diy solar

diy solar

[Project] Extending EV range by offsetting auxiliary 12v loads using solar

Just read a bit online on the differences between continuous and intermittent relays.
In fact, most sources talk about solenoids, not relays, regarding this difference.

Anyways, although this is a high (100/200A) current relay, I would doubt it is really suitable for a car's starter.
I think car starters draw a much higher current than just 100 or 200 amps.
This is judged by the cold-cranking-amps (CCA) rating of some car batteries. These are usually a few hundred amps, not 100.
So I am hoping that this relay is just a high current continuous one.
But I will be checking this just to make sure.
I will try to power it without load for a few hours, and inspect it to see if it gets hot (or burns altogether). I will try this with a small lifepo4 5Ah battery, and will attach a 30 amp circuit breaker just in case the relay burns and shorts out.

Thanks for the info.
Better be safe than sorry...
 
You can shop on Grangers or McMaster-Carr to get an idea of specs. Grangers is often too expensive.
I buy most things through eBay.

DC relays can have something called an "economizer", which allows high current to close the relay then switches to low current to hold.
A light bulb in series with the coil might do that - it is 1/10th the resistance initially, then increases in resistance as it heats up.
A PTC thermistor would behave similarly.
The 12V out from this relay could open a low-power relay that had been shorting a resistor in the coil circuit.

One that I use (100A 120VAC switching, 48V coil) has two coils in series, and an extra contact to short out the coil.

It is possible this relay is OK. 4W isn't a huge amount.
 
thanks for the updates! good work on the centralized integration board ?

re: intermittent vs continuous duty relays:

does relay heat up when activated for more than five minutes?

could use the shunt to measure relay coil power use?

no matter what material it’s constructed of, if the outside case does not heat up noticeably after half an hour of being engaged, then even if it’s ostensibly rated for intermittent use, i personally would probably experiment with it for tens of minutes or hours at a time with a cautious eye and maybe even a temperature sensor on the relay for over temperature safety disconnect

this mcp9808 has served me well, for an unsolicited inexpensive i2c sensor recommendation if needed :) the temperature measurement is 13-bit internally according to data sheet.

this project is really cool imho!

intelligent system to augment vehicle power source with VIPV! turning sci-fi into sci-re ?
 
DC relays can have something called an "economizer", which allows high current to close the relay then switches to low current to hold.
the gigavac mx12 relay/contactor specified for 200A continuous duty, yet only has 1W and 1.4W continuous power use to stay closed. the two models with that <2W power figure indicate a control circuit to achieve the lower use. datasheet
1637090749462.jpeg
that specific model’s amp rating is overkill too, but i just think it’s neat! that 200A switched with ~1W :) maybe a diy solution will emerge someday to do this for less cost
A PTC thermistor would behave similarly.
The 12V out from this relay could open a low-power relay that had been shorting a resistor in the coil circuit.
oo neat, taking notes.. take relay, bench dc supply, slowly ramp up voltage to coil. starting from 0V, going up. every 0.5V note whether the relay has closed. maybe temperature of relay case too? once it’s closed, go up maybe 1V and call that the Activation Threshold Voltage?

then begin reducing the voltage 0.5V at a time to find the voltage that the relay opens again when the solenoid magnetic field becomes too weak. Drop Out Voltage?

finding a PTC that would start with the Activation Threshold Voltage, and then after 10-60 seconds change resistance to cause the relay coil to receive slightly above the Drop Out Voltage.. i’m not that good with ohms law to do that right now ?

one last thing to look for when testing, if the relay begins heating up a lot when at the lower voltage, maybe the lower voltage needs to be increased to press the internal conductor harder to reduce resistance.

randomly thinking “out loud” so to speak?
One that I use (100A 120VAC switching, 48V coil) has two coils in series, and an extra contact to short out the coil.
interesting! there’s so many different circuit topologies to handle the solenoid control, thank you for sharing
It is possible this relay is OK. 4W isn't a huge amount.
indeed; the project does technically specify an intermittent use case (only closed during certain, changing, vehicle/environment conditions)

love this project!☀️???
 
Congratulations, you've now engineered an optimizer! Based on a data set of one sample, under ambient conditions.

Repeat data gathering with a temperature chamber, and multiple samples from different manufacturing lots.
Perform statistical analysis and design based on 3-sigma or other desired limits.
Perform FMEA.
Don't forget that 100A load on battery might cause coil voltage to drop, causing it to open, at which point voltage rebounds ...

:)
 
Thank you both for the informative replies!
One more thing to take into consideration is that a relay in a vehicle (as opposed to a static environment) would have to deal with vibrations. The conductor would have to be pretty stable and withstand vibrations and some jolts now and then.
So it should be "pressed" hard enough by the voltage to handle this.

I have started testing the relay...
It seems to start at almost half an amp, and slowly goes down to around 300ma after a few minutes. I'm not sure this relay has an optimizer though. I took you advice and connected it through a shunt.

I'll post my result soon. I want to test it for at least 20 minutes to see how things go.

For now I can say that after about 5 minutes there is a spot on the relays casing that is getting warm, but doesn't feel exaggerated.
 
OK, bad news...
After 50 minutes the relay started heating up pretty bad.
The current draw went down though.
Here is how it started:



20211116_225140.jpg
Then after 4 minutes:

20211116_225525.jpg

and after 51 minutes:
20211116_234257.jpg

And here are some thermal images of the relay itself after 51 minutes:
img_thermal_1637099120761.jpg
img_thermal_1637099128033.jpg

This is worrying but also interesting at the same time. I say this because I found some very similar high-current relays on Amazon, and they were all continuous ones.

So I wonder if this is normal behavior or maybe this is really just an intermittent relay.
 
It looks like the relay contacts have high contact resistance so it heats up, current handling rating is probably faked. Check the Vdrops between the contacts of the relay.
 
It seems to start at almost half an amp, and slowly goes down to around 300ma after a few minutes. I'm not sure this relay has an optimizer though. I took you advice and connected it through a shunt.

Measure voltage at the relay too, and calculate resistance.
Temperature coefficient of resistance for copper is 0.4%/degree C, so would have to rise 167 degrees C to explain drop from 500 mA to 300 mA just due to temperature change.

Oh, looks like you do have voltage and current measurements (is the voltage across the relay, or other wiring too?)
Try the math on temperature of copper coil, see what you get.

Nice you're got a thermal imager. 88 degrees C wouldn't necessarily be unacceptable for the windings, but higher than 60 degrees is probably too much for plastic case. (would like touch-safe temperatures.)

It looks like the relay contacts have high contact resistance so it heats up, current handling rating is probably faked. Check the Vdrops between the contacts of the relay.

Was any current going through contacts in this test?
 
Measure voltage at the relay too, and calculate resistance.
Temperature coefficient of resistance for copper is 0.4%/degree C, so would have to rise 167 degrees C to explain drop from 500 mA to 300 mA just due to temperature change.

Oh, looks like you do have voltage and current measurements (is the voltage across the relay, or other wiring too?)
Try the math on temperature of copper coil, see what you get.

Nice you're got a thermal imager. 88 degrees C wouldn't necessarily be unacceptable for the windings, but higher than 60 degrees is probably too much for plastic case. (would like touch-safe temperatures.)



Was any current going through contacts in this test?
No, this test was without any load.
I just wanted to see if this is a continuous relay or not.
The voltage displayed is the voltage going to the relay.

The thermal camera I have is a small device you connect to you phone. it's a UNI-T 120 Mobile.
A nice little device that less expensive than a full blown thermal camera. I'm still learning how to use it.
 
Last edited:
The pictures show two heat spots @ 88.4c (temperature rise of 63c) on one side of the relay, the relay only has one coil so it surprised me that it shows two spots of heating and one large spot at 72.4c on another side of the relay.
BTW, I need to look into the UNI-T 120 Mobile, you mention.
Is this the one you have? Thanks.
 
Last edited:
OK, so now I'm even more confused:
I just did a similar test to a small 40A relay. Again, without load, just to see if it heats up when on for a while.

After only 20 minutes or so this little thing was also very hot.
So either this is also an intermittent relay (which would be very strange) or my testing is somehow flawed.

The current draw on this little on is of course alot less.

Here are the results:
Starting current
20211117_000115.jpg
After 24 minutes:
20211117_002546.jpg
img_thermal_1637101616383.jpg20211117_000131.jpg


Maybe these relays shouldn't be run without loads?
 
The pictures show two heat spots @ 88.4c (temperature rise of 63c) on one side of the relay, the relay only has one coil so it surprised me that it shows two spots of heating and one large spot at 72.4c on another side of the relay.
BTW, I need to look into the UNI-T 120 Mobile, you mention.
Is this the one you have? Thanks.
Yup, that's the one.
 
OK, so now I'm even more confused:
I just did a similar test to a small 40A relay. Again, without load, just to see if it heats up when on for a while.

After only 20 minutes or so this little thing was also very hot.
So either this is also an intermittent relay (which would be very strange) or my testing is somehow flawed.

The current draw on this little on is of course alot less.




Maybe these relays shouldn't be run without loads?


Powering a load won't make it any cooler.
Connections to all terminals, and heavier wire coming from the terminals, would serve as heatsink and make it cooler. Try that.
 
It's been a while...
I've been busy, but finally all the parts I needed arrived, including the Lifepo4 battery!
You can read my little review about the battery here:

In short, the capacity is less than I expected, but otherwise it's a nice battery.

I finally got to wiring the DC-DC converter box (which will sit in my glovebox as mentioned).
It's not as elegant as I wanted it to be, but overall I'm happy with it.
Of course I made some mistakes along they way, causing me to throw a away a few lugs/terminal connectors.
Here are some pics:
20211201_224233.jpg20211201_230029.jpg20211201_230041.jpg20211201_230048.jpg

I added the temperature sensor of the Juntek shunt just before closing. I thought it would be nice to monitor the temperature inside the box. I mounted it close to the heatsink of the DC-DC converter, but not directly on it.
I also drilled some holes on one side, and larger hole on the other side (where the fan is) to allow for airflow when the fan will kick in.
The yellow lead sticking out is the relay trigger lead, which will be connected to the car's ignition sense/switch (just like a dashcam is wired to turn on).

I ended up getting a 80A relay that seems to be appropriate for continuous use.
It has some nice large contacts, and thick wires on the relay socket (unlike the 40A relay I previously ordered and tested).

I tested it for a while. Seems to get warm, but not too hot. It consumes less than 200ma during use, so that's nice too.
It's this one:

On each end of the box is an XT60 connector, male on the output (to car) side, female on the input (from battery) side.

I played around with this setup. The Juntek shunt works nicely, sorry I didn't take pictures of the display during testing. All seems to work well, the DC-DC converter's fan kicked in only when pumping out around 12 amps.
I calibrated the DC-DC parameters like this:

Under voltage protection: 11.6v (BMS of the battery cuts off at around 11v)

Over current protection: Around 10 amps.
This is less than I initially planned (15 amps), but seems like a better and safer setup. I am mostly using 12AWG wire, but the fuse piggy back that will connect to the car's fusebox is only 16AWG.
In any case, at 10 amps things should run alot cooler with less heat generation. Also, I wanted to keep the C rate of the battery well under 1C (20A), so 10 amps at the out would mean a few more amps at the input (depending on battery voltage).
In short, this setup will offset around 150 watts, instead of the 200w I planned.
I might change this in the future, if I see that all goes well and not too much heat is generated.

Output voltage: 14.5v
I think this should be enough to overcome the car's onboard DC-DC converter which outputs around 14.4v
I will probably have to play with too when the booster box will be connected to the car.

So... what next?
  1. Connecting the booster box to the car.
  2. Testing and more calibration using just the Lifepo4 battery (without solar).
  3. Mounting MPPT SCC, second Juntek shunt, mini fuse box and Lifepo4 battery in this little plastic ammo box:20211126_114410.jpg
  4. Mounting and connecting solar panels inside the car. It's been cloudy lately, so not much use of solar for now.
  5. Finalizing all connections inside the car and final testing.

As you can see, there's lots of work to do. I didn't expect it to be this much, but hey, it's mostly fun!

More updates to come...
 
As suggested on this thread and in other EV forums, I was convinced to swap the 12v SLA battery in my EV for a 12v Lifepo4.

The advantages are clear: less weight and flat voltage curve should help my EV be more efficient.

I thought of doing this as a second stage of my ongoing project, but as I found a fairly cheap 12v automotive Lifepo4, I decided to go along with it in parallel to my efforts of the buffer battery setup.

So about a month ago I ordered this one:
Costing about $235 shipping included.

It's a 12v 45Ah Lifepo4. In theory it has the ability to directly replace a car's 12v SLA starter battery. I say in theory because although it's rated for hundreds of CCAs (cold cranking amps) which is needed to start an internal combustion engine, I wouldn't count on a high lifetime in this scenario. I've seen teardowns on YouTube of better known Lifepo4's that claim to be suitable as a starter battery (such as the LiLead 110). They really aren't much different than your solar oriented deep cycle Lifepo4's.
The only difference maybe in the BMS programming, which will tolerate a very high current surge for a few seconds.
The LiLead for example uses standard EVE prismatic cells. These are rated at standard 1-2C rates discharge. There are some more professional models that use cylindrical Lifepo4 cells, or even better high C rated headway cells. But these are majorly expensive (around $1k).

Anyways...
An EV 12v battery doesn't need alot of cranking amps since there is no cranking going on. It's just needed to power the contactors of the HV battery and from then on the onboard DC-DC converter supplies all the 12v power (in addition to charging up the 12v).
I got this specific model and not a standard deep cycle Lifepo4 (which costs roughly the same) because this one has a wider temperature range, which I thought would be better suited for a battery under the hood of the car (and not indoors). The seller claims that they use Lifepo4 pouch cells for these types of "starter" batteries (which personally I'm not too fond of), instead of prismatic which they use in their energy storage models.

So... I finally received it!
Of course the first thing I did was to charge it up and do a proper capacity test. Just out of curiosity and to test how it works before installing in my EV.
Here was a disappointment: It tested out only to 20Ah instead of 45Ah!

I did this charge-discharge test twice @0.2C rate but got similar results both times:
20211206_061740.jpg

The second time I charged it up, I connected a shunt to see how many Ah went in up to 14.6v (as recommended) and sure enough it was only 20Ah.

Capacity is not so critical in an EV because it's use is very little as mentioned. But I was hoping that I would get closer to the advertised capacity, which would allow me to run some auxiliary loads straight from the 12v system for a while before draining too much.

In any case I contacted the seller and opened a dispute on Aliexpress.
The seller claims this is OK because starter batteries are "rated differently". I personally think it's B.S. he scammed me with false advertising stating its a 45Ah battery.
We will see how this dispute goes (I requested a refund of $100).

Capacity aside this little battery is actually pretty cool! It's got a nice look to it, with rubbery casing (doesn't look/feel like cheap ABS plastic) and properly sized terminals for car batteries (positive larger than negative so that you won't mix them up).
It's also so lightweight! Only 3.4kg !
I don't know if 3.4kg should have hinted to me something about the accuracy of the 45Ah claimed capacity or not. Does this sound logical in terms of Lifepo4 energy density?

I went ahead and removed the 12v SLA. Darn that battery is heavy!
It's a 60Ah SLA, just like you get in a fuel burning ICE vehicle. It's amazing how manufacturers try to reduce costs by putting off-the shelf ICE batteries in EVs even when they aren't really suitable for an EVs needs.
20211206_193257.jpg20211206_193305.jpg20211206_193350.jpg
I used a suitcase style hand held scale and it came in at 16.5kg !
That's a weight saving of 13kg in favor of the Lifepo4!

It was a pretty simple swap once I found the correct bolts to open in the battery tray.
The Lifepo4 is a bit smaller than the SLA so it has a bit of room to move in the tray (sorry I didn't take pictures of it installed as I was in a hurry). Not too much though, but I will address this soon using some foam padding to get a snug fit.
I was actually surprised to see that the SLAs voltage was at 12.25v after removing (which was after a short drive, so it should have been fairly charged up).
I don't think the battery is close to end-of-life, because it's less than a year old of my ownership as the EV was bought new.
Maybe it's just a sign if how inefficient SLAs are...

I tested the Lifepo4 in the car before driving away and all was good.
I did about 150km of driving with the Lifepo4 and it seems to work perfectly!
I had the 12v SLA in the trunk (with tools) just in case something would go wrong.
But now I'm more confident with it, so a did a little drive without the SLA in the trunk.
I don't know if it's only psychological or not, but the front end of the car seems to feel lighter and a bit more responsive.

I have a battery monitor connected permanently, and the Lifepo4 was holding nicely around 13v when the car was in "auxiliary mode" (HV battery disconnected and therefore DC-DC converter off). I turned on the headlights, radio and even A/C fan and it held up well.

Sorry for this post being so long, but I wanted to document my experience and maybe it will help others too.

I hope the Lifepo4 will hold up well in all seasons, but only time will tell.
 
great work!

shame the aliexpress battery tests at just under half nameplate ampere hour label. hoping you can get something back for the difference.

so cool that you replaced the stock lead acid battery with lifepo4!

thank you for the update on project

i imagine the charging efficiency increase and flatter voltage curve ought to give back a few percent of energy :) awesome work!
 
As suggested on this thread and in other EV forums, I was convinced to swap the 12v SLA battery in my EV for a 12v Lifepo4.

The advantages are clear: less weight and flat voltage curve should help my EV be more efficient.

I thought of doing this as a second stage of my ongoing project, but as I found a fairly cheap 12v automotive Lifepo4, I decided to go along with it in parallel to my efforts of the buffer battery setup.

So about a month ago I ordered this one:
Costing about $235 shipping included.

It's a 12v 45Ah Lifepo4. In theory it has the ability to directly replace a car's 12v SLA starter battery. I say in theory because although it's rated for hundreds of CCAs (cold cranking amps) which is needed to start an internal combustion engine, I wouldn't count on a high lifetime in this scenario. I've seen teardowns on YouTube of better known Lifepo4's that claim to be suitable as a starter battery (such as the LiLead 110). They really aren't much different than your solar oriented deep cycle Lifepo4's.
The only difference maybe in the BMS programming, which will tolerate a very high current surge for a few seconds.
The LiLead for example uses standard EVE prismatic cells. These are rated at standard 1-2C rates discharge. There are some more professional models that use cylindrical Lifepo4 cells, or even better high C rated headway cells. But these are majorly expensive (around $1k).

Anyways...
An EV 12v battery doesn't need alot of cranking amps since there is no cranking going on. It's just needed to power the contactors of the HV battery and from then on the onboard DC-DC converter supplies all the 12v power (in addition to charging up the 12v).
I got this specific model and not a standard deep cycle Lifepo4 (which costs roughly the same) because this one has a wider temperature range, which I thought would be better suited for a battery under the hood of the car (and not indoors). The seller claims that they use Lifepo4 pouch cells for these types of "starter" batteries (which personally I'm not too fond of), instead of prismatic which they use in their energy storage models.

So... I finally received it!
Of course the first thing I did was to charge it up and do a proper capacity test. Just out of curiosity and to test how it works before installing in my EV.
Here was a disappointment: It tested out only to 20Ah instead of 45Ah!

I did this charge-discharge test twice @0.2C rate but got similar results both times:
View attachment 74844

The second time I charged it up, I connected a shunt to see how many Ah went in up to 14.6v (as recommended) and sure enough it was only 20Ah.

Capacity is not so critical in an EV because it's use is very little as mentioned. But I was hoping that I would get closer to the advertised capacity, which would allow me to run some auxiliary loads straight from the 12v system for a while before draining too much.

In any case I contacted the seller and opened a dispute on Aliexpress.
The seller claims this is OK because starter batteries are "rated differently". I personally think it's B.S. he scammed me with false advertising stating its a 45Ah battery.
We will see how this dispute goes (I requested a refund of $100).

Capacity aside this little battery is actually pretty cool! It's got a nice look to it, with rubbery casing (doesn't look/feel like cheap ABS plastic) and properly sized terminals for car batteries (positive larger than negative so that you won't mix them up).
It's also so lightweight! Only 3.4kg !
I don't know if 3.4kg should have hinted to me something about the accuracy of the 45Ah claimed capacity or not. Does this sound logical in terms of Lifepo4 energy density?

I went ahead and removed the 12v SLA. Darn that battery is heavy!
It's a 60Ah SLA, just like you get in a fuel burning ICE vehicle. It's amazing how manufacturers try to reduce costs by putting off-the shelf ICE batteries in EVs even when they aren't really suitable for an EVs needs.
View attachment 74847View attachment 74848View attachment 74850
I used a suitcase style hand held scale and it came in at 16.5kg !
That's a weight saving of 13kg in favor of the Lifepo4!

It was a pretty simple swap once I found the correct bolts to open in the battery tray.
The Lifepo4 is a bit smaller than the SLA so it has a bit of room to move in the tray (sorry I didn't take pictures of it installed as I was in a hurry). Not too much though, but I will address this soon using some foam padding to get a snug fit.
I was actually surprised to see that the SLAs voltage was at 12.25v after removing (which was after a short drive, so it should have been fairly charged up).
I don't think the battery is close to end-of-life, because it's less than a year old of my ownership as the EV was bought new.
Maybe it's just a sign if how inefficient SLAs are...

I tested the Lifepo4 in the car before driving away and all was good.
I did about 150km of driving with the Lifepo4 and it seems to work perfectly!
I had the 12v SLA in the trunk (with tools) just in case something would go wrong.
But now I'm more confident with it, so a did a little drive without the SLA in the trunk.
I don't know if it's only psychological or not, but the front end of the car seems to feel lighter and a bit more responsive.

I have a battery monitor connected permanently, and the Lifepo4 was holding nicely around 13v when the car was in "auxiliary mode" (HV battery disconnected and therefore DC-DC converter off). I turned on the headlights, radio and even A/C fan and it held up well.

Sorry for this post being so long, but I wanted to document my experience and maybe it will help others too.

I hope the Lifepo4 will hold up well in all seasons, but only time will tell.
You gotta love Lifepo4!
This chart is from my battery meter installed on my car. You can clearly see that with the SLA battery, voltage would drop down pretty close to 12v, and the onboard DC-DC converter would need to charge it up more.
The Lifepo4 (swapped at Dec. 06) drops to around 13v, and therefore need less charging :cool:

Capture+_2021-12-08-09-42-52.png
 
@meetyg, you have proven that the upgrade from PbA to LiFePO4 has improved system voltage stability! wow! thank you for proving out this concept thus far. the documentation in this thread intensely motivates me to continue with my self chosen research path as well. big up.

thank you for these updates. your methodology is commendable, and i think your sensitivity to safety while still making progress is also incredible.


as it happens, the 12V PbA battery in my car died recently, and sourced a 12V LiFePO4 equivalent capacity pack to replace it. still doing due diligence on the self install. your graph of voltage vs time has massively reinforced the decision to include continuous voltage and shunt ampere logging at some timescale like 1 per minute at least.

thank you, and i know you’re not done yet ;)

☀️??
 
Back
Top