is under 70 pounds, for a weight savings of over 300 pounds, and with much less voltage sag. I used 12 of these on the Borg which is linked in my signature, and if allowed it will lift the front wheel during acceleration. It is more than adequate.

While I don't recommend hobby Lipo for inexperienced users, the difference in performance between lead acid and lithium batteries is game-changing. Ebikes and electric vehicles would not be practical (or fun) today without advanced lithium batteries. For a trike such as yours I would look into used Electric Vehicle modules such as those from the Leaf or Volt if you want increased acceleration and a lighter machine.