A better alternator delete
 

I think alternator deletes are a really great mod. They give you a sizable gain for just one modification. I've had them on just about all my daily drivers for the last decade. However, they just are not usually cheap. First off you need a deep cycle battery. If your starting battery is fine, its hard to swallow $80 for a new battery. If its dead, its not too big of a deal. Then, you need a charger for this battery. Let’s say this charger costs you $50 (I'm pretty sure mine cost more). Now you're up to $130 for a single modification. Next, you need to actually do the mod which can range from inserting a switch in some wiring to coming up with a small electrical circuit like I've had to do on my Mirage. Lets say this costs another $10 for a total of $140 for an alternator delete. Yay, you get a nice bump in MPG, but that $140 can get you a lot of gas too, right? What if there were another way?

You've probably already seen my post on battery hookup. If you haven't, I highly suggest it because it seems like a great place to get some lightly used lithium batteries. They're keeping these things out of landfills and getting us DIYers a great source for lithium batteries. With the prices they have I've been thinking more and more about what I can do to improve mpg via battery use without spending a small fortune.

My idea for today is a portable lithium battery pack that essentially doubles as an alternator delete. What I'm talking about is a smallish lithium ion battery pack that is 15V+. This pack has a voltage regulator on it that sends the 15V+ to your car's 12V battery via a PWM signal. The power out of the portable pack will keep the car's system voltage just above the voltage set by the alternator, thus disabling it. This is the same idea I used on my Mirage to disable the alternator and it works great. When the lithium pack is depleted, your car goes back to using the lead acid starting battery as normal.

I'm pretty sure we've talked about this before... kind of. A lot of you apartment dwellers don't have access to outdoor charging. This portable pack solves that problem. Take it in at night and you can charge it in the warmth of your building. This also solves the problem of some lithium cells not being able to be charged at freezing temperatures. When you go out to the car in the morning, simply plug the pack into your cigarette lighter (or other plug of your choice) and you're ready to go.

Another benefit of this is that your car's system voltage is actually where it is supposed to be, not at 12.X volts. This means no more dim headlights at night, or slow wipers. Also, no hacking car wiring. This thing literally can plug right in, and come right out.

On top of all that I also think this is doable for less than the 'old' way of doing an alternator delete. Here is my guestimates for cost:

lithium pack - $30 (from battery hookup)
5S bms - $13 (price from amazon)
5S charger $17 (price from amazon)
voltage regulator - $15 (DIY circuit? rough guess)

Total: $75


So, roughly half the price. If you use a cigarette lighter plug you can literally use this thing on any car you jump into.

I like this idea!



I'd be surprised if nobody in the history of EcoModder has ever done this before (I once thought about doing it with a DC-DC converter powered by 24v).



But the difference today is the availability of cheap used cells and cheap electronics to make it work - almost plug 'n' play.



Planning to try it out?

Well, it was discussed in the Virtual Alternator Delete thread.

I planned to give it a try, then I had a kid, bought a house, still renting another house... also I hardly ever drive anything other than the company car now that they provide unlimited fuel for it.

In 2 years when I have to turn it in I'll re-evaluate the feasibility of the Runzheimer program where I would get compensated some amount for using a personal vehicle for business.

Aha, I'll have to find that one. I must have missed it. :thumbup:

Ah, "virtual" was misspelled in the title. Here it is.

https://ecomodder.com/forum/showthre...ete-36517.html

I think "the better alternator delete" the alt stays in the car unbelted and electrically isolated and can be quickly returned to service if needed.

This does seem like a great idea, but for couple possible concerns:

- Charging a lead acid battery is less efficient than lithium. IIRC only about 80% of the charge to a lead acid is actual charge, and that number lessens when reaching the float stage. And if you're not floating a lead acid, it's life is going to be shortened.

- Do you know that the 5S Amazon BMS is a good one? There are a lot of cheap BMS out there that fail on short notice. And then since you're relying on the BMS to manage the battery, if it goes kerplunk, you likely won't know of any battery issues until it's too late. Obviously LiFePO4 is much more robust but it won't be good for the life of the battery. From what I've seen, a good 4S-5S BMS is anywhere from $40-$70. Then again if it were to fail, it would be much cheaper to replace the battery than to spend $ on a higher quality BMS.

My thoughts are:

Why not just replace the lead acid with the lithium, and use a switch for the alternator like you've been doing? There would be fewer charging inefficiencies due to going from battery to battery, and you'd also save some weight by eliminating the lead acid battery altogether.

....but that's just me being an extremist ecomodder, where if I do an ecomod I have to go all the way for efficiency and leave nothing on the table (80 PSI tires.. :rolleyes:). Other than that it seems like a great idea. Maybe replace the normal 12V starter battery with a small 12V lithium battery in addition to this? but then most small lithium batteries don't have the cranking amps available.

This would definitely be a much simpler way of doing it, although I surmise that the efficiency gained would be slightly less than doing a direct lithium replacement and alternator cutoff. But since we're talking about, what, a 5%-10% gain from deleting the alternator, the difference between these methods is likely <1%.

Here is a 32.95V 2.9ah pack with a built in BMS for $25. Considering 90% usable capacity @ 14.4V gives around 6ah of capacity. Two of them brings you to $50 for 12ah@14.4V and you don't have to worry about rigging up your own BMS. A cheap DC-DC converter to trickle charge your 12V battery from the larger pack.

I'm not too sure about a PWM voltage regulator. I really don't know anything about them other than that I've read that PWM solar charge controller's lower the voltage but don't up the amps (i.e. a 18V/2.75A/50W solar panel through a PWM charger puts out 2.75A@14V (38.5W), which wastes over 20% of the energy as heat. Now I have no basis for assuming that other PWM voltage regulators work like that, but that's just what I've read. A small, 100W DC-DC converter isn't that expensive and usually much more robust.

My idea which I posted in the other thread is to have 2 battery harnesses in parallel in the cabin. I take a freshly charged lithium battery and plug it into the harness, and remove the previously used one to charge it indoors. Each time I take a trip, I'm starting out with a freshly charged battery, and not subjecting it to cold temperatures while it gets recharged. When the battery drains to a certain voltage, the alternator kicks on and maintains voltage.

No brain power required with this, as the only thing to remember is the freshly charged battery each time.

I bought a 5aH battery to experiment with... this thread might push me over to the point of testing. Maybe I'll work on the wiring tonight and connect a harness to the Acura.

The other thread:

Virtual alternator delete.

I decided to give this thing a shot. I like Redpoint's idea of using a lithium battery in parallel, but I would like to give this idea a go. I have somewhat changed the design a bit though.

https://ecomodder.com/forum/attachme...1&d=1572548797

First off, the voltage regulator will be a 300W dc-dc buck converter that I found on Amazon for $14. This will take the 17-21V from the lithium pack and convert it down to right around 14.5V for charging the lead acid battery. The voltage will be set to right above what the alternator charges the battery to.

https://ecomodder.com/forum/attachme...1&d=1572548845

I am also going to ditch the BMS in favor of individual cell chargers. I was able to purchase 5 12V 3A power supplies for $3 from bgmicro.com. These power supplies will put out 12V to a 5A dc-dc buck converter that is set to whatever cell voltage I want, probably 4.0-4.1V. This will balance each cell every time it is plugged in to charge. As long as it isn't discharged too far, it shouldn't be a problem.



The last piece major of the puzzle is the batteries. I still haven't figured this out yet. Battery Hookup is out of the cells I really wanted to use, and they are also out of stock on the second choice as well. These were mainly my choices as their prices were very low. As I piece things together I'll keep an eye on what they have. If all else fails, I'll just build a pack from 18650 cells.

All of those conversions can't be efficient. :/

Guess it's not that important if you're charging it with grid power.

My guess is each conversion is probably about 80% efficient. If you guys have suggestions I'm all ears. This just seemed like a easy and inexpensive way to get done what I am attempting. I'm no electronics expert.

That is what is great about Redpoints idea, it's very simple and has less components. The only issue I see with it is that it uses a very small percentage of the capacity of the lithium battery, and the battery is the largest cost. This is my attempt to reduce the cost.

I see adding a lithium in parallel as doing nothing but good. While it's plugged in the 12v lead acid will cause a constant small drain on it, turning some number of watts into heat. If you try this I'd be highly interested in what kind of current you see coming out of the lithium to float the lead acid battery at ~14.5v.

I disconnected the Pb battery when I attached the Li-ion battery because I could hear electrolyte decomposing from the higher voltage.

Yeah, the lead acid battery can not handle a 16.8V LiPo cell connected in parallel. We could eliminate the lead acid battery as Redpoint did in his test. That makes it nice and simple, swap out the lead acid for a LiPo.

There are some great benefits to be had with going this route. First off its really simple. You shouldn't need a BMS. The alternator will not overcharge these cells. Also, the charging efficiency of the LiPos is far superior to the lead acid battery being charged near max SOC. Lead acid batteries are said to only be ~50% effective near full charge.

Light on the Earth: Online Solar Manual for Homeowners

On the other hand, LiPos are said to be 99% efficient! This is a huge efficiency boost once the alternator is in use. Of course we also have the weight benefit of going from a 40-50 lb battery to half to a quarter of that.

And, there is also one big downside to this method as well, and that is usable battery capacity. I'm not too crazy about running my car's electronics at much over 15V. I know lightbulb life is negatively effected at higher voltages. I'm not sure what else is though.


Lets look at an example. Lets say we charge the pack up to 15.5V, and the alternator will start charging once the voltage gets below 14.5V. I grabbed this discharge chart from the Sonata 5300 cells that I wanted to use for this project. The two red lines represent a 4S 15.5V charge, and 14.5V. The colored verticle lines show us our usable capacity at their respective amperage draw. This shows us that the usable capacity of these 5.3Ah cells is only about 1.2-1.3Ah. That means we'll actually need a very large pack to get any reasonable usable capacity out of it.

https://ecomodder.com/forum/attachme...1&d=1572627350

I think the ultimate solution is basically what mpg_numbers_buy said. You have a very small 4S LiPo pack made of power cells (cells designed to put out a lot of power) that can start your car easily. It could be quite small and light. Then, you have another pack made of energy cells (cells designed for lower power output, but usually higher cycle life) that can charge the 4S LiPo pack through a dc-dc converter. This pack would be grid charged. It would be sized as large as you need to give you the required alternator off time. Once this pack runs out, the alternator kicks in. However, (of course there is a downside) it has the complication of BOTH systems.

If you do the balancing yourself, it really isn't all that complicated.

With a lithium battery, especially a homemade one, you will want to relocate your 12V to the cabin. Leave the DC-DC connected to the 12V, and make connections going to the larger battery that you can easily remove.

The complication would involve installing it neatly, or if you wanted to engage regenerative braking.

Excellent points! I agree.

There is still the problem for most of us and charging lithium below freezing temperatures. This is why I am going to keep the lead acid 12V, and do a lithium jumper pack. The lead acid car stay in the car and handle the cold and the lithium can be brought in to stay warm.

Interesting suggestion... if you're providing the boost in voltage via buck converter, you could wire the brake switch to disconnect it so the alternator sees battery voltage and turns back on. You'd get regen back into the battery whenever you tap the brakes. The only downside is the alternator would stay on while you're stopped with your foot on the brake.

If instead you could signal the buck converter to disable whenever the car goes open loop, that would be even better. It would go open loop when DFCO and charge the battery, and then re-enable the buck converter once it goes closed loop at a stop.

Hi,
It may sound primitive but how about adding a secondary battery just for the high energy consuming loads like headlights, cabin fan et cetera?and then charge it at home?it will need some extra wiring and a battery.
Has anybody done that?

You've just suggested what this whole discussion is about.

Well, I believe what he means is, breaking the electrical system into separate partitions.

In that case, it mostly defeats the purpose, which was to get the alternator to turn off. I don't see the benefit of creating islands of power.

I thought it was to eliminate the power rob the alternator produces under load and still drive the car under sub standard conditions

I got the lihium battery charging dc-dc converters over the weekend. I plugged them into a 12V power supply and tweaked the output to be 4V and 4A output. I initially had it set to 5A, but it seemed like it was getting a bit warm for my liking. Granted, I was pumping 5A through alligator clips haha. That might have had something to do with it. I'll revisit that later when things are more finalized.

I am hoping to get the AC to DC power supplies today, and then I can test the charging side of this thing.

I got the ac-dc power supplies tonight. They're larger than I thought they'd be at about 5.25" long. However they are very light weight. I'm thinking I may take them out of the cases when I get to putting everything together. It will be better for cooling and size.

https://ecomodder.com/forum/attachme...1&d=1573178617



Here is the specs on it. It actually puts out about 12.3V which is fine for this purpose.

https://ecomodder.com/forum/attachme...1&d=1573178669



So, I hooked up one of the power supplies to the dc-dc converter. Then, I hooked the dc-dc converter up to the multimeter set to measure amperage and let it run for a while.

https://ecomodder.com/forum/attachme...1&d=1573178823



I wanted to see how hot the dc-dc converter would get just running at 4A for a while. It turns out, it runs pretty darn hot. I'm definitely sure I don't want to run it at 5A. The diode (the thing just to the right of the inductor coil) gets very hot. I measured the best I could with my IR temp gun and the highest reading I got was 216F after a few minutes running. The chip says SK54 on it, so I googled that. The first data sheet I found says max junction temperature is 300F. The capacitors were also pretty darn warm at ~145. A quick google search says electrolytic capacitors are good to about 180F. So, I guess I'm okay? Electronics guys feel free to jump in with some feedback.

https://ecomodder.com/forum/attachme...1&d=1573179068

I've been running a set up like this for a couple of years on a three of my cars. On my 'best' set up, solar panels charge my OEM battery to 100% float every morning. The alternator sees the full battery and doesn't charge other than regen.

For the first ~6 miles the alternator just sits at 11.7 volts (no charge). After that (and several engine restarts) the ECU sees the battery as getting slightly low and starts to put a slight charge in at 12.4v. It then cycles lightly depending on how much power the solar panels produce. On a sunny day in the open it's possible to almost drive entirely alt. less.

In two years I've saved ~140k alternator Wh. That's around 11 gallons over ~12k miles (around 3%). However you trick the alt into not charging should give similar results (assuming your method has enough capacity to reduce your alternator duty cycle to around 20%). All the same, the difference is too small to pick up in tank-to-tank fill ups. 1.5% gains would likely be more typical if you can't charge at your destination.

At it's simplest, I just keep my Fiat battery on a garage mounted solar charger, that keeps the alternator off for the first few miles just the same, but I don't have kWh numbers for that one. Sure is a simple 'mod' though :thumbup:

Check if the output GND is connected directly to the input GND.

If so you cannot connect the drivers in series to charge your cells without shorting them.

The cooler you keep the components the longer they will last. Especially the capacitors, they must be pretty crappy to get as hot as you measured.

Yay, Battery Hookup had a deal on 18650 cells. So, I got a case of 120. That will be more than enough for this project. I was thinking I would use 40 cells for this pack.

The cells are in packs of three. I'll have to disassemble them and figure out how I want to reassemble them once they are here.

https://ecomodder.com/forum/attachme...1&d=1573242078

Pretty good deal! Guessing this is for a higher voltage booster pack, 4S30P to charge a standard 12V?

I am not sure exactly what configuration they will end up in. I do not think I will need the whole 120 cells for this project. The first pack I will be making is probably going to use about 40 cells. 5S8P is what I am thinking. I'll use a 300W dc-dc converter to charge the lead acid battery at just over what the alternator puts out.

The next step now is figuring out how to charge them. I have my doubts that my power supplies are isolated as Teoman mentioned. So I may have to come up with something else.

Check them for sure, but typically off-line supplies (AC to DC) are required by law to have isolated outputs, so you should be able to charge individual cells in the stack.

Thats good news. I'll have to crack one open and double check. The cases seem mostly empty. I imagine they could have been much smaller.

It looks like it is isolated to me. The big yellow taped up thingy's part number is a transformer when you google it.

https://ecomodder.com/forum/attachme...1&d=1573271163

Those connectors look like laptop connectors. The laptop batteries have internal temp semsors and balance circuits...

If those are indeed the packs you have why not try to source the other half of the connector?

If i were you, i would try to source a connector and 3d print something to make the cells pluggable.

Wish they were in 4S format.


They communicate over I2C (if they are similar to laptop packs).

The chip that manages them was made by Texas Instruments (last time I had a look) and was able to pump out all kinds of usefull data like Wh remaining.

An arduino project would be an interesting endeavor.



OR you could say enough with high tech cr*p!

In that case, I would form the cells as parallel chains that are balanced capacity wise. (There was a website for that that jhurgia recommended).

Find a good connector that connects the strings in series. This connector would also double as your charging port. So depending on what you connect to this connector you can get 4S output or all cells parallel output. Then you can have single power supply charging without the need for balancing.

I Hope I was able to make myself clear.

If the yellow transformer is across the AC plug, they are isolated, that is why the transformer is there. It's cheaper to build a buck power supply that is transformer less but not isolated. The diodes are getting hot because you are running max current the device was designed for, ditto on the capacitors because they are supplying current during the "off" portion of the cycle. I wouldn't run more than half the rated current without having heat sinks if you want any longevity or "surprise" oops events.

I would also be inclined to use the modules as is with the bms on top if they allowed charge to 4.2 @ cell. Which gives you 12.6v

I'm interested to see what creative methods you end up using, as well as feedback on those cells (they're a good price, so if they live up to what they are sold as, I'd want to pick up some).

You'll need more than 20A to keep everything running, unless you've modified the snot out of your car to minimize the load. (I needed 30a to keep my car running without putting any extra loads, like the wipers, or the parking lights). Of course if this isn't your intention, then you can ignore this.

I really ought to get back on to this (similar) project.

They are modem batteries I believe. Printing something up so I don't have to disassemble them sounds like a great idea. I'm not sure I can make that work, but I will definitely look into it.

I also tested for continuity and there was none.

They are rated for 5A, and I have them running at 4A. I figured 80% wasn't too much to ask, but perhaps it is. 3A would still be fine for these purposes I think.

If that is possible, I could look into using a boost converter.

I think it is funny you are attempting to murder your alternator and MetroMPG is trying to bring his back from the dead.