Bonanza "Bulldozer" Dual PhaseRunner AWD

[Alan B]

The Saddle Block Clamps work well but do take a bit of space to hold on boxes and batteries. But they get a great grip on the tubes and don't deform them.

The circuit Boards have arrived, the parts arrived, and the parts even fit!!! Need to set up the Mill to cut these boards apart and build them up!

Also got some 3" by 1/8" thick fiberglass (probably too flexible for battery support board), and some 1/8" thick aluminum in 3" and 4" width (these are probably good for battery support to replace the red oak)!

I'll set two of the charger interface boards for charging, and the third I will set up for CellLogging. Strap the CellLog on there and I can check my pack cell voltages by plugging into the charging DB9! Very quick and easy!!

These boards also have a couple of mounting options for the PowerPoles. It is going to be much neater. And I can use some of that 3" wide fiberglass for under-board protection.

 

[torker]

Can't wait to see some pics of that setup. I am not running any lvc and just using the rc chargers to balance. OK for now but need to implement something better soon.

 

[Alan B]

I'm taking pics, just have not uploaded them yet.

Interesting you should mention Low Voltage Cutoff. I'm planning to use some of the SMD LVC boards from Geoff57 with my boards to provide the LVC function as well as the DB9 balance charging / checking connection. I'm not planning to use the CellLog for LVC, though that can be done. They draw a bit more power from the batteries than I like, and I don't have enough of them, plus they require optical isolation of their outputs. Jeremy mentioned some time back that occasionally the CellLogs get "hung", so they may not be the best choice for the most reliable LVC.

I cut apart the boards and installed the connectors. Everything fit.

 

[Alan B]

Keyswitch with 3 positions and key removal only at position #1 will handle up to 4 amps; and 50 amp Electric Vehicle fuse:

ON-ON-ON toggle switch that can be used for a Kill switch that will turn off when switched either direction away from the central ON position:

Vetrap by 3M as used here for holding batteries. Comes in 5 yard roll (that is stretched length). Comes in different widths and colors. I've been using 4" black, this roll is 3" blue. Available at large pet stores (for wrapping horse's leg injuries, etc). Sticks only to itself, does not leave a gooey mess on batteries, etc:

Boards before cutting with parts pre-fitted:

Boards during cutting on the Mill using carbide PCB rasp:

Boards made up. The three boards toward the right side are the battery interface boards (6S3P). These mount on the battery pack. The one on the left is a charger interface board. The JST-XH extension gets mounted to this board and interfaces this board to the charger. A DB-9 extension cable goes from there to the battery interface board on the bike during charging. A charging cable goes from the banana jacks on the charger to the PowerPole to carry the charging current. The DB-9 carries only balance current:

 

[torker]

That's awesome. So then you have another connector going on that carries the charge power to the batteries? I said I worked on electronics. I didn't say I was any good at it. hihi Alan will know what hihi means..

 

[Alan B]

That's awesome. So then you have another connector going on that carries the charge power to the batteries? I said I worked on electronics. I didn't say I was any good at it. hihi Alan will know what hihi means..

Yes. The PowerPole carries the 25V charge current to the 6S group, and the DB9 carries the balance signals. One PowerPole is visible in the photo above. All three boards will have that.

I made up the first charger interface board. Cut up a JST-XH extender cable and soldered it to the pads on the board, and clamped it down with tiewraps to the holes put there for that purpose. A Banana plug to PowerPole cable and a DB9 extension cable will be added to that to make up the charging cable system.

 

[torker]

Ok I get it. Will be good to see it hooked up in charging mode. Nice clean work btw.

 

[dbaker]

Very, clean boards Will you offer a group run for ES folks?

 

[Alan B]

Ok I get it. Will be good to see it hooked up in charging mode. Nice clean work btw.

Thanks.

Very, clean boards Will you offer a group run for ES folks?

Thanks. PC boards can make things quite nice. If there is interest we should be able to make them available.

 

[Alan B]

HobbyKing does it Again

Just unpacked and tested a box of Turnigy Lipo from HK USA. Four 6S 5AH bricks, and all are okay. 3.84/3.85V/cell. That's 14 good packs and zero bad ones. Spread over a number of orders.

And a big silver star to the CellLog. Excellent way to check a pack (though it would be nice if it would boot faster).

Fantastic. Now I can go ahead with upgrading to 3P as well as installing the new DB9 charging cabling. Unfortunately, I probably have the flu, so things are going very slowly. It is a beautiful idyllic afternoon here in the Bay Area. If I felt better I would go for a ride!

I went to make up my first DB9 charging cable, and ran out of 18 gauge red-black. So I will have to get more.

 

[Alan B]

I ran an end to end test on the new balance setup. Battery to battery interface board to DB9 cable to charger interface board to CellLog. It passed by correctly reading out the cell voltages (and not shorting any cells).

I got some more black/red wire for the charge current cable, perhaps today I can make up the charging cable.

 

[Ykick]

(though it would be nice if it would boot faster)

+1, 'got 2qty to brigade when checking all the bricks in my OVEN...

button battery could likely be fitted to maintain power? but probably less trouble/expense to just use 2 and it's nice to have a spare log8.

 

[Alan B]

(though it would be nice if it would boot faster)

+1, 'got 2qty to brigade when checking all the bricks in my OVEN...

button battery could likely be fitted to maintain power? but probably less trouble/expense to just use 2 and it's nice to have a spare log8.

That is a good tip, using two to alternate so while one is booting up you are busy hooking the other one up. Efficiency engineering.

Beautiful day here, slept through most of it. Starting to feel better but still not up to normal, whatever that is.

Was hoping the large screen Cycle Analyst would arrive today, but no dice.

 

[Alan B]

Spent a little time on this today and made some progress.

DB9/PP Battery Interface installed. I decided to have the charging cables on the curb side, opposite where the RJ's were pointed. Here you can see the older RJ45/PP board on the left and the newer DB9/PP board on the right. The DB9 board is mounted on red fiberglass. The mounting to the battery board is via tiewrap, the battery support board will be redone soon and a better mount will be made for the interface boards; this will do for now:

Moving in a bit closer, you can see the charge current PowerPole is held down with a loop of wire, and there is also provision for a small bolt through the middle of the connector. The charge current cable is also tiewrapped to holes in the PCB:

Now with charging cables installed, the DB9 carries balance currents and the PowerPole carries charging current up to 10 amps. More could be accommodated with heavier wire:

 

[Alan B]

Installed the other two DB9/PP adapters. Tested them with the CellLog. They seem to be working fine!

 

[Alan B]

Controller "OFF" Drain

So what is the drain of a 12 FET controller when turned off? I connected the new charging setup to the charger and the bike and let it run. I figured since I had not used the bike since last charging it that it would cycle off very quickly, but it would give a test of the wiring. But the first bank took 1.28 amp hours! The second took 1.27!! What's going on??

The only thing connected to the pack for the past week has been the controller's main power inlet. So this charges the capacitors, but doesn't run anything. Supposedly. But apparently there is some leakage there, to use 1.3 amp hours. Let's see. One week is 168 hours. 1.3 amp hours divided by 168 hours is 8 milliamps. Not much, but it ADDS UP. So make SURE to unplug the controller because that's more than 10% per week lost just sitting there. That means 8 weeks from full to dead!

Hmm. Makes me want to get that keyswitch system working soon!

At any rate, now charging group 3 with the new charging cable setup. Working nicely! Incidentally, it takes 33 minutes to balance charge 1.3 amp hours with the Turnigy Accucel 8150 charger at 6 amps into a 6S 10 amp hour pack. Charging beyond that should be predictable at the 6 amp rate. Should make a chart estimating time vs amp hours.

 

[Alan B]

Here is the new DB9/PP balance charging system on the battery pack, all three groups:

This is the charging station. The RV Converter puts out 13V at up to 30A to the Accucel 8150 charger. This connects to the DB9 extension cable via the charger interface board. The balance lines connect to the charger's JST-XH connector on the right. The charge current is carried through the red/black wire from the banana plugs on the left and then are tiewrapped to the interface board for mechanical support:

The other end of the charging cable plugs into the bike. Here mine is parked vertically in the garage getting fed:

Any questions?

 

[Alan B]

Incidentally, if anyone wants the RJ45 charging setup, drop me a PM, I will sell below cost (SOLD). Otherwise I'll probably rob it for parts. I have a few more of the JST-XH parallel boards, also. It seems to be working fine, I have not had any more problems with popping traces since the initial installation issue. I must have done something to it when I was installing it the first time.

It surprises me a little that the standby leakage of the controller is as high as it is. This one measurement is almost 2% per day. Better unplug those battery packs!!!

 

[Partycat]

I would think that to be a good idea anyways with the LiPo, not leaving the whole thing sitting there plugged in like that for any length of time. If a 2%/day capacity loss was enough to worry you about discharging it, then it is spending a bit of time unattended !

I'm fairly sick of Anderson PowerPoles at the moment. I've installed almost 50 sets of them on my bike at this point, and I wish I would have left the bullet connectors alone, they seem much easier. The TriCrimp tool broke in about 5 crimps, and trying to crimp that fine stranded wire in the 45A connectors is just a massive pain.

Also I neglected the orientation on a couple of the hall wires (ran out of coloured housings and went with red/black...) and ran the bicycle backwards into myself

 

[Alan B]

Welcome to ES, Partycat.

I have the older West Mountain Radio crimp tool. The Tri-Crimp came along later at a lower price point, perhaps there is a reason... I also have the SB-50 jaws for the WMR tool so I can crimp SB-50's and PowerPole 75's. It does seem that any connector we choose gets annoying after installing a few dozen of them. The 4mm bullets are soldered with no strain relief, all stress is on the solder. They allow plugging things together that are polarized backwards like two batteries plus to minus and minus to plus (BANG).

As you can see I am trying to minimize the number of connectors that I change on this build. So I stuck with the Turnigy 4mm stock bullets for the batteries, JST-XH for the balancing, etc. I'm trying to use each connector for an appropriate application. Making circuit boards is a lot easier than installing a bunch of connectors and making splices, etc. At least in my opinion.

In this case the PowerPoles are used for carrying the charge current and the DB9 for the balance current. They are easy to plug/unplug and are rated to carry these loads easily.

My main power from battery to controller is SB50 on this build, plus there are a couple of PowerPole 30's for lesser loads including precharge and lights (which are not installed yet).

I generally unplug the SB50, but did not last weekend, so gathered this bit of drain info. You are right in that it won't matter if the rig is being used every day or two. But patterns change and remembering to disconnect when the exception occurs is not easy. I may go ahead with a keyswitch and FET setup that brings the leakage waay down and does away with the manual precharge, or at least makes it easier to do.

 

[adrian_sm]

Any questions?

How quickly is it charging the batteries?

I noticed that changing the main RED/BLACK lines to shorter, heavier gauge wire made a huge improvement in how quickly it charges overall. It think the voltage drop was enough that it was artificially restricting the recharge rate, as the voltage the pack saw was lower than it could handle. Might be worth a quick test next time you charge.

- Adrian

 

[Alan B]

Hi Adrian,

My charger is reaching 6.0 amps (which is about all it can do on 6S), so the impedance of the charge lines is not an issue. The charger is the Turnigy Accucel 8150. So far that's the only charger I've tried.

I did change the wires from about 4 feet long to about 7 to 8 feet in this last upgrade, and they are 18 gauge red/black Radio Shack speaker wire. This reached 6.0 amps during my charging test.

Depending on the charger, the results might be different. Which charger, and what length of what gauge wire did you use?

 

[Alan B]

Keyswitch Considerations

I made a few more designs of keyswitch electronics. I don't like having to unplug the batteries, and I don't like 2% per day leakage. My last Mouser order included a keyswitch. There are so many it is difficult to choose one. I picked one that had three positions, 45 degree indexing, and the key is removable only at the counterclockwise most position. It has two poles, gold over silver contacts and 8 terminals. It is rated to 4 amps at 125 VAC. The manufacturer's number is JD-7509B.

The double pole nature of this switch and 8 terminals allows for a lot of possibilities for a circuit. Clearly the main current cannot pass through this switch, but the precharge current could. A 100 ohm precharge resistor limits the current to less than one amp.

So the circuit above takes a minimalist approach and runs the precharge through the switch. Two poles could be paralleled for this which increases the current capability some as well.

The switch is set up to use all three positions. Position 1 is the key insertion/removal position. Position 2 is Precharge. Position 3 is Operate.

In the Operate position power is applied to the gate of the FETs that carry the real current. These should be the same type as the ones in the controller, and the same or greater in number in parallel. I will probably use four in parallel in the final circuit because they are not that expensive and I don't want them to get warm and require heatsinking. I'd rather buy an extra FET or two than deal with heatsinks and heat.

Additionally in the Operate position power is applied to the auxiliary circuit and the controller power circuit. So these loads cannot operate during Precharge.

If this switch is operated quickly to cut the Precharge short the main FETs will come on and complete the Precharge. There will be no spark, but the current surge might be high. The turn-on of these FETs is not going to be very fast due to the gate resistor, so it may be pretty soft already. We'll have to do some measurements to see. In any case there can be no spark, so the inrush is not hard on the connectors. The capacitors are the only other concern and they generally do fine even when we hot-plug our controllers, so it is not likely a problem. The FET turn on can be modified by changing the gate resistors. Rather like starting a car, the way we operate the controls affects the results. Some testing and adjustment may be needed to finalize the values.

So this is pretty close to as simple as it gets. The next more complicated design would be to use a FET to carry the precharge current. Another even simpler possibility is to play with the FET turn-on parameters and drop the pre-charge resistor, using the main FETs for this by turning them on slowly enough. It is a little hard to predict that behavior, so either some modeling or some trial and error would be required. I'll have to look at that a bit more.

 

[Alan B]

Here is the keyswitch mentioned above and a fuse designed for Electric Vehicle use with an 80 volt rating. They are available at various currents, this one on the light end at 50 amps. According to their data sheet it takes about double the current to trip these, so this should trip at about 100 amps:

 

[Alan B]

Just got word back from ebikes.ca, Large Screen Cycle Analyst production is having some problems (even though the website said IN STOCK), so shipping is delayed. Oh well, keep using the old one for now...