Antiscab's NWH10 Prius plugin Lithium conversion

Post up a thread for your EV. Progress pics, description and assorted alliteration
antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Sat, 19 Apr 2014, 20:14

some more progress.

I discovered the main fuse is inside the service disconnect, and the wires going in are just bolted down, so varying the lengths or even putting different wire in is quite doable:

Image
Image
Image


This is the pin out for the wires going from the cars original battery ECU to the battery
Image

0 = pack negative
1 = point 1, etc

voltage is measured at 20 points evenly throughout the pack
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Mon, 21 Apr 2014, 02:05

more progress,

most of the electronics mounted to the battery box - new holes drilled for the battery ECU wires

Image

Image
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Mon, 21 Apr 2014, 02:06

theres 10mm polycarb material being used as a spacer so all the batteries sit at the same height, depite there being bolt heads in the bottom of the battery box
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Antiscab's NWH10 Prius plugin Lithium conversion

Post by BigMouse » Mon, 21 Apr 2014, 05:33

Grade 8.8 bolts on the fuse? That's a bit overkill! So what's the latest plan for this project? Where are you headed with it?

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Tue, 22 Apr 2014, 01:24

wiring for the original Battery ECU is done

just need to put all the balancing boards on all the cells, then put a single cell charger on each cell for a top balance, then put it all in

oh, and mount up the on board charger so it can be a PHEV
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

User avatar
Bluefang
Groupie
Posts: 90
Joined: Wed, 13 May 2009, 22:37
Real Name: Derek Hohmann
Location: Gold Coast, Queensland

Antiscab's NWH10 Prius plugin Lithium conversion

Post by Bluefang » Tue, 22 Apr 2014, 03:10

How are you going to trick the voltage readout of the 20 points? I would guess that the voltage at the measured points on the lithium pack would be different to the original pack?

Have you ever thought about doing a full standalone control conversion? Ie run the electronics from a DIY controller and the engine from a standalone ECU like a Megasquirt.
www.facebook.com/ausemoto
Building Electric vehicles to be enjoyed and give people a EV grin.

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Tue, 22 Apr 2014, 03:17

Bluefang wrote: How are you going to trick the voltage readout of the 20 points? I would guess that the voltage at the measured points on the lithium pack would be different to the original pack?


I won't have to fake it - the lithium voltages will be close enough to the old nimh voltages

Bluefang wrote:
Have you ever thought about doing a full standalone control conversion? Ie run the electronics from a DIY controller and the engine from a standalone ECU like a Megasquirt.


That is a bigger project than you would imagine - I would need two motor controllers, the megasquirt, and an additional controller to decide where the power is going to come from when I push the accelerator pedal

It took one of Toyota's R&D departments all of 1995 and 1996 to figure that out, I'm not game on doing the same :D
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

suziauto
Groupie
Posts: 94
Joined: Fri, 20 Mar 2009, 16:55
Real Name: graeme Manietta
Location: brisbane
Contact:

Antiscab's NWH10 Prius plugin Lithium conversion

Post by suziauto » Wed, 21 May 2014, 01:25

Hi Matt,

Very interested in this conversion as we also have a NWH10 .. we converted it from sticks to Gen2 cells, they were used and after 10,000k's they've lost capacity .. our options is another set of gen2 , revert to the new sticks or swap out the whole pack to Lithiums and we'd prefer that and make it a PHEV. The charger looks great .. hope to see how well it works..

Cheers Graeme Manietta

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Wed, 21 May 2014, 04:54

Well I'm hoping to have mine running next week, so I can report back how it goes.

It should have been running last week, but I made an error putting the battery together, so have had to pull it all apart again. the first time is always the hardest and most time consuming
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

User avatar
Bluefang
Groupie
Posts: 90
Joined: Wed, 13 May 2009, 22:37
Real Name: Derek Hohmann
Location: Gold Coast, Queensland

Antiscab's NWH10 Prius plugin Lithium conversion

Post by Bluefang » Wed, 21 May 2014, 05:49

Really going to be awesome if it works with the lithium batteries. I guess you wouldn't be able to use the Lithium batteries that range between 3.5 and 4.2V? Some really nice cells can now be bought due to the ramp up of production for Tesla etc.
www.facebook.com/ausemoto
Building Electric vehicles to be enjoyed and give people a EV grin.

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Tue, 03 Jun 2014, 05:22

Huzzah, as of today the car is now running

I still have to change the 12v battery, as it is very dead

and do rather a lot of tidying, but it runs

more testing to follow this week

Bluefang wrote: I guess you wouldn't be able to use the Lithium batteries that range between 3.5 and 4.2V?


I don't know how the car would handle the voltage range.

cell count has to be a multiple of 20 to keep the car's BMS happy

80 cells would be too low in voltage (the car would overcharge the battery while driving)

100 cells may have too high a voltage, at best you may find you have no regen until the battery is completely flat

Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Tue, 03 Jun 2014, 23:41

I drove the car home today, after putting some new tyres on and a new 12v battery

it drove fine (even on the 3 year old petrol)

disappointingly, the car appears not to be using much of the batteries' capacity.

I was expecting the car to determine battery SOC by voltage only, but it appears I was mistaken

I'm meeting up with Simon to clear out all the error codes, hopefully that will correct the issue somewhat.

I'm also planning on installing a cycle analyst to get some more information
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

User avatar
BigMouse
Senior Member
Posts: 600
Joined: Thu, 28 Oct 2010, 02:39
Real Name: Vincent Tannahill
Location: Silicon Valley
Contact:

Antiscab's NWH10 Prius plugin Lithium conversion

Post by BigMouse » Wed, 04 Jun 2014, 00:13

The SOC/voltage relationship between NIMH and LFP are very different. If it uses voltage, then you'll see the SOC not actually changing much at all, even if it is being used. Do you have an ammeter? You might get lucky if the pack voltage sits in a "sweet spot" where the controller will both draw from and regen to the battery. This could happen without a noticeable change in voltage until you reach an extreme (either almost full or almost empty).

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Wed, 04 Jun 2014, 23:22

BigMouse wrote: If it uses voltage, then you'll see the SOC not actually changing much at all, even if it is being used. Do you have an ammeter? You might get lucky if the pack voltage sits in a "sweet spot" where the controller will both draw from and regen to the battery.


The battery guage on the car instrumentatoin swings between full and empty the same as it did when the old nimh battery was connected.
That leads me to suspect it is doing an Ah count, with a low expected capacity.

I'm installing a cycle analyst tonight to see what's happening on the battery side.

The voltage on the 330V 20Ah nominal LiFePO4 battery should always be high enough for the car to think the battery is full (the original battery was a 288v 6Ah nominal nimh battery)

The car does regen and electric assist normally, I just have a hard time believing a single hard acceleration would really use 5kwh :p
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

User avatar
Simon
Senior Member
Posts: 431
Joined: Sun, 19 Aug 2007, 19:38
Real Name: Simon
Location: Perth WA
Contact:

Antiscab's NWH10 Prius plugin Lithium conversion

Post by Simon » Thu, 05 Jun 2014, 17:38

My Prius did the same thing with the brand new NiMH cells. It wouldn't use hardly any electric power and would go from "full" to yellow on the battery gauge with one hard acceleration. Once I cleared the error codes it ran fine.

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Tue, 17 Jun 2014, 17:43

I can't clear the codes, as the battery ECU isn't seeing the voltage from one of the cell groups, so I have to try another battery ECU

I finally hooked up a cycle analyst yesterday (that was an ordeal in itself)

I discovered the car will push the battery voltage up to 390V if it thinks the battery is empty

curiously the car won't declare the battery as being full at this point - so it does look like it does an Ah count

I haven't gone for a drive yet with the cycle analyst (still some wiring floating around in the boot that needs tidying)
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Sun, 22 Jun 2014, 01:25

the cycle analyst is sort of working - I had to modify it to go up to 400v.

anyway, the A and Ah works

I'm still driving with the fault codes, so the car is still minimising use of the battery (capacity wise), but I wanted to see what the car would do when it doesn't trust the battery.

I have measured max discharge at 70A, only momentarily, or during a hard acceleration.

max charge rate during regen is 45A.
max intentional charge rate is 15A

all that the GBS 20Ah cells should be able to handle.

The car is presently only using 0.1Ah of capacity - which is why a single acceleration would appear to flatten the battery.

another curious thing is the car is assuing the battery is lossy Ah wise, so it will gradual charge the battery. if the battery gets too full because the lithium battery is not lossy, the car gets very unhappy.

my work around is to use my Vectrix to discharge the prius battery.
not quite what I had in mind for a plug in prius :p

I will try my spare battery ECU - to see what the car does when it does trust the battery
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Sat, 28 Jun 2014, 20:04

So I swapped the Battery ECU

The spare Battery ECU also reports cell group 17 over 20V and cell group 18 at 0.00v

so I must have a wiring issue

Driving around with the car in fault mode - the car puts on average 1Ah of charge more than it takes out, per 100km

I suspect the value would be lower if the car were working properly, but still proplematic.

before I started looking at this conversion, I was looking at putting a pack of 10Ah headways and not going the plug in route.
the idea being the headways would last longer than a new set of nimh cells.

It looks like I need a way to tell the car the battery is in fact full.

Once I get the wiring issue sorted out, I can look at disconnecting the battery ECU from the 12v, so it forgets.

when this happens the battery ECU gets the car to charge the battery up to ~385v and then sets the gauge to full.

Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

gtyler54
Noobie
Posts: 26
Joined: Mon, 13 Aug 2012, 00:54
Real Name: George Tyler
Location: Hamilton NZ

Antiscab's NWH10 Prius plugin Lithium conversion

Post by gtyler54 » Tue, 12 Aug 2014, 22:33

Hi Matt, although I did join this forum years ago, can't actually remember when, I found this thread in a google search for a wiring diagram for the NHW10 engine ECU.
   I have been working on NHW10's for many years, done i would guess 50 repairs. I have both a rennacs and a Tecu scanners that read and can also talk to the computers. Why do you think that the car does not do charge counting for SOC calculation? What I have seen makes me thing it does. For one thing, you can have different SOC at the same, or even reversed voltage. Have you read the SOC from the battery ECU?
   I have been think of doing this myself, also with headway cells. I was thinking of 6 cells between taps instead of your 5 because of the maximum cell voltage. I have found some conflicting data on headway cells, some say min 2V per cell and some 2.5V. the battery can charge at least to 370V as this is the highest I have seen. It could be more. with 120 cells this would mean 3.083 V per cell maximum, and a minimum of 2V per cell. If the 2V per cell is actually ok all may be fine as the car stops discharge at 240V, average charge on the cells will be on the low side, you won't be able to fully charge the pack as the car will limit it to something below 370V I think.
   With 5 cell strings between taps there will be 100 Cells. At 370V this will give 3.7V maximum, which is also where I think the car will stop charging. minimum voltage will be 240V/100 = 2.4V which I think is better. Even if 2V is ok there is not leeway. I guess you have a similar situation in either direction!
Maybe a BMS system of our own, working on what used to be the thermistor strips? At the moment there is a fixed resistor in series with all the thermistors, if we had a relay contact in series this would cause the ECU to drop the contactors if the voltage goes above 360V?
gtyler

gtyler54
Noobie
Posts: 26
Joined: Mon, 13 Aug 2012, 00:54
Real Name: George Tyler
Location: Hamilton NZ

Antiscab's NWH10 Prius plugin Lithium conversion

Post by gtyler54 » Tue, 12 Aug 2014, 22:50

my plan was to make a device on the computer bus that resets the state of charge to 100% continuously. I can do that with the cheap russian scanner, so if I can capture the codes it sends I can program a Aurdiuno or a PIC to do this.
gtyler

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Thu, 14 Aug 2014, 04:41

gtyler54 wrote:
   I have been working on NHW10's for many years, done i would guess 50 repairs. I have both a rennacs and a Tecu scanners that read and can also talk to the computers. Why do you think that the car does not do charge counting for SOC calculation? What I have seen makes me thing it does. For one thing, you can have different SOC at the same, or even reversed voltage. Have you read the SOC from the battery ECU?


I too have a TECU scanner and have read the live data

The reason I thought it was by voltage was because I once found a lookup table for the NWH10 battery - SOC vs voltage vs current

I thought since it was using a hall sensor for current measurement and wasn't using the batteries full capacity, it was more likely to be using a lookup table rather than AH counting

on my own car I have a break on the sense line to battery point 3 from the negative (note the TECU software lists the battery measure points numbered from the positive, which confused me to no end)

gtyler54 wrote:
   I have been think of doing this myself, also with headway cells. I was thinking of 6 cells between taps instead of your 5 because of the maximum cell voltage. I have found some conflicting data on headway cells, some say min 2V per cell and some 2.5V. the battery can charge at least to 370V as this is the highest I have seen.


highest I have measured is 380vdc

the car gradually keeps recharging the battery as it's expecting it to be lossy. except lithium isn't, so the battery just kept getting full.
once the voltage reaches 380v, a battery error is flagged and the battery is dicsonnected.

my GBS 20AH cells were top balanced before I installed them so they survived.

I installed a cycle analyst which revealed max momentary discharge was 85A.

max regen current was around 65A from memory

I'm using 100 cells, with each module composed of 5 cells

gtyler54 wrote:
Maybe a BMS system of our own, working on what used to be the thermistor strips? At the moment there is a fixed resistor in series with all the thermistors, if we had a relay contact in series this would cause the ECU to drop the contactors if the voltage goes above 360V?


dropping the contactor results in a car that is not driveable regrettably
gtyler54 wrote: my plan was to make a device on the computer bus that resets the state of charge to 100% continuously. I can do that with the cheap russian scanner, so if I can capture the codes it sends I can program a Aurdiuno or a PIC to do this.


That would solve my biggest problem of getting the car to discharge the battery on average rather than recharging it on average while driving

Would you be able to make a device that defaults to setting the SOC to 100% repeatedly until the lowest module voltage reaches 14.2v (worst case 4 x 3.3v + 2v or just 3v average)

at that point it would be best to set SOC to 10% or 20% and stop updating so the car just acts like a normal prius until next restart.
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Thu, 14 Aug 2014, 04:43

gtyler54 wrote:
   I have been working on NHW10's for many years, done i would guess 50 repairs. I have both a rennacs and a Tecu scanners that read and can also talk to the computers.


do you happen to know what connector the battery ECU uses for the voltage sense wires?

I gave up looking for one and just stole the female off one of my spare battery ECU's to use on my wiring tester - that was how I finally worked out the battery numbering on the software was backwards
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Thu, 14 Aug 2014, 04:57

This is the break out connector I ended up making (took me all day friday last week)

Image

I'm having another circuit board to go on top that leads to 20 volt meters, so I can quickly test a number of things rather than fiddling with a multimeter
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

gtyler54
Noobie
Posts: 26
Joined: Mon, 13 Aug 2012, 00:54
Real Name: George Tyler
Location: Hamilton NZ

Antiscab's NWH10 Prius plugin Lithium conversion

Post by gtyler54 » Thu, 14 Aug 2014, 13:33

Sorry, I don't know what the connector is. you do know that the battery is numbered backwards? i.e. 1 is the most positive. do you have the "tecu"scanner software? total cost about $20. http://vdiag.net/tecu3.html
http://www.ebay.com.au/sch/i.html?_saca ... c&LH_BIN=1
for the cable.
gtyler

antiscab
Senior Member
Posts: 2235
Joined: Mon, 26 Nov 2007, 05:39
Real Name: Matthew Lacey
Location: Perth, WA

Antiscab's NWH10 Prius plugin Lithium conversion

Post by antiscab » Fri, 15 Aug 2014, 04:37

I actually have two cables and the TECU software (one of them is actually in the above picture)

in what sort of time frame are you likely to make a device that auto updates the battery SOC in the ECU?
Matt
2011 Blade Electron mk6
2007 vectrix - 145'000km
1998 Prius - needs batt
1999 Prius - needs batt

Post Reply