4Springs' Brumby

Post up a thread for your EV. Progress pics, description and assorted alliteration
User avatar
jonescg
Senior Member
Posts: 3601
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: 4Springs' Brumby

Post by jonescg »

Looks like you might be missing out on a bit more than you think at the top end. But still, that range isn't great for the size of the battery.
AEVA National Secretary, WA branch chair.
francisco.shi
Senior Member
Posts: 423
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: 4Springs' Brumby

Post by francisco.shi »

I did some measurements a while ago because the state of charge did not seem to match what was on the data sheet.
What I did was measure
I started at about 3.2v Then I put the battery on charge and measured the voltage with respect to the Ah that had gone into the battery. Then worked out a percentage based on the total Ah that had gone into the battery. I did this using a solar inverter. The problem I had was similar to you, I was supposedly getting less capacity than what I expected (from the battery not supplying the charge for long enough) Then I found that it was the estimation of charge that was the problem. At the time I did not have enough solar production to fully charge the battery. It would always go flat at night.
Here is what I got:
(some numbers are back to front because this is from the code I used to estimate the SOC)
4165-4115 90-100%;
4115-4075 80-90%;
4075-4055 70-80%;
4055-4035 60-70%;
4035-4015 50-60%;
4015-3985 40-50%;
3985-3925 30-40%;
3925-3855 20-30%;
3855-3785 10-20%;
3785-3475 0-10%

So if you are stopping at 4v then you are probably only using about 50% of the capacity
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

Now I have some figures from charging this pack, to compare with the discharge figures.

When I started driving (discharging), cell voltages were around 3.95 V. I need to fine tune the charging algorithm in the BMS to get them up a bit higher. I drove for 106 km, which I estimate to be about 20 kWh based on previous experience.

When I started charging (about 15 minutes after I finished driving), the cell voltages were 3.12 V to 3.26 V.
While charging I was monitoring the kW used on the AC side. It was about 3.3 kW.

Here is a graph of the first 50 mins of the charging cycle. Y axis is one point for every 15 seconds, so 100 is about 25 minutes.
Charging 50min.png
Charging 50min.png (150.93 KiB) Viewed 810 times
The cell voltages have clustered back up nicely by the end of this graph (50 minutes). 3.3 kW * 50 minutes = 2.75 kWh delivered by this stage.

The charge went for 660 minutes before the BMS cut off the charger. At that point the cells had a voltage range of 3.92 V to 3.95 V. 3.3 kW * 660 minutes = 36 kWh.
So this does match up with my earlier estimate of a 35 kWh pack. Don't know why I only went 106 km though!
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

francisco.shi wrote: Sun, 09 Aug 2020, 10:23 I did some measurements a while ago because the state of charge did not seem to match what was on the data sheet.
...
So if you are stopping at 4v then you are probably only using about 50% of the capacity
Now you tell me!
If I use your table with my charging figures above, I only charged to about 40% (3.95 V)

I went back to check the voltage of my charger. Apparently it will charge to 175.2 V, not the 172.2 V I reported here. This is according to a post I made about it on the forum here: viewtopic.php?p=46341#p46341.
175.2/43 = 4.07 V per cell.

So I adjusted the BMS to suit and let it continue charging. It was charging for an hour, and the voltage had only moved 0.03 V, so I was thinking that Francisco's table must be correct. But after a while I realised that the charger current was tapering off as the pack went up above 141 V (3.98 V per cell). I've turned it off now, I'll set it going again tomorrow when I can keep an eye on it.
User avatar
coulomb
Site Admin
Posts: 4305
Joined: Thu, 22 Jan 2009, 20:32
Real Name: Mike Van Emmerik
Location: Brisbane
Contact:

Re: 4Springs' Brumby

Post by coulomb »

4Springs wrote: Sun, 09 Aug 2020, 18:16The charge went for 660 minutes before the BMS cut off the charger. At that point the cells had a voltage range of 3.92 V to 3.95 V. 3.3 kW * 660 minutes = 36 kWh.
So this does match up with my earlier estimate of a 35 kWh pack. Don't know why I only went 106 km though!
But did the charger take the full 3.3 kW for all 11 hours, or nearly all? Usually the charge current and therefore power will taper off towards the end.

But if Francisco is right, you're only charging to ≈50% of the battery capacity; in that case I would not expect a tapering of charge current.

[ Edit: added "current and therefore power" after "charge" ]
Nissan Leaf 2012 with new battery May 2019.
5650 W solar, 2xPIP-4048MS inverters, 16 kWh battery.
1.4 kW solar with 1.2 kW Latronics inverter and FIT.
160 W solar, 2.5 kWh 24 V battery for lights.
Patching PIP-4048/5048 inverter-chargers.
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

coulomb wrote: Sun, 09 Aug 2020, 18:40 But did the charger take the full 3.3 kW for all 11 hours, or nearly all? Usually the charge current and therefore power will taper off towards the end.
Probably the full rate for the 11 hours, but with room for a little doubt.
There are actually two chargers, and I can only monitor the consumption of one of them at a time. Charger 2 was drawing 1.2 kW when I checked it early on. Charger 1 was drawing 2.1 kW, and I monitored this one for the second half of the 11 hours. Charger 1 drew the full 2.1 kW for the whole time until the BMS told it to stop.
The tapering I reported on didn't commence until about 20 minutes after I restarted the charger with a higher voltage limit in the BMS. It came down to about 0.7 kW. I'll set this going again tomorrow and see what it does.
User avatar
brendon_m
Senior Member
Posts: 681
Joined: Sat, 28 Oct 2017, 11:00
Real Name: Brendon McCarrol

Re: 4Springs' Brumby

Post by brendon_m »

Ideally you really need a power meter with a total used/history to work out how much power the pack actually absorbed (minus losses). I've tried guessing SoH on my cars by using instantaneous power and extrapolating from there but it was such vague results I ended up buying a different meter.
Also knowing power factor is important and whether the meter accounts for it. I've had 2 meters in series with each other, one accounted for power factor and the other just assumed pf of 1 for everything and I've ended up with very different results on some consumption tests (like one literally reading twice the other, but that was especially bad pf)

There may not be one big reason for the short range but a list of smaller ones that add together like not fully charged, not fully discharged, too many launches on the trial trip, bit of unnoticed headwind, cold weather sapping range, handbrake been on for a while and is now dragging, wheel bearings stiff etc.
Only way to find out is more testing I guess
francisco.shi
Senior Member
Posts: 423
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: 4Springs' Brumby

Post by francisco.shi »

As the charger gets close to its maximum output voltage it may not be able to push the required current so you may not be charging the battery with as much current as you think.
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

I've made some adjustments to the BMS, and the charger is now charging to the limit, whereas previously the BMS was turning it off at a lower limit. This added a few extra kWhs to the pack, but it is still not fully charged.
I took it for another test drive, and managed a more respectable total of 140 km.
The end of this trip was designated by the lowest cell dropping below 3.1 v for a couple of seconds. A light comes up on the dash to warn you. This happened under acceleration, but by another km or two it was occurring continuously. I think this is a reasonable indicator of pack almost depleted.
I have a new Zeva Fuel Gauge Driver. I had previously set up the 'full' and 'empty' points to match my fuel gage. This device determines when to indicate empty based on an estimate of Amp hours. When empty, I adjusted it until the gauge pointed to empty, and this was at the 210 AH setting. So this gauge was estimating that I had used 210 AH for my trip, or around 33 kWh (using a nominal 3.7 v per cell).
I then charged the battery, using just one charger, and with this plugged into my power meter. Here is the graph:
Smappee Power Log.png
Smappee Power Log.png (13.86 KiB) Viewed 673 times
The big dip is when I needed to turn off the charger to re-arrange the cars (I had to move the Brumby, to access the Kona, to get it out of the way of the Outlander...).
The graph starts to taper at 9:00. At this stage the total (not shown on the graph) was 33 kWh. Eyeballing the graph, I'll call it 'practically finished' at 14:00. The total at this stage was 38 kW. 'Completely finished' (18:00) is 39 kWh.
Note that there is a constant draw even when finished. Close up it looks like this:
Smappee Baseline.png
Smappee Baseline.png (18.51 KiB) Viewed 673 times
There are four battery warmers, cycling on and off as they come up to temperature. This averaged at about 100 W overnight. This can be taken as a maximum, since the cells were warm during the charging proper, and the warmers wouldn't have used this much power.

So 'Practically Finished' was about 38 kW. Subtract 2 kW for the battery warmers and misc stuff gives 36 kWh, which is not too far from my original estimate of 35 kWh, and the Zeva gauge of 33 kWh. I might call it 35 kWh. I could get a bit more in with a charger set to a higher voltage.
With a trip of 140 km, this equates to about 25 kWh per 100 km. I was thinking that this was pretty bad, but I've just had a look through some efficiency data from 2015, and found that 25 is actually pretty good for the Brumby. The table I'm looking has the best value as 22, but with a more typical value being 30. Or 39.8 when I towed a trailer with 19 hay bales on it!
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

brendon_m wrote: Sun, 09 Aug 2020, 08:32
4Springs wrote: Sun, 09 Aug 2020, 08:14 I drove it yesterday until the lowest cell reached 3.1 V under load. Then I drove the 1 km or so home after that. I managed 106 kM.
Did you let the pack sit after that to see what the voltage settled to?
This time I did, and found that the lowest cell had come back up to 3.17 v after about 30 mins. Highest was 3.29 v.
User avatar
jonescg
Senior Member
Posts: 3601
Joined: Thu, 21 Jan 2010, 23:05
Real Name: Chris Jones
Location: Perth, WA.

Re: 4Springs' Brumby

Post by jonescg »

4Springs wrote: Tue, 18 Aug 2020, 09:32
The big dip is when I needed to turn off the charger to re-arrange the cars (I had to move the Brumby, to access the Kona, to get it out of the way of the Outlander...).
Darryl Kerrigan: “Oh Steve, could you move the Camira? I need to get the Torana out so I can get to the Commodore.”
Steve Kerrigan: “I’ll have to get the keys to the Cortina if I’m gonna move that Camira.”
AEVA National Secretary, WA branch chair.
T1 Terry
Senior Member
Posts: 1155
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: 4Springs' Brumby

Post by T1 Terry »

Does your BMS sense high cell voltage to stop the charging as well as the total battery voltage? The BMS we built for house battery use has the high cell voltage stop as a secondary protection and drives a timer. The timer can be set anywhere between 1 minute and 1 hr, it then senses for another high cell voltage alarm and repeats the charging off cycle. If there is no high cell alarm, charging resumes. This allows the charge to settle through the cell/s so a true 100% saturation charge can be achieved.

T1 Terry
I do like your idea of a warning light if a cell goes low voltage, maybe a counter added so after a total of say 1 minute of the light being on, limp mode is activated so you can at least pull off the road and wait for the cell balancers to shift a bit of the capacity to the low cell so you can limp to the nearest rescue point, be it a power point or a spot where it can be loaded onto a trailer
Green but want to learn
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

T1 Terry wrote: Tue, 25 Aug 2020, 10:50 Does your BMS sense high cell voltage to stop the charging as well as the total battery voltage?
Yes. The way it was set up with LFP cells was that it would turn the charging current down when the first cell reached the voltage at which it started bypassing. The lower charge current was about the same as the bypass current on the BMS cells, so this gave the other cells a chance to catch up. The charger was disabled as soon as every cell was in bypass.

With NMCs, the balancing is done at any voltage, so the end of charge situation is different. It is spending hours in the last top-up phase. I've found that the cell voltages are the same (about 4.02 v) 24 hours after charging is finished.
T1 Terry wrote: Tue, 25 Aug 2020, 10:50 I do like your idea of a warning light if a cell goes low voltage, maybe a counter added so after a total of say 1 minute of the light being on, limp mode is activated so you can at least pull off the road and wait for the cell balancers to shift a bit of the capacity to the low cell so you can limp to the nearest rescue point, be it a power point or a spot where it can be loaded onto a trailer
At the moment the light is it on this system. A cell voltage needs to be below the threshold for a few seconds before the light comes on. After that notification it's up to the driver to not kill the pack!
I do have a low-low voltage threshold, but I don't do anything about it. This iteration of the BMS has some unused outputs, so I could hook one up to the drive contactor and disable it if a damaging voltage was imminent. This is a pretty severe thing to do to a driver though...
T1 Terry
Senior Member
Posts: 1155
Joined: Thu, 30 Sep 2010, 20:11
Real Name: Terry Covill
Location: Mannum SA

Re: 4Springs' Brumby

Post by T1 Terry »

With NMCs, the balancing is done at any voltage, so the end of charge situation is different. It is spending hours in the last top-up phase. I've found that the cell voltages are the same (about 4.02 v) 24 hours after charging is finished.
With the balancers operating all the time, how do you determine the difference between cell voltage and cell capacity? Is the cell voltage and charge capacity linear like a lead acid cell, assuming ther is no surface charge involved?
I was just thinking that if there was an alarm signal when a cell hit X voltage, it could stop the charging and allow the balancers to do their thing, then the balancer turned off and the charger started again with the pattern repeating until all cells were at X voltage at rest.

T1 Terry
Green but want to learn
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

T1 Terry wrote: Wed, 02 Sep 2020, 13:22 With the balancers operating all the time, how do you determine the difference between cell voltage and cell capacity? Is the cell voltage and charge capacity linear like a lead acid cell, assuming ther is no surface charge involved?
I'm not sure that the difference is of any consequence, and yes, I'm assuming that the voltage is reasonably linear.
T1 Terry wrote: Wed, 02 Sep 2020, 13:22 I was just thinking that if there was an alarm signal when a cell hit X voltage, it could stop the charging and allow the balancers to do their thing, then the balancer turned off and the charger started again with the pattern repeating until all cells were at X voltage at rest.
Remember my balancers are just resistors. With LFPs, when a cell reached 3.5 V I'd slow the charger down to about the same rate as the resistors. So it slowly topped up the low cells until they were all the same voltage.
With NMCs my charger tapers the current off to almost nothing, so the effect is about the same.

I drove 60 km today. The cells were 4.02 to 4.03 V when I started, and 3.67 to 3.68 V when I got home again. So they seem to be balanced over a wide range.
francisco.shi
Senior Member
Posts: 423
Joined: Mon, 17 Sep 2018, 16:30
Real Name: Francisco
Location: Brisbane, Strathpine

Re: 4Springs' Brumby

Post by francisco.shi »

What is the cell voltage when fully charged?
User avatar
4Springs
Site Admin
Posts: 671
Joined: Thu, 23 Dec 2010, 01:14
Real Name: Christopher Walkden
Location: Selbourne, TAS

Re: 4Springs' Brumby

Post by 4Springs »

francisco.shi wrote: Wed, 02 Sep 2020, 20:15 What is the cell voltage when fully charged?
4.02 - 4.03 V
So not full (4.166 V). This voltage should equate to about 90% if you believe the published specs, or about 50% if you believe Francisco's table. I'm putting in about 35 kWh out of the rated 40 kWh, or 87%.

There are two ways I can see to increase the capacity of the pack. The first is to remove two 'cells' (four in parallel, so really 8 cells). This would let the charger charge the cells to the highest capacity. I'd get 100% of a reduced capacity pack, which equates to about 38 kWh.
The second option is to purchase another charger, set to a higher voltage. This would get me about 40 kWh. My medium term plan is to purchase a 3 kW charger to add my two existing chargers, giving a total of about 7 kW. So since my plans include the purchase of a charger anyway, I think this is the way to go. The maximum pack voltage would be 179 V, which is inside the specifications of the Kelly controller (180 V) and the DC/DC converters (can't remember but I'm sure I looked it up). I'm not sure about the Masterflux compressor though, I can't seem to find the paperwork to tell me which model I have. Current models look like they are designed to go up to either 165 V or 260 V, so hopefully it is one of the latter.
Having only one charger that goes up to the full voltage would give me a handy way of doing a default '80%' charge. So in normal circumstances I just plug in the current charger as normal (into a 10 A socket). If I wanted to charge faster, or to charge up to 100%, I'd plug in the new charger (into a second, 15 A socket).
Post Reply