Search all of my sites with Google

Wednesday, December 23, 2009

2QD Repairs, Modifications, Thoughts

The repair of the controller has turned into a modification session, too. So it is taking considerably longer than I originally expected or intended.

I'm upping it to 48V capability, changing out a few transistors and whatnot, as well as installing the whole thing into a better case that will provide heatsinking over it's whole surface. The case used to be for a Jensen 300W car AC inverter that doesn't output anything, which I gave up on fixing for now (I can always rehouse it later).

The extruded aluminum case is not very large but is big enough to hold the 2QD board with the power end cut off, 3 very tall low-ESR caps (75V 1500uF each) off that giant UPS board instead of the dinky one I could fit on the 2QD PCB, so much better power filtering, and plenty of surface area to bolt multiple MOSFETs to. Plus I can seal it up for watertightness and still vent all the heat outward, unlike the plastic PacTec project box I had it in before, with the tiny heatsink out the end of it.

I'm still waffling back and forth on whether to just use the TO220 style FDP038AN06 FETs I already use, or the higher-power-handling but higher RDSon TO264 style NTY100N10 FETs I used so well for so long on the original rebuilt ScootNGo controller.

Either way, I have to hand-wire the FETs to the control PCB, as there is no way to fit the whole PCB in there in a way that lets me bolt the FETs directly to the casing, without removing the FETs from the PCB. So I cut the entire end of it off from the gate resistors on, and will just run wires to each FET and then bolt them to the inside of the case, using insulating pads to keep them electrically isolated between the upper and lower halves of the halfbridge.

The case also has a spot for an automotive blade fuse. I am not using it now, but I may wire it in later, once I have an easy way to set a limit for the current on the 2QD. Right now the 2QD limits current via a voltage drop measured across the MOSFETs, so it depends on their RDSon. Since the RDSon of the MOSFETs I am using now is so low, around 3.5mOhms (actually half that since I've got parallel pairs), the current limit is apparently very high. Beyond 153 Amps, at least.

There is a current limit fine adjust resistor, and it's coarse adjust next to it, so I just have to figure out the circuit so I can work out what potentiometer I can install there to give me a good range of limitation I can use, for as low as 5A to as high as 50A (more if I have MOSFETs installed that can handle that). Then I can also use this to test for various different limitations, such as batteries that can't supply as much current, like my Li-Ion and a pair of well-used NiMH packs someone on ES is sending me (which might put out 7A on a good day).

Also so I can test for how well it performs at the actual legal max of 750W. ;)

I've also been charging all the SLAs I have around here for the last week or more. Some of the UPS ones simply won't charge well, but I am still letting them trickle charge in hopes they might recover at least a bit. I think that with the two newer ones that came in that most recent powerchair case, I will have 7 good ones in the 17Ah range, and 2 not very good ones, and one dead one. Plus the three 31Ah U1s I'd started with but changed out due to weight when I got the 17Ah UPS batteries. One U1 seems a little weaker than the other two.

I haven't done a full discharge test on any of them yet, but will do that using the Turnigy meter to get an idea of how many Ah I can realistically expect out of each one, including the three already on the bike.


  1. Do you have any circuit in mind to disconnect the battery test load at a given voltage? Not quite as nice as an automated test process but at least the Turnigy's meter will hold the AH reading (with a separate supply for the meter) until it's convenient to check up on it.

  2. Not yet, but probably something similar to what I suggested to someone as an active cutoff for their BMS on a LiFePO4 pack, to keep the BMS from actually destroying an already low-charge-state pack that can't be charged (for whatever reason) for a while after a ride.

    Basically, it's a relay to the BMS so that when it drops below some safe threshold just below LVC, the relay no longer holds closed, and the BMS no longer sucks power from the battery.

    Should be a one-transistor circuit, with a couple of resistors, probably. Maybe four, if you need a base and/or collector or emitter resistor. Plus an SPST or SPDT relay of minimal power handling ability. All it has to do is have a voltage divider on the transistor's base that turns it on only above the threshold value.

    The base would be energized from the BMS's pack voltage connection. The emitter would connect to BMS ground. The collector connects to one side of the relay coil, and the other side of the relay coil connects to the BMS's pack voltage connection.

    The NO relay contacts would connect the BMS itself to the pack voltage.

    If necessary I'd have a separate input power lead from the charger jack to the relay that would reengage it for charging up using the BMS, but I think the way I describe it above, it should automatically reengage once the charger is connected to the BMS.

    That same circuit could be used as an active cutoff for any load, too, but the relay would need to be significantly larger to handle the actual load. In BMS use it does not need to handle the load current, as the BMS does that, or the load is on a separate cable from the BMS while the BMS just monitors the pack and alerts the user or turns off the controller at LVC via the e-brake line(rather than having all the discharge current run thru the MOSFETs on the BMS)


Alternate suggestions or improvements to anything that's been posted is very welcome, and extreme detail is preferred to brevity.

Keep in mind that unless you leave an email address in your comment, I haven't any way to reply to you except to reply to your comment here. That means if you want a reply, you'll have to come back to *this* blog entry and it's comments to see my reply to you, unless you leave some method of contact within your comment.