EB3XX controllers battery/phase/regen amp settings vs shunt

[zombiess]

I'm running a heavily modded Lyen EB318 controller (details here http://endless-sphere.com/forums/viewtopic.php?f=2&t=33847 and )

I also added a piece of 12 gauge copper wire in parallel with the shunt which dropped it down to around 0.5xx mOhm. The controller expects to see a 1.5 mOhm shunt from what I can tell. Directly from Lyen it comes with 3 shunts and I calculated my shunt at 1.2xx mOhms so I needed to set my battery/phase amps a little lower than desired when programming.

I've read that after modding the shunt you get a stronger regen, but that's not the case for me. I'm still only getting about 15A of regen after doing the shunt mod for other troubleshooting purposes. On the EB2XX controllers this worked so these new EB3xx controllers must be using a different measurement method. I have tested this on both a 9C 2806 and a Croatian Hubzilla motor.

I know when I was playing around with the EB2XX boards there seemed to be a connection between the battery amps and phase amps but I'm not as familiar with them on here as others. It appears to me modding the shunt and changing the battery current setting on the EB3xx boards has no bearing on the phase amp setting and they are completely independent of one another.

My current setup is running 60A battery and 100A phase which means I'm programmed for 20A battery 100A phase 1:5 ratio and now my take offs are much smoother and have scary torque. I haven't tried higher battery amps settings yet to see how this is effected but will post back when I do. I do know that at 100A battery 80A phase it pulls really hard but has a little stutter on take off until about 10mph if I'm not gentle on the throttle. I'm going to try upping the battery current to 100A (programmed for 33A) while keeping the phase current at 100A (programmed for 100A) and see if this changes.

Anyone care to speculate?

[Alan B]

Sounds like the phase amp limiting algorithm is unstable. When you set it high enough it doesn't kick in. When it kicks in it oscillates.

Is there a sensor for phase amps? If not it must calculate them from battery current.

[Alan B]

I would think you would have to multiply the phase limit the same as the shunt calibration. So your phase limit may be really high now. The limit in that case is probably not the controller and the FETs may be at risk.

[zombiess]

Sounds like the phase amp limiting algorithm is unstable. When you set it high enough it doesn't kick in. When it kicks in it oscillates.

Is there a sensor for phase amps? If not it must calculate them from battery current.

I looked for a sensor especially since I noticed an inductor type trace coming off each phase output that is not present on the EB212 boards, but I don't see anything that looks like a hall sensor, but then again everything is all surface mount. I didn't see any traced leading back to the ECU either. The inductor trace leads to an electrolytic which then goes back to the gate drive totem pole circuit. Without spending hours reverse engineering it I'm not sure how it's working (got too many other things to worry about).

I would think you would have to multiply the phase limit the same as the shunt calibration. So your phase limit may be really high now. The limit in that case is probably not the controller and the FETs may be at risk.

Phase is definitely limited by the setting. If I follow the 2.5x or even a 3.0x multiplier with the controller programmed for 20A battery (60A actual) and 60A phase it has problems accelerating from a dead stop and poor acceleration. Upped phase to 100A and kept battery at 20A and it's an entirely different animal and can loop me off the back from a dead stop. Programmed at 60A phase it has great acceleration and throttle response once moving above 15mph. Programmed at 100A phase the acceleration and throttle response seem the same to me once moving. Only change seems to be starting off and low speed.

I think the controller might be reading the actual phase output or making a damn good guess because changing the shunt should have changed the way regen kicks in.

Full size pics available of the phase section from an EB212 and an unbuilt EB312 board available here http://dynamic.opticalanarchy.com:8080/ ... ntrollers/

All the trace action on the EB312 is happening on the front. Sorry the on pic is a bit blurred but I just did some quick hand held shots. If you need detailed shots I can bust out the tripod/DSLR/macro lens and get you better pics.

[Alan B]

If there's no sensor other than the battery current shunt then that's what it must use.

We need better controllers. But we already knew that.

[zombiess]

If there's no sensor other than the battery current shunt then that's what it must use.

We need better controllers. But we already knew that.

Take a look at the pics and see if you notice anything I'm not seeing, you are probably more qualified than me, many of the part numbers are visible in the shots. I just find it interesting that regen didn't change at all after modding the shunt which leads me to believe they are somehow monitoring or calculating phase output really well.

On this same controller before the shunt mod I was running a programmed 45A battery (63A actual), 125A phase on my 9C 2806 and it was smooth with tons of torque. After shunt mod, same issues as hubzilla motor unless phase is cranked up like now.