FW: (ET) batt's, new gauge points, and a faq draft for discussion

["Elie, Larry (L.D.)" <lelie ford com>]

I feel a bit humbled.  Over a week ago, after posting a quick suggestion 
on charge voltage, I was corrected by Harry Landis and Marty Canada.  
Although I have worked with automotive charging systems for years, and 
have 4 ET's at the moment, I began to doubt some voltages I had read.  In 
private E-mail, Harry was quite persuasive on the levels I was claiming 
AFTER charging.  I have been working with ET's for years, and they have 
not all been in the same condition.  That troubled me.  So, last weekend, 
I decided to check the gauge on some tractors.  I had before, but, hey, 
let's be sure.  I used a good HP power supply, and a DVM that does true 
RMS, although the AC ripple from this supply is tiny, to make sure that I 
read what the gauge read.  Turns out my 'best' gauge did NOT agree with my 
old numbers.  These are quite different.  The error increased with 
voltage, at least until in the 'charge' range.  In the past, I usually 
only got out the DVM at the end of charge to make su!
re the batteries were all the same.  I have indeed charged to VERY high 
voltages but in most cases, at the time when I started mowing day-to-day, 
I just read the gauge with my marker notes on the face.  Those face notes 
were based on bad data.  I'm sorry.  I still stand by all the numbers on 
charging for cars (something I do know about) and the point at which I 
stopped charging, because I often measure it.  But all my ET OPERATION 
numbers were just TOO HIGH.  That said, going back over my typical charge 
numbers; I end my charge at 42V to 43.5V, (in the white region) as 
measured by a good DVM, on all tractors, and may have never exceeded 45V.  
I have measured over 40V one hour after charge, but probably never 
significantly more.  Probably never that high on the next day.  I normally 
am seeing about 39V a day or more after charge, as measured by a good DVM; 
these last 3 represent different values than I have given in the past.  
These more reliable values should be represented !
on any FAQ.  That said, 38 or 39V at the end of the charge cy!
cle (with the charger still on) is too low to get to the right cell 
voltages except possibly with a perfectly balanced set of new batteries.  
That was and still is my main point.  For more on that, see below.  Based 
on other e-mail, I think most of this is agreeable to Harry by now.  Are 
other gauges different?  Perhaps.  4 Tractors (and one was bad) does not 
make for good statistical data.  Thank you to those who corrected me.  
Comments are welcome.  More data is welcome as well.  I hope any faq is 
based on data and not opinion, or at least, not on just one opinion.

I'm going to find a 2" round Digital/Analog so I can really keep track.

Open for discussion.

Comments, suggestions?

Preliminary FAQS on Meter Voltage (Fuel Meter)
Standard Charger, charging voltage and battery voltage
Meter Current (Load Meter)

The Fuel Meter on the GE, WheelHorse and New Ideal electric tractors are 
similar.

As far as we know, of the units based on a '36V' battery system use the 
same 2" gauge.  The gauge is a Volt-Meter.  It's purpose was to tell the 
user roughly how much charge was still in the battery pack in operation, 
and to give some indication of how if the batteries were taking a charge.  
Unfortunately, after 30 years, many of these gauges have gotten moist on 
the inside and either failed or have had the needles the workings run on 
go bad.  Often they are not operable.  Even worse; some gauges require 
voltages higher than intended to show a given level of charge.  New gauges 
are available from Bill Gunn.  Some people have replaced them with 
standard 2" gauges, but to do so one must understand something of the 
scale of the gauge.  It does NOT start at 0V.  The following is a table of 
needle reading and measured voltage for a gauge in correct working 
condition.  'E' stands for empty, green is normal operating condition, 
with 'F' standing for full and white is charge only ra!
nge.  The numbers are when the needle splits the line between colors.

                                                                
Needle just moves                                       19.8V           
Bottom of red                                           26.7V
Top of red, bottom of green, "E"                33.1V
Top of green, bottom of white, "F"              39.0V   
Top of white, bottom of upper red,              45.5V
Top or upper red                                        47.3V

An earlier table has been posted (by me) with numbers on an old gauge.  At 
the bottom of the gauge, the needle required an extra volt to register.  
In the operation mode, it required an extra 2 volts for the same position. 
 At the bottom of the charge region, the error was 6 volts!  If your gauge 
reads far off the above values, you probably should replace it.

It becomes clear from this table that the term '36V' tractor is a bit 
misleading.  By this chart, 36V is almost 'empty'.  That is not really 
fair, as there is a lot going on in the tractor.  The gauge is was a 
compromise, and one must understand a bit about both the charger, the 
batteries and loads to understand how to interpret it.

First, the battery pack consists of (6) deep discharge wet cells, each 
containing 3 individual cells, all wired in series.  '36V' is a term used 
to describe what it is, but is not a very complete description.  The 
following table gives normal voltages for each cell, battery and for the 
complete pack by capacity or charge level (this is typical for these 
tables, the values below 40% may change depending on manufacturer, at room 
temperature with good batteries)***

Cell    Battery Pack    Charge Level
2.12    6.35            38.1            100%    *
2.08    6.25            37.5            90%
2.07    6.21            37.3            80%
2.05    6.16            37.0            70%
2.03    6.10            36.6            60%
2.01    6.03            36.2            50%
1.98    5.95            35.7    40%
1.96    5.88            35.3            30%
1.93    5.79            34.7            20%
1.89    5.66            33.9            10%
1.75    5.25            31.5            0%      **

*This battery has been fully charged to the design limits at a HIGHER 
voltage and has settled to this point in time.  See below.

**Essentially, although a single battery has some charge in it, it can no 
longer do effective work.  You may get the ET to work below 31.5V total 
with 5 batteries well positive and one battery 'reversed'.  The sum is 
less than 31.5V.  I have indeed seen this.  See below.

The first question people may ask is:  Why the '36V' battery pack is not 
'full' until one reaches more than 38V?  The answer is that each cell has 
been given a nominal value of 2V.  When batteries in cars were 6V, the 
error in that statement was minor.  With 6 batteries, the error is more 
obvious.

The second and more important question is: Why did GE mark the gauge so 
badly to this table?  Well, they weren't really marking to fit the table.  
Nor did they do too badly.  Several things come into play.  The state of 
charge table is talking about room temperature voltage before use, not 
during load or charge.

First, let's talk about the bottom of the gauge.  The fuel gauge would be 
used part of the time under load.  Since a normal ET load is 50 to 100 
Amps (see the load faq) this is enough to lower the apparent voltage at 
the meter under operation.  This is why one can mow until the gauge drops 
to the bottom of the red or even lower, which from the 'charge table' 
above is BELOW 0% of capacity.  It isn't really less than 0%; if the load 
were turned off under these conditions and one were to wait, the voltage 
would return into the lower part of the red range in a few minutes, and in 
a half-hour may even return into the green.  The battery has not 
'recharged' but has indeed recovered a bit.  If at that point it is indeed 
at the bottom of the green, that is only 20% of capacity from the chart 
above, much to low to do much except perhaps drive itself back to a wall 
plug.  Regardless, the meter reading under load is always LOWER than it 
was sitting.

Second, let's talk about the high-end, the charge end of the gauge.  Why 
did GE need a gauge that displayed anything above 38 or 39V?  A typical 
assumption is that if 2.12V per cell is charged, than all one needs to 
provide is 2.12V per cell.  Unfortunately, it isn't that simple.  If you 
apply 2.12V per cell average, and then turn off the charger, you would 
find that the batteries dropped to less than 2V per cell in an hour.  Why? 
 2.12V is the room temperature voltage one can measure some time after 
charging that represents a full charge.  Charging warms up the batteries.  
Additional voltage must be provided to do that.  In addition, each battery 
consists of 3 series connected cells.  There are 6 of these batteries in 
series.  The cells are not identical.  One cell may see or 'take' more 
voltage than another.  The charger must make sure the lowest voltage cell 
ends up with enough voltage to make sure that 2.12V is in that cell some 
time after the charger is turned off.  Wit!
h 36 cells, and with owners who may have used different ages and brands or 
sizes of batteries, and different internal impedances, GE decided to go 
with a bulk-charger capable of bringing voltages significantly above what 
might be needed for perfect batteries.  This is not unusual; most if not 
all bulk chargers at auto parts stores provide 2.5 to well over 2.6 V per 
cell, and these are only concerned with 6 cells, not 18.  Many tables 
suggest 2.58V per cell at room temperature may be required.  Many of these 
chargers go into a low current or 'trickle' mode at the higher voltages.  
Even then, at the beginning of the 'charge' section of the GE fuel gauge, 
the batteries are receiving an average of some 2.2V per cell; at the top 
of the white range we have 2.53V per cell, more than enough to warm them 
and cause them to begin to gas (release water vapor and perhaps even a 
little hydrogen gas) at these levels.  Actually, some bubbling may be good 
in that it mixes the electrolyte, th!
ough one could just bounce the ET around a bit to do this.  T!
oo much gassing is bad.  So is too much heat.  That is why GE included a 
timer to limit the damage one is capable of doing.  There are better 
solutions, today, but this one was inexpensive and does charge the pack 
very quickly.  Speed of charge appears to be GE's main intent.  The GE 
charger's rating is about 20 amps, and since the recommend charge rate for 
flooded batteries is capacity (in amp hours)/10, for 180 to 230 amp hour 
batteries, the GE is just about as fast as you would want to use without 
shortening battery life.  It can draw up to 14A off your house line.  With 
most fuses/breakers set to 15A, it really isn't too bad of a design for 
charging fast.

  Proper questions about charge would be:  "What was the charged voltage?" 
meaning what was the room temperature voltage of the battery pack hours 
after charging but before use, or:  "What was the voltage when you stopped 
charging?"  meaning what was the ending voltage when the ET charger was 
turned off.  These numbers may indeed be about several volts apart.  
Another helpful question might be:  "How long were you charging?" but this 
one implies that the charger was working correctly; most of the ones I 
have used were not until they were repaired; a leaky cap will not allow a 
full charge in any length of time.  Numbers that would not help tell us 
about the level of charge would be:  "What was the voltage when you began 
charging?"


So, where do I stop charging?

Good question.  GE recommended a time system based on the age of the 
batteries.  Some of us wait until the charger produces a voltage at 
roughly the start of the 'charge' section of the gauge.  Other find this 
voltage (39 to over 45V!) to be unsettling.  If one has a leaky capacitor 
on the old GE charger, you will probably be unable to get beyond 38V, 
which at the end of a charge cycle will probably only yield 36V or so (30% 
or less) charge.  That isn't really enough.  A typical charge on a set of 
more than 1 year old batteries will require something of the order of 42+V 
at the END of the charge cycle (with the charger still on) to fully 
balance the cells.  New batteries will be better balanced and require 
less.  Several year old batteries will require more yet.  Don't let the 
voltage go above the charge limit into the red; you will boil the 
batteries dry and damage them.

Balancing cells.

Often when the fuel meter shows 'low', only one or two batteries are 
really low.  Why?  The lift is wired to 18V (3 batteries) and the lights 
to 12V (2 batteries)  Especially if the lift is used a lot (the 
rear-tiller for example) 3 batteries will be lower than the others.  It 
takes extra time to get the voltage high enough to get the charge up in 
all the cells.  Another way of handling this problem is to use a well 
isolated trickle charger on the 'lower' batteries one or 2 at a time.  
Either way, the ET begins to act very badly with a low battery.  Under 
very bad conditions, one can even reverse-charge a battery.  Reverse 
charge describes a condition where the + end of the battery becomes - and 
vise-versa.  Leaving the battery in this state will permanently damage it. 
 The ET charger probably may not bring a reverse charged battery back; a 
good 6V charger may be required.

What about when the ET is not used for long periods?

The ET charger is mostly a Bulk charger.  It dumps in voltage as fast as 
it can.  Absorption (or taper) chargers are a bit smarter; the voltage 
remains fairly constant but the current tapers to a lower lever to prevent 
heat damage.  Most better auto chargers are absorption chargers.  Float 
(often called maintenance or even trickle) chargers bring to voltage to a 
set 'charged' level and then taper back to hold at roughly 100% of the 
charge level above.  Be careful; many so called 'trickle' chargers are 
really absorption chargers.  If you want float, look for float or 
maintenance.  For long term use, a float or maintenance charger is best.  
There are controllers, like that by Harry Landis, which operate by turning 
on the standard ET charger for a fixed time, based on the resting voltage. 
This will dump in a slug of current whenever necessary to keep the pack 
topped up.  This can indeed recover whenever the battery is sensed to be 
low.  You don't want the batteries to freeze, s!
well and crack.  Leaky batteries damage the tractor.

Freeze?
Yep, batteries can freeze.  The following is the GE table:
100% charge     -86F
75%                     -42F
50%                     -16F
25%                     -2F
10%                     +7F

In the north, if you park it for the winter with a full charge, it is 
fairly unlikely you will freeze a battery.  That said, a good charger can 
add years of life to the batteries.

Water level
Check water level in the batteries periodically.  Check each cell.  
Especially if you have been charging to a high voltage to equalize cells.  
Cells lose water, even in normal usage.  The batteries here carry extra 
water over the plates compared to a car battery.  Water must NOT be below 
the plates, ever, if the battery is to be used.  Distilled water is best, 
but de-ionized will work in a pinch.  If the water is low, you will 
probably overheat, warp and short out the plates.  Mixed and charged, the 
electrolyte is acid so be careful.



Tractor Loads

There is no load gauge on an E8M or E10M.  The load gauge is a linear 
current (amp) meter.  An amp meter measures the voltage drop across a 
piece of wire of known resistance (shunt).  The GE is marked by color, not 
current.  The shunt is a 20" piece of #6 gauge copper wire.  The 
resistance of the shunt should be .000658 ohms.  A table of load according 
to the gauge follows.

Bottom of green/0% = 0 mV.                              (Still 0)
Top of green/bottom of yellow = 36 mV           (55 amps)
Top of yellow/bottom of red = 64 mV             (97 amps)
Top of red/100% = 90 mV                         (137 amps)

The table values courtesy Steve Naugler.  I don't guarantee the condition 
of the gauge, but I suspect they are about right.  Since a change of only 
1/5" on the shunt length would show up as up to an amp change, these 
values are probably no better than an amp or so.  They compare well to a 
clip-on hall-effect probe, but it is pretty hard to change load slowly 
enough to put the gauge needle right on a line and keep it there long 
enough to be sure.  They are realistic based on how long 220 amp-hour 
batteries last in the application.  It's also a bit humbling as 760 
mechanical watts (and for a decent motor, 1000 electric watts) only puts 
out a single horse-power.  At 100 amps and 36V we are only doing around 
3600 watts; ~<4 hp of work!  I think that's about right.  Also note, 
normal loads are less than 50 amps, and large loads (snowplow in heavy 
snow) are less than 100 amps.

Larry Elie