Open ReVolt as Charger

[wjdennis]

NihaoMike, that is exactly how I charge my Li-ion cells each night. It is a constant current 30A until the first cell reaches 4.2V. Then the charger gets turned down to 2A and shunt regulators hold all cells from going over 4.2V. When every cell reaches 4.2V, I turn off the charger.

So for the first 80% of the charge cycle, it is constant current.

But back to my original question. While driving and having one pack supply current to the other, I'm not really talking about charging the cells. It's just that Pack A would supply a constant 70A to Pack B.

During flat driving, Pack B would be supplying 30A and Pack A 70A.

When going uphill, Pack A would continue to supply 70A, but Pack B would supply whatever additional current was needed, say 120A.

When going downhill, Pack A would continue to supply the 70A and add a little energy back into Pack B.

But neither pack would be "charging". Both packs would be using up their energy during the drive, with Pack A occasionally shuttling a little of its energy into Pack B during downhill driving.

dcb, if you're reading this thread, I'd like to hear more about why you think lithium is inefficient when using constant current as I've described. I've certainly never heard that before.

Thanks.

Bill

[MPaulHolmes]

Hello! Well, there's no reason the controller has to charge at a constant current if that's a problem. You can program any sort of algorithm you want with feedback and PI loops and who knows what!

[dcb]

Originally Posted by wjdennis

dcb, if you're reading this thread, I'd like to hear more about why you think lithium is inefficient when using constant current as I've described. I've certainly never heard that before.

Unknown battery x is assumed inefficient, it was not clear you were using lithium. Most of the DIY crowd here is still pretty low-tech on the batteries so that where where I am coming from.

A constant current is a fairly trivial circuit (basically a glorified voltage regulator) so using a complete motor controller would be a bit of overkill that still needs a fair bit of thinking over to make it into a charger. But there is a charge profile typically applied to lithium ion, i.e. if cell voltage is low, and to prevent runaway, so a processor (i.e. avr) is probably approproate to condition the applied charge (and discharge) to suit the needs of the specific batteries.

Of course there are still losses in conditioning the transfer to a "constant current", they can be reduced by using a switching instead of dissapating approach, though this will introduce ripple, if that is an issue.

What batteries #s are you using?

[wjdennis]

The pack that will be supplying the constant 70A comprises 34 200Ah ThunderSky LCP cells. The back that will be receiving the 70Ah comprises 40 60Ah ThunderSky LFP cells.

I'm happy to try to build something simpler than Paul's controller to handle this. Chargers on the market that handle 70A are expensive ($4000 or more). The ReVolt is inexpensive in comparison, and the design was already done. Could I build something for even less than that?

Thanks.

Bill

[jyanof]

Originally Posted by wjdennis

The pack that will be supplying the constant 70A comprises 34 200Ah ThunderSky LCP cells. The back that will be receiving the 70Ah comprises 40 60Ah ThunderSky LFP cells.

If I understand this right, the 1st smaller high power pack essentially acts as a very large capacitor to supply power during the variations in required driving power. The 2nd larger pack would essentially then provide an quasi-average power over the whole drive.

Something to consider is that the current configuration of the controller as a buck converter will only allow you to charge a battery that is of lower voltage than the supply battery. Current flows from high potential to low potential and the controller only makes the supply voltage lower.

I forget the vpc difference of LCP vs LFP, but you show different cell quantities. To use a buck type converter, your LCP pack must have a higher voltage than the LFP pack.

If this isn't the case, there are other more complicated options such as a boost-buck converter. This'll have a boost circuit that'll boost the voltage higher than the supply voltage. Then, you'd have a buck stage to reduce the high voltage to the desired range.

Another thing to consider is that most buck converters require inductance in the output loop to keep the current from rising too fast. For a motor controller like Paul's, the motor has inductance (though, there's the problem with large motors not having enough inductance and only certain controllers can handle them. From what I've read, a curtis 1231c will blow within minutes when connected to a warp 11).

Anyway, if a battery is your load, you'd likely have to add inductance in order to control the charging current. It might be tough to find large inductors capable of handling 70A, but maybe you can make your own from thick wire and some iron laminates.

It's an interesting setup to think about and I'm sure you could get it to work with some engineering.

[wjdennis]

Thanks for the comments, jyanof. You nailed it: the second pack is like a big capacitor bank.

Regarding the inductor: this topic actually started on another thread before I moved it here, and that's where I stated that I'd have to add an inductor in series with the batteries, as well as probably adding another diode in series on the positive lead, too, just to guard against the possibility of overvoltage on the input side.

The LCP cells are a nominal 3.6V. They charge up to 4.25V, which is 144V in my case. The LFP cells have a nominal voltage of 3.2V, and charge up to 3.65V per cell, which is 146V. So the two strings are about equal at full charge. The LCP cells settle down to about 4V/cell after charging, or 140V. The LFP cells settle at about 3.35V/cell after charging, or about 136V. Under load, I expect the LFP cells to sag about 8V at 50A. Based on the internal resistances of the two strings, I expect to need between a 15V and 20V difference in the two packs to get the full 70A. So when the current draw on the LFP cells goes to 2C sometimes, they should sag maybe 16V, and the full 70A will be applied. At lighter current draws, I may not be able to get the full 70A. I might need to reduce the number of LFP cells to 39 or 38.

Only real-world testing will tell.

Bill

[jyanof]

Some more thinking out loud about controller logic...

i guess it would still work of the two battery voltages are the same - you can program the micro to just limit the current from the large pack when it increases above 70A.

The controller would start out 100% on so that both battery packs are at equal voltage. (if they start out equal, there should be no current flow).

The two packs will supply their own respective currents as power is drawn (maybe proportionally to internal resistance?), but the voltage of the two packs will remain be equal as long as the current is less than 70A on the large pack.

The controller will limit large pack current if it increases above 70A forcing the small pack to supply the rest.

it seems like the small pack will drain faster until its resting state voltage is lower than the large pack - i guess it'll be charged by the large pack at this point?



it seems like an involved problem to work out in my head, but i guess it comes down to voltages. Keep the voltages on the respective packs within their respective limits and all is good. I can't say how performance would be... I think unless you have a significantly high voltage difference between the packs, you'll end up sharing load between the packs.

Now i'm thinking about the 70A figure... with my 144V pack, cruising at 45mph requires 50-70A and 65mph requires 80-100A. Lets say you need 70A from your battery system at cruise - if all 70A comes from your large pack, you won't be able to charge your small pack. If you need 100A at cruise, 70A will come from the large pack, and 30 from the small (assuming everything is setup to allow for this).

Are you intending to charge the small pack while cruising or just while standing still? or are you typically traveling at speeds that allow for it? I imagine your current draw is a little less, but its gotta be in that ball park...

anyways, just more thoughts...

[wjdennis]

jyanof,
Thanks for your insightful comments. At 65 mph, my car pulls about 100-110 amps. My commute is 50 minutes each way. The last 6 miles coming home at night entails a 1700-ft. climb (of course 1700 ft down in the morning). So I'm hoping that the smaller (60Ah) pack will supply about 30A on average and the larger (200Ah) pack will supply 70A.

Once at work, I'd like to turn the controller's current current limit down to about 20A and leave the 200Ah pack charging the smaller pack for about 1 hour, putting 20 Ah back into it. This will give the smaller pack some extra energy to supply for the 1700-ft climb at the end.

So when I finally reach home at night, the 200Ah pack should have supplied 160Ah, about 80% DOD. The 60Ah pack will have supplied 50A on the way home, so also be at about 80% DOD. This is assuming an hour commute each way. With a 50-minute commute, I'm hoping the DOD will be a little less, maybe 70%-75%.

Bill

[wjdennis]

Originally Posted by wjdennis

Black Panther wrote: The idea is that the second pack would be smaller capacity, but newer tech with a high C rating and the same nominal voltage. if preserved as a fresh pack until that hill, when you paralleled them, the second pack should supply the majority of the current simply by the balance of each packs capability to deliver the current. your low C pack will tend to sag as it approaches its c rating and when paralleled to a high C pack which maintains the voltage, the low C pack will not be able to push as much current because the paralleled voltage is higher than what it can push that much current in to.

Without the second pack connected, the voltage would drop more under load and your low C pack delivers more current at the expense of battery stress.

With the high C pack connected, the high C pack will not stress until much higher current loads, so it can keep the paralleled voltage at a respectable level until you have nearly depleted them.

I hope this clarifies it somewhat.

Black Panther,
I'm not saving the second pack until a hill. Cruising at highway speed requires 100A. I want the lower-C pack to deliver no more than 70A, ever. So while cruising, the higher-C pack will be delivering 30A, and the lower-C pack 70A through the controller. When the hill comes, the higher-C pack will deliver more amps, with the lower-C pack remaining at 70A through the controller.

Bill