Need LOTS OF help with an upgrade I've ben working on.

[Mcmattwich]

So I agreed to help upgrade a bike I was contracted to build a battery for. I have built my own bikes, so the required controller upgrade (was going for a higher voltage battery so was necessary) seemed straightforward enough. But it's been an array of problems since . Here are the stats as I know them:

1. Bike is sondors original fat tire, 750w motor upgrade done prior to me

2. Had kt 52v controller originally with lcd3

3. Now running 62.9v battery, with a controller by NB power on amazon, 72v 40 amp, there's no number on it, but it says DC moto by lithium power on it. Sw900 screen


Problems

1. On take off from a dead start with throttle, or high level PAS, the motor wants to make a grinding noise. I'm guessing too high of an amperage, however, setting p14 on the display is not letting me change it.

2. The speedometer is wayyyyyy off. Like it's showing 62mph at a trot.

Please help. I'm extremely handy if any physical changes need to be made to the controller, but I've had this thing for a while. He's patient, but I don't want to be that guy that failed to accomplish something I was supposed to be good at, I've never had this much trouble with any build or upgrade in my life. And it seems like simple stuff

 

[AHicks]

That buzzing noise is often heard on square wave controllers especially when accelerating, and often very noticeable when accelerating from a stop. I wouldn't have one personally. Unfortunately, KT offers both square and sine wave. The sine wave is silent.

If this new controller works like a KT controller (trimming PAS power using P14?), if P14 doesn't allow enough adjustment for you, go to P5 and trim the amount of available amperage a bit. You can then go back and use P14 to trim.

The speedo problem is often P1 I believe. Plug in something different there to verify. I've had to ride next to another bike to get that setting right.

 

[Mcmattwich]

This grinding noise isn't the controller, it's not the original controller he wanted to go with, and the first one I put ondid the same thing (sabvoton clone). But it was also giving problems with jumpy starts. I know the noise you mean, and this one IS noisier on takeoff than the original. But this one isn't a KT the original was; the ktlcd3 wouldn't work on this even, which is why I went to the sw900.

It's worth noting that it IS literally a grinding sound from the hub motor, the kind that IMMEDIATELY makes you think oh god, the planetary gears. But a check showed them to be perfect.

Batteries are more my thing as a side gig, but I thought I had this in hand. Nothing has seemed to go right with it, but I'm staying quite hopeful. And I'm going to be professional in all endeavors I do for people, period, so if I have to, I'll replace the whole damn system on my dime before I let this thing have someone thinking I won't do right by people.

Also thank you for your reply, I was scared I'd get nothing because this site is so full of people! But it's always my go to, just listening to past posts has solved so many problems for me

 

[Mcmattwich]

Also my P1 is brightness?

 

[AHicks]

Sorry, can't help you then. I know the KT stuff fairly well, but clueless regarding what you are looking at.

The buzzing generated by the square wave will be heard at the motor. The motor is buzzing, not the controller. Loudest when starting from a stop or accelerating, tapering off during cruise/holding a steady load.

 

[E-HP]

What's the no load (wheel off the ground) current/power?

 

[Mcmattwich]

Hm I didn't know it would be heard at the motor, but again, this definitely made the same noise on a sabvoton, which is sine wave. Plus, this is like metal on metal sound, though there's no evidence of that being possible as nothing is damaged. Visibly anyway. Inside the motor. The best way to describe it is when you brake a car that has completely eroded the rotor.

The wattage (I'm assuming that's how I'm intended to find the amps it's giving) is problematic; it displays the watts on the SW900, but they are presumably off. There's no voltage setting on it for 62.9 (17s) so I assume that what it's reading is not correct.

 

[E-HP]

The wattage (I'm assuming that's how I'm intended to find the amps it's giving) is problematic; it displays the watts on the SW900, but they are presumably off. There's no voltage setting on it for 62.9 (17s) so I assume that what it's reading is not correct.

The setting only dictates the low voltage cutoff. What's the no load power?

 

[Mcmattwich]

So what lifted off the ground, and then hit the throttle and see what it jumps to?

 

[E-HP]

So what lifted off the ground, and then hit the throttle and see what it jumps to?

Correct. Initially the wattage may go up, but once it's spinning, it should settle to 30W-40W or so.

 

[Mcmattwich]

Ok I'll check. While I do that, how high should it go? And if the lvc is the only Factor affected, how would the bike know what to display for the wattage if there is no voltage known to this equation? I mean I get that it can tell the amps, but without knowing a nominal to multiply it by how would it be able to get the watts? A 52 and 48 volt battery could both be within the same range, and still above the lvc for example

 

[E-HP]

Ok I'll check. While I do that, how high should it go? And if the lvc is the only Factor affected, how would the bike know what to display for the wattage if there is no voltage known to this equation? I mean I get that it can tell the amps, but without knowing a nominal to multiply it by how would it be able to get the watts? A 52 and 48 volt battery could both be within the same range, and still above the lvc for example

One of the input wires going to the display is battery voltage, so actual voltage known to the display. Nothing to do with the parameters set with the display. Current is imputed via a shunt in the controller.

 

[Mcmattwich]

Yeah I get that. But the wattage being produced is based off the equation v x a=watts right? If it's only reading the voltage of a battery as it sits, then doing that equation for a 48v and a 52v, both of which are options in the parameters, would show different results, but a battery at 50v could be either of them, so how would it know?

Btw doing the check right now

 

[Mcmattwich]

Ok so about the lowest it drops to is allegedly 200 watts

 

[E-HP]

Ok so about the lowest it drops to is allegedly 200 watts

Your phase and/or hall sensor wire combination is off. You're somewhere in the two right boxes, from what you're describing. High no load current/power indicates a mismatch.

I think you can fry your motor if you run it like that WOT.

 

[Mcmattwich]

Ok yeah I haven't continued. It made the noise and I stopped. I'll give them a switch around. So it's because non industry standard color coding pretty much? Not that there is an industry standard, just saying they're different

 

[ZeroEm]

by Mcmattwich » Aug 16 2021 12:30am

Yeah I get that. But the wattage being produced is based off the equation v x a=watts right? If it's only reading the voltage of a battery as it sits, then doing that equation for a 48v and a 52v, both of which are options in the parameters, would show different results, but a battery at 50v could be either of them, so how would it know?

He answered your question, don't think he will repeat it. Your equation is correct but it is not being used. Think of a water meter. The shunt detects the flow and reports it. The voltage setting is for other functions, For one it describes the type of battery pack so a guesstimate of SOC.

 

[Mcmattwich]

Ok I mean it's not important to the actual process and solution here, which I'm grateful for. But it's not quite answered either. The sw900 I have can be set from 24v to 72 in the settings. (But 17s isnt one) The display reads out watts, that's it for the power measure. I get how using the voltage would have the SOC, if I switch it between 60 and 72, I see the battery meter go down.

I'm not disagreeing that the display isn't using the voltage to determine the watts. I'm saying that if it's not factoring in what voltage battery it is reading in any way, on a display that could feasibly be numerous sized of a pretty large range, then I want to know how it displays watts, because it needs the other side of the equation to know it.

 

[E-HP]

I'm saying that if it's not factoring in what voltage battery it is reading in any way, on a display that could feasibly be numerous sized of a pretty large range, then I want to know how it displays watts, because it needs the other side of the equation to know it.

The display is basically a wattmeter. I can display battery current, derived from the voltage drop across the shunt, and it gets battery level voltage through the harness. So, with those basic inputs, it can display voltage, current, and watts.
The battery level, like the thing on your phone, uses logic to determine 100% vs 50%, based on what kind of battery you have (the voltage, relative to the packs nominal voltage).

It's a cheap controller, so they give you set options, e.g. 48V, 60V, 72V, etc., and only the common serial numbers (13S, 16S, 20S, etc.). It uses that information to determine the low voltage cutoff. If you have a programable controller you can program the cutoff independently, but for most cheap controllers, you get a few set points.

PS. I've never used your display, but the display/controller interface usually has at least 5 conductors. Battery (+), GND, signal wires for send and receive, and a conductor for turning the controller on and off.

 

[markz]

Ditch the KT, get a generic controller sensorless and run that besides its always good to have a spare.

Not wishtanding, go through every parameter thrice over and be sure you understand what each parameter is and in the meantime triple check your wires.

 

[Mcmattwich]

Okay will do. See this turned out to be important after all because it seems to me that something I thought I knew about batteries I was incorrect about, I always took the voltage in that equation to be the nominal vault of the battery, what it was considered like a 48-36 etc. wattage is actually measured based on the current state of charge voltage?


See I'm glad I pressed. All right guys I'm going to go give this thing a shot with the switching of the wires, I understand all of the parameters and what they do, except for one, in the manual it says that 15 is unused.

 

[ZeroEm]

The labels are generalize, they give you a good idea of what you have. Like a 48V pack is a general term, it means 13 cells in series. Nominal voltage of most cells is 3.6. 3.6 x 13 = 46.8. That does not fit in to the 12v car battery world, four car batteries is 48v.

 

[Mcmattwich]

I get that, I build battery packs. I just was under the impression the voltage for figuring watts was a constant in the equation

 

[Blacklite]

I get that, I build battery packs. I just was under the impression the voltage for figuring watts was a constant in the equation

They are all instantaneous measurements, so at a particular moment in time. Voltage (potential difference) is amount of energy per unit of charge. Current (amps) is amount of charge per unit of time. Multiplied together the charge terms cancel and you get energy per unit time, which is power (watts).

 

[ZeroEm]

Opps :lol: I miss spoke. Pardon my presence.