Do the old "infineon" controllers gradually reduce current to avoid the LVC?

[DanGT86]

Anybody remember if the old "Infineon" controllers reduced current gradually to prevent the low voltage cut from being triggered? I didn't think they were that smart. I thought they would crank out full current until they hit the low voltage cut set point.

Just dusted off an old Lyen 9fet trap controller to test ride a project bike. Its an old EB209 board for you fellow ES old timers. Since the programming software is extinct I have no idea what the settings are. From memory and the past motor I used it with its likely somewhere around 35 battery amps and likely 70-100 motor amps.

I have always run it on 20s Lipo packs so I'm sure I had the LVC value set somewhere around 65v. Might have been as low as 60v since I was always using a CA3 to handle the actual LV cut rather than letting the controller shut me down.

When I was riding it around today at full charge It was pulling 33amps. By the time I was sagging down to 66v I was only drawing 15amps battery at full throttle from a dead stop.

Maybe I am missing something obvious but what would be causing the reduction in current proportional with voltage? Seems to me that at low speed and from a dead stop it should always pull the same battery current regardless of input voltage?

 

[neptronix]

I don't remember them doing that, at least the EB2/3 models. Do you have an EB4?

 

[amberwolf]

Anybody remember if the old "Infineon" controllers reduced current gradually to prevent the low voltage cut from being triggered? I didn't think they were that smart. I thought they would crank out full current until they hit the low voltage cut set point.

None of the ones I've had do, whether "generic" (like a couple from Methods) or custom-built (like Lyen), or others from various unknown sources. They just shutdown (stop powering motor) when the voltage dips low enough, pulling as hard as they can up till that point, and work again as soon as the voltage rises above that point.

The first one I ever had that could reduce power near LVC is the Phaserunner.

Since the programming software is extinct I have no idea what the settings are.

AFAICR the software can't read the settings anyway--it can only write new ones. But if you want to try it out, Anpaza's thread

XPD: open-source keywin e-bike lab replacement

Hello guys (and girls, any here? :shock:)! Last week I had some fun with a couple of Infineon controllers recently purchased from Ed Lyen, and while playing with it I managed to intercept the serial protocol between the software and the controller. The protocol is quite simple, and since I was...

endless-sphere.com

should still have a version there somewhere.

When I was riding it around today at full charge It was pulling 33amps. By the time I was sagging down to 66v I was only drawing 15amps battery at full throttle from a dead stop.

If your'e not using a CA with this one (measuring current and voltage with ??) that might be doing this limiting, then I don't know--the *watts* would be dropping with voltage, but the current should only drop if the resistance of the system is high enough to cause that to drop as the voltage drops.

This might happen if the battery itself is very high internal resistance...but it would cause pretty severe voltage drop under any load if this was true.

If the motor or connections to the motor / controller are high enough resistance, current would drop as voltage drops, but taht would be such a high resistance it would not perform normally even when at full voltage.

 

[DanGT86]

Thanks AW.

I am using a CA3 to monitor it but the CA does not show any limiters active. I also lied to it about my battery pack to drastically alter the LVC point from the CA. Also tried it in Bypass mode to eliminate CA limiting features. I suppose I could physically remove the CA throttle and go straight to the controller just to see what happens.

The voltage sag is pretty bad since this is an old tired pack I use to test ride projects. It was hanging out at 73v and sagged to 66v under load. But even with the voltage sag it seems like it would always pull full current until the controller's LVC is hit.

This thing runs great but the max possible current just drops gradually as the pack drains. The CA3 watt limit is set to 9999watts.

At first I was thinking maybe the battery voltage sag was low causing me to hit full speed for that voltage and that full speed was low enough to just not require full current. But the dead stop launches should rule that out.

As for the software I used to use XPD but the required files or drivers are dead now. I just meant I could have reprogramed it to known values and forced the LVC down to like 40v to see if it still happens.

I believe one user on that XPD thread made an open source arduino device but that stuff is all voodoo to me. I don't think there is any easy way for the avg semi computer literate guy to program the old controllers.

Weirdly the software I got from Lyen with my controller as well as the Keywin ebike lab or whatever it was called also dont work anymore. IDK exactly how something went back in time and destroyed them all but it is sad. I guess maybe a driver that didnt come along for the ride to windows 10.

I'll bypass the CA physically and see how it acts.

 

[DanGT86]

Just found all my old XPD controller programming files on my old laptop. Below is a screenshot of the most recent date of any of the files so I'm guessing that is what I had the controller set for. Thanks august of 2020.

I just bypassed the CA directly with the throttle. I then did some full throttle runs against the brake and saw the current spike to 33amps on the CA and the voltage drop to 55v.

Went for a very short ride outside and the power is oscillating badly and only pulling about 10 amps. The voltage during this was showing steady in the low 60s.

Based on this data I have a few theories:

• The CA was not likely the culprit
• The LVC is not actually 56.7v. This controller had input resistor mods to run a 75v pack. I believe I had to lie to LVC.
• My true LVC is likely closer to 66-70 v because I was scared/respectful of the Lipo I used to run. 70v=3.5v per cell
• The oscillation is the controller triggering the LVC then recovering then hitting it again.
• When the voltage was still 70ish the oscillation was probably imperceptible and just looked like low avg current.
• The CA refresh rate may be set too low to see this oscillation. I'll turn that up and try the grin analogger.

So now I have some more things to test but I am closer to an answer. In other good news my modded 12FET Lyen controller is set for 130phase amps and 65battery according the most recent file I found. That is the high end of ideal for this bike. I thought I was going to have to trash that old controller since I had no way to program it. I guess the last motor I tested with it was this one!

These trapezoidal speed based controllers with the CA3 are pretty old tech but there is no thrill like the violent acceleration they provided. I feel like these are the carbureted V8s of the ebike world.

Now its time to swap out this tape for a legit battery and mounting solution and do some wire management. Most of my projects stop evolving at the tape and velcro strap stage.