Index Home About Blog
From: John De Armond
Subject: Re: Installing second battery
Date: Tue, 21 Sep 1999 01:13:24 EDT

DrSilver wrote:
> I bought an '86 Ford F-250, and I'm fixing it up for fishing expeditions.  I
> put a camper shell on the back and I seem to be accumulating lights and
> other electrical doodads at an alarming rate.
> I'd like to install a second battery to run all this stuff.  Can anyone
> point me towards instructions on how to do this?

Visit your RV dealer (or some car parts places) and get a heavy duty
battery paralleling relay.  This relay looks quite like a starter
relay but has a lower power winding that can be energized
continuously (a starter relay can't.)  Install the new battery and
parallel it to the existing battery through the relay.

The relay should be energized only when the alternator is actually
generating.  The proper way to do this is to pick up the signal (if
it exists) from the alternator or regulator that has voltage on it
when the alternator is generating.  On the Toyota-based camper I
just installed a system on, the regulator had a terminal that had 12
volts on it when the alternator was charging.  Many alternators have
a separate half-wave diode pack that supplies a signal when the
alternator is charging.  A bit of poking around with a voltmeter
should find the right place.

If you can't find a suitable signal from the alternator, you can
install a separate oil pressure switch that will energize the relay
whenever oil pressure is present.  Just tee it in with the existing
oil pressure switch/sender.  

Another possibility if your engine has fuel injection is to pick off
voltage to the fuel pump.  The ECU typically will run the fuel pump
for a few seconds when the ignition key is turned on to build fuel
pressure and then de-energized until the engine is cranked.

The object in this exercise is to make sure the batteries are
paralleled only when the engine is charging.  If they are paralleled
when the engine is stopped (if the relay signal is taken from the
ignition switch), then both batteries are discharged.

I recommend installing a SPDT momentary contact switch that when
pushed, will manually energize the relay while breaking the circuit
from the regulator.  Label this switch "emergency start".  This is
your built-in jumper cables :-) and will allow the accessory battery
to be used to start the engine if you leave your lights on and run
down the truck's battery.

The diode battery splitters that used to be popular do not work well
(at all) with most integral regulator alternators (most everything
modern.)  The reason is that the 0.7 volt drop across the diode
prevents the battery from properly charging.  The old systems worked
because the external regulator took its voltage measurement from the
ignition switch instead of the alternator terminal and automatically
raised the alternator terminal voltage whatever amount was necessary
to bring the ignition switch voltage back to 13.8-14 volts.


From: John De Armond
Subject: Re: Elec connection, truck to trailer
Date: Wed, 10 Nov 1999 00:28:59 EST
Newsgroups: rec.outdoors.rv-travel

budhop wrote:
> Mike:
> Thanks for your input.
> I have a spare isolator from a boat upgrade. I'll look into the suitability
> of using that as an improvement and to cover the possibility of having a
> senior moment and leaving the two systems connected during extended trailer
> use.

A diode type isolator will remove any possibility of charging the
trailer battery because of the 0.7 volt drop across the diode.  The
only way to isolate the trailer system and still charge the battery
is to use a relay isolator.  Most relays I've seen are actuated by
the vehicle's ignition switch.  The batteries are paralleled when
the switch is on and isolated when the switch is off.  This works
but is not optimal because if the ign switch is accidentally left
on, both batteries are discharged. The optimal connection is to
actuate the relay with a signal that the engine is running.  Some
alternator voltage regulators output such a signal.  Another option
is an oil pressure switch.  This is what I used on my MH.  You'll
need the type switch that CLOSES on pressure increase.  I don't
remember what model engine this comes from - I just looked in the
catalog at the parts house until I found one.  A third, slightly
more expensive option is to use a racing RPM switch.  MSD and
Mallory make nice switches that will work.


From: John De Armond
Subject: Re: Elec connection, truck to trailer
Date: Thu, 11 Nov 1999 00:50:34 EST
Newsgroups: rec.outdoors.rv-travel

Don Dickson wrote:
> Neon John wrote:
> >
> >
> > A diode type isolator will remove any possibility of charging the
> > trailer battery because of the 0.7 volt drop across the diode.
> Then a lot of people are being fooled because the most common isolator
> is the diode type. The battery will still charged even with the .7v drop
> because the alternator output is in the 14v range. It just makes it
> harder to get a 100% charge on you trailer battery. 

Sorry, that's not the way it works.  The alternator terminal voltage
is mostly irrelevant (see below) and is a dependent variable.  The
independent variable is the voltage at the regulator reference
terminal.  The alternator voltage simply is whatever voltage it
takes to reach the regulator's setpoint plus the voltage drop
through the B+ wiring. On an old car with small wiring such as my 68
Fury, the alternator terminal voltage is over 17 volts when charging
at near full output.

The issue is where the voltage regulator measures the system
voltage.  All external regulated alternators and some internal
regulated alternators take the system voltage measurement at the
ignition switch (actually the +12v terminal of the regulator but
since there is essentially no current in that lead, it is the same
as at the ignition switch.)  Many modern alternators, particularly
the 1 wire self-excited alternators, take the voltage measurement at
the alternator output terminal.  This is a cost saving design
"feature".  Evidence of this feature is an unusually heavy B+ wire
to the battery, used in an effort to minimize the voltage drop
across this wire.  Internally referenced regulators do a much poorer
job of battery management and battery life suffers as the result. 
My wife's toyota camry is an example.  The reference voltage is 14.4
volts. With a typical complement of accessories operating, the
battery voltage is the proper 13.8-14 volts.  But when the lights
are off and the AC is not running, the battery terminal voltage is
14.4.  This car eats a battery every 2 years or so as a consequence.

Diode isolators will work on alternators with an external voltage
reference.  The reason is the 0.7 volt drop across the diode is seen
as just another voltage drop in the system wiring. The regulator
simply increases the field excitation to the alternator so that the
reference voltage is again 13.8-14 volts.

Diode isolators will NOT work on internally referenced self-excited
alternators.  This includes most foreign cars that use the Bosch
pattern alternator and most late model US designs.  The ONLY
isolator that will work with these alternators is the relay
isolator.  Lacking a factory manual or other listing of suitable
alternators, the easiest method of determining what kind of
alternator one is dealing with is to put a heavy duty diode (such as
a diode isolator) in series with the alternator B+ terminal.  If the
B+ terminal voltage RISES, then the diode isolator will work.  If it
stays about the same, then the regulator is internally referenced
and the diode isolator will NOT work.

It is the path of least resistance and hassle to just use the relay
isolator.  No testing is required and one knows that it will always
work.  it is not unusual for ham radio operators to install a second
battery in their cars to run their rigs from and because of that, I
have been involved in dozens of dual battery installations.  I've
found that the diode isolators are not worth the effort.  The relay
isolator works on ALL vehicles.

Simply sticking a diode in the +12 volt lead going to the trailer,
as I interpreted the recommendation of the original respondent to
be, is a situation different from this discussion.  Such a diode
will introduce the additional voltage drop only in the trailer
wiring and will cause the battery not to charge.  This drop is
different from ohmic voltage drop across the wiring because the 0.7
volt drop is essentially constant regardless of current.  If the
wiring causes a 0.7 volt drop at, say, 30 amps, the drop would be
only 0.35 at 15 amps and zero volts at zero amps.  The charge
current to the trailer battery will be reduced until the drop and
the battery voltage equals the 13.8-14 volt regulator reference. 
The battery will be charged, only more slowly.  Not so with the
fixed diode drop.

>In fact there
> probably is even more resistance(voltage drop) in the wire going to the
> trailer if you don't make sure to get at least one size larger than the
> standard gauge that installers use. The guy who put mine in looked at me
> kind of funny when I told him to go to the heavier wire. It made a whole
> $2.50 greater cost.

True.  However, as I note above, the ohmic drop in the trailer
wiring is distinctly different than the junction drop across the
diode.  Ohmic drop slows the battery charging; junction drop stops

>  The
> > only way to isolate the trailer system and still charge the battery
> > is to use a relay isolator.
> This is a more efficient way to do it. Not the only way. I have a relay
> isolator myself but because it has a moving mechanical part it will at
> some time eventually fail. I usually disconnect the power from the relay
> if I don't intend to use the trailer for several months. That way the
> relay isn't constantly turning on and off for no reason other than the
> ignition is being turned on and off.

But the starter solenoid, a much less robust relay, is being cycled
essentially the same number of times (one time per ignition-on, plus
or minus) and the failure rate of these solenoids is practically
nil.  I carry a relay in my spare parts kit but I've never used it
over the last 15+ years of owning vehicles with 2 batteries (radio
cars and now a MH).  I suspect you put the relay at more of a risk
of damage by pulling the wire off than just letting it operate

Couple other thoughts.

If a diode isolator fails open, a common failure mode, the charging
system to the vehicle battery is also disabled and the alternator is
put at risk.  This is because the vehicle charging current must also
go through the isolator.  If a relay fails open, the most common
failure mode, the only result is the house battery is not charged. 
Relays fail-safe while diodes will commonly fail-dangerous.  Another
reason I prefer the relay.

I mention this only because I've seen three instances of this in the
last year.  A starter solenoid, though it looks like an isolation
relay, is NOT interchangeable.  The isolation relay is designed to
be energized continuously.  The coil typically draws an amp (about
14 watts).  A Ford-type solenoid is designed for intermittent
energization.  The coil typically draws 4+ amps (about 56+ watts). 
The Ford-type will quickly overheat and fail.  A starter solenoid
can be used in a pinch if the coil is powered through an ignition
resistor.  This will drop the voltage to a safer level.


From: John De Armond
Subject: Re: Elec connection, truck to trailer
Date: Fri, 12 Nov 1999 04:04:20 EST
Newsgroups: rec.outdoors.rv-travel wrote:
> Good information John...
> Although you may have had good luck with your solenoids, you may be
> interested in seeing a comparison to a new type of solid state relay
> isolator.  It only uses much less current to turn on.  And, since it
> monitors the voltage, you can even mount it in the camper...
> They have some resistance comparisons here:

Thanks for pointing me to that URL.  After reading a few pages of
this web site, I decided to take an engineering look at this product
and the company's claims.  I think the results are quite interesting
and illustrate why when a small company claims to have invented a
better wheel, they usually have not.  It also shows misleading
advertising using selective and distorted date at its worst.  Pull
up a chair - this will be a long one.

This company very carefully tries to present selective test data
favoring their product with not reporting other relevant data not so
favorable and while not disclosing how the device actually works. 
Nonetheless, with a careful review of all the data presented and
some educated guesses, we can figure out a lot of things.

They claim a very low "on" resistance, too low to be using power
MOS-FETs for switching.  We may then assume that the switching
device is either a bipolar transistor or an insulated gate bipolar
transistor(IGBT).  The claim that the relay requires only 12 ma to
operate indicates that IGBTs are used, as that would be insufficient
drive for even a darlington pair bipolar.  IGBTs are new and "cool"
so I'd expect an outfit like that to be attracted to 'em.

A heavily driven IGBT will act like a resistor at low currents but
will become more of a constant current limiter as the current
approaches the saturation level for the applied gate drive.  This
causes its effective resistance to rise which leads to higher power
dissipation, which will eventually lead to thermal runaway and the
ultimate destruction of the device if allowed to run to completion,
quite unlike a relay.  And when overloaded, the semiconductor will
fail catastrophically.  Again, quite unlike a relay.

Inconsistent specs

On their page, they claim a
voltage drop of "< 0.05 volts at 5 amps".  That implies an internal
resistance of 10 milliohms or less.  We'll assume they're reporting
the lowest figure they can support with data and use 10 milliohms. 
Not bad.  However, elsewhere on that same page, they claim that at
120 amps the 75150 will dissipate "about 36 watts".  This computes
to a resistance of 2.5 milliohms! They show 5 milliohms of
resistance on page in the
graph. Hmmm. Which is it?

Misleading specs. 

The chart and discussion on page is what really got my
attention.  They claim that a mechanical relay's coil draws 1 to 3.5
amps (we've already addressed this) and that the contact resistance
might vary from 1 milliohm (best case, new relay) to 500 milliohm
(worst case, old relay).  This would correspond to from 0.12 volts
to 60 volts drop at 120 amps (the number of amps they quote
repeatedly). What is very interesting is that they charted their
data taken from a test current of only 0.5 amps.  A half amp is
effectively like dry switching for a solenoid relay.  That is, there
is not enough current to burn off surface films on the contacts and
so the apparent resistance will be erratic.  This disappears when
the current rises to several amps.  BUT!  A half an amp would make
their solid state device look artificially good.  Hmmm.  I didn't
like the look of that chart nor the discussion so I decided to test
a relay secured from my spare parts bin.  My Fluke 8800a lab DVM had
just returned from the cal lab so I decided to press it into
service.  this instrument can measure millivolts to 3 decimal

For my first test, I set the relay up on my bench with a power
supply capable of supplying a regulated 5 amps DC.  I configured a
Fluke 88 DVM to measure the current and the 8800A to measure the
voltage drop across the relay contacts using a 4 wire Kelvin
connection which cancels out the test lead resistance.  A separate
power supply supplies coil power to the relay. 

The first measurement was to check the coil current.  At 12.5 volts,
the coil drew 0.7 amps or 8.75 watts.  Significantly lower than
their claim.

For the next measurement, I actuated the relay 15 times in a row and
recorded the current and voltage drop.  I plugged the test values
into an Excel spreadsheet to compute the effective resistance and
power dissipation.  Here is the table of figures:

Test 1:			
Amps	MV dc	Ohms	        Watts	Watts @ 120 amps
5.00	2.968	0.0005936	0.01484	0.356
4.99	115.000	0.0230461	0.57385	13.800
4.94	2.861	0.0005791	0.01413	0.343
4.94	58.540	0.0118502	0.28919	7.025
4.93	1.867	0.0003787	0.00920	0.224
4.93	5.542	0.0011241	0.02732	0.665
4.94	5.586	0.0011308	0.02759	0.670
4.92	4.712	0.0009577	0.02318	0.565
4.92	51.502	0.0104679	0.25339	6.180
4.89	29.400	0.0060123	0.14377	3.528
4.88	26.502	0.0054307	0.12933	3.180
4.88	7.201	0.0014756	0.03514	0.864
4.70	3.310	0.0007043	0.01556	0.397
4.87	5.995	0.0012310	0.02920	0.719
4.87	5.125	0.0010524	0.02496	0.615

I computed the power at 120 amps so that the numbers could easily be
compared to their claim that their device dissipates "about 36
watts" at 120 amps.  These results are quite interesting.  The first
thing we notice is that even at its worst, the relay drops almost no
voltage and dissipates almost no power.  The variable resistance
shows that a) this relay is still on the edge of operating dry due
to the low current and b) this relay's contacts were somewhat dirty
because it had been stored for quite some time.  Note that by the
end of the test, the voltage drop was becoming quite consistent.
Test 2: 
Amps	MV dc	Ohms	        Watts	Watts @ 120 amps
74	0.04082	0.0005516	3.02068	4.898
74	0.04460	0.0006027	3.30040	5.352
74	0.04200	0.0005676	3.10800	5.040
74	0.04820	0.0006514	3.56680	5.784
74	0.09000	0.0012162	6.66000	10.800
74	0.07402	0.0010003	5.47748	8.882
74	0.06024	0.0008141	4.45776	7.229
74	0.04867	0.0006577	3.60158	5.840
74	0.03833	0.0005180	2.83642	4.600

For test 2 I decided to test the relay at a higher and more typical
current level.  I used my DC welder to supply the current and a
Fluke 80i-410 DC clamp-on amp clamp to measure the current.  Here
the figures become more consistent.  The higher drop is due to the
slight heating of the relay contacts which causes their resistance
to become slightly higher.  The power dissipated and the voltage
dropped is still completely insignificant.  This is in stark
contrast to the "bad" numbers presented on their web page.

We've proved claims 2 and 6 on page to be false.  Claim 1 is
valid only if the relay is improperly configured and really has
nothing to do with the relay itself.  Claim 3 is irrelevant as far
as I can tell for RV use.  Claim 4 is of no consequence either way. 
Claim 5, "During extreme cold weather, solenoids or relays can fail
to operate, especially if not environmentally sealed" is hilarious
and an indication of the stretch this company will go to in order to
promote their product.  What makes it so funny is that the starter
uses a relay of essentially identical construction.  If the cold
keeps the relay from working, it will also keep the vehicle from
starting!  And if the relay has frozen, the power dissipated in the
coil will thaw it quite quickly.

Some things they didn't mention:

* 	As they indirectly acknowledge in claim 5, the device has a
positive temperature coefficient of conduction.  What this means is
as the temperature rises, either from ambient temperature or
internal heating, the resistance of the pass transistor increases. 
This causes more heat which causes more heat and so on until the
transistor either fails from thermal runaway or else protective
circuitry operates.  Either way the relay quits working, hopefully
only until it cools.

*	Semiconductors are easily damaged from high voltage.  Something as
simple as having a spark plug wire fall against the wiring and
sparking during service operations might do the trick.

*	Since the device dissipates significant power, it must be mounted
where it can receive cooling air.  With a relay it really doesn't
matter much where it is mounted.

*	Semiconductors are easily damaged by reverse polarity voltage. 
Such as hooking a battery up backward or accidentally reversing the
jumper cables.  Voltage reversal will almost always smoke the diode
pack in the alternator and many times the ECU if the engine has
one.  Nothing like smoking your expensive relay in the same
incident.  Speaking of price,

*	Price. Last time I was at the Shipp's RV sales, I noticed that an
isolation relay costs about $13.  According to their web page, the
smaller device costs


And the larger one,


Wow!  I can buy a lifetime supply of mechanical relays for that kind
of money.

In Summary

Their data is internally inconsistent, test conditions are chosen to
make other products look their worst while hiding the products'
warts and they make irrelevant comparisons while hyping their highly
overpriced product.  Think I'll pass on this one.


From: John De Armond
Subject: Re: Elec connection, truck to trailer
Date: Tue, 16 Nov 1999 01:54:05 EST
Newsgroups: rec.outdoors.rv-travel wrote:

> Although I don't believe it will change your opinion, I must point out
> an error in your calculations.  In your tests, you miscalculated the
> power levels for an equivalent resistance at 120 amps.
> >
> > Amps  MV dc   Ohms            Watts   Watts @ 120 amps
> > 5.00  2.968   0.0005936       0.01484 0.356
> > 4.99  115.000 0.0230461       0.57385 13.800

Ooops,  sorry.  Knew I should have done those calcs by hand and not
relied on the spreadsheet.  

> These should be 120*120*resistance which equals 8.55 Watts and 331
> Watts respectively for the first two above. Your figures appear to be
> result of multiplying the new current by the old voltage drop. Of
> course, this calculation is not an accurate method on a relay for such
> a different current.
> Anyway, it appears to me that your "engineering look" is not entirely
> without bias.  Many of your own claims appear exaggerated to strengthen
> your own position on the subject.

Claim?  I'm not selling anything.  I'm pointing out that irrelevant
claims (see below) are being made to justify a product that costs
over 10 times the product it is trying to replace without showing
any significant benefit and with several major deficits, probably
the most significant of which is the possibility of thermal
runaway.  I'll be interested in seeing if they respond to your
mailing.  I got the distinct feeling of technical obfuscation as I
read their site.  They could present the same product with a lot
less hype using conventional technical terms.

> What seems ironic, is that your test data, even at 5 amps, tends to
> support the erratic resistance data given rather than "prove" false
> statements.  Pass if you like, but I'm glad that I didn't. I still
> think it's a pretty neat product.

Variable contact resistance over the range of interest for a good
relay IS irrelevant to this application.  We're charging a battery,
ferchristsake!  I found comparing their good relay to a bad
mechanical relay to be particularly disingenious.  Now if we compare
a bad mechanical relay to a burned out solid state relay....

Like I said in the oil discussion, whatever makes you feel good is
what is good for you.  For me, I'll buy an extra mechanical relay
and put the $100+ in my pocket for some other toy.


From: John De Armond
Subject: Re: Need help with battery isolator
Date: Wed, 08 Mar 2000 16:09:25 EST
Newsgroups: rec.outdoors.rv-travel

sbourg wrote:
> In article <>, Neon John
> <> wrote:
> >
> >You'll be a LOT happier if you install a relay between the main and aux
> >battery instead of an isolator.  The relay, energized by (preferably)
> >an oil pressure switch or your ignition switch, simply parallels the
> >batteries when the engine is running and separates them when the engine
> >is off.  You don't have to figure out where the alternator's sense
> >terminal is or worry about the voltage drop across the isolator or
> >worry about diodes burning out or any of the other problems that go
> >along with solid state isolators.  The relay you want to use is called
> >a "battery isolator relay" and looks like a round GM-type starter relay
> >except it has a coil rated for continuous duty.  DO NOT use a starter
> >relay - the coil will quickly burn out, as it is not rated for
> >continuous duty.  My local full retail+ RV store charges $13 for the
> >relay.  The only downside, that the relay is mechanical and may
> >occasionally fail, can be addressed by buying a second relay and
> >tossing it in your spare parts box. Still cheaper than an isolator and
> >MUCH less work.
> >
> The relay may be easier to install, but it can allow heavy
> current to flow between batteries when one is in a state of
> greater discharge than the other. This is stress on both the
> batteries and cabling - and could blow a breaker. 

Maybe it could (I don't think so but lacking 100% proof) but in
practice it doesn't.  Some current will flow if one battery is
completely dead but the current isn't enough to hurt anything.  One
can get an idea of the magnitude of this current by observing the
current flow when one jumps off a completely dead vehicle.  A few
minor sparks but certainly nothing that could possibly damage any
component in the system.

> The isolator is
> an elegant solution which absolutely prevents this and allows
> better distribution of charge current where it is needed. The
> diodes are likely to blow only if something is miswired or
> shorted to ground. Else they will outlive the mechanical relay
> contacts by a wide margin with no degradation in performance -
> which is inevitable in a relay.
> The controversy over which method is better is ongoing, but all
> the relay has going for it is cheaper initial cost, and in some
> cases - simpler hookup.

A relay is always a simpler hookup.  The diode isolator is a very
inelegant kluge that seems to have a life of its own.  It wastes
power (a little less than a watt for each amp flowing), it is
relatively fragile (just as fragile as the diodes in the
alternator), is subject to destruction from even a momentary reverse
polarity (such as when one reverses jumper cables) and is difficult
to make work correctly.  It might be acceptable if the charging
system is engineered (and not just kluged together) from scratch to
accommodate the 0.7-1 volt drop across the isolator but it is a pain
in the ass in the aftermarket.  This is 20 years of experience

Old external regulator alternator systems work just fine.  Internal
regulators may or may not and often times, when "not", the problem
is a just barely undercharged battery that seems to go dead every
week or so.  The problem is that some internal regulators STILL
sense the voltage at the alternator terminal even though it has a
sense terminal.  This is probably due to an internal connection that
guarantees the alternator will charge at some rate even if the sense
lead becomes disconnected.  The alternator used on 80's and 90's
vintage Volvos were horrible in this regard.  If the alternator
insists of sensing at its terminal, the charge voltage at the
battery will be 0.8-1 volts low.  The battery(s) will be charged but
only partially.  An indication of how much of a problem this is is
to take a look at the insert that comes with the most popular brand
of aftermarket isolator.  It lists dozens of configurations to try
to work around this problem.

Another problem that rears its head when the alternator is required
to produce near full output over a long period of time, such as when
driving an inverter, is that the isolator gets very hot to the point
that it may burn out.  100 amps at 1 volt drop is 100 watts and
that's a LOT of heat for the skimpy heat sink  that comes on most
isolators.  This is 100 watts of power that is NOT available to the
inverter.  Diodes almost always fail shorted.  When one leg shorts,
the battery on that leg gets all the charging current while the
other oen gets practically none.  This will fool the average joe
into believing the battery is bad.

I fought these damn things for years, first on emergency vehicles,
then on ham radio installations and finally on RVs before it finally
dawned on me that the relay solves all these problems.  The relay is
trivial to install.  simply run a suitable gauge wire from the
existing battery to the relay and thence to the new battery. 
Connect the coil, preferably through an oil pressure switch (so the
relay is only on when the engine is running) or to ignition switched
12 volts. There is effectively no loss through the relay and the
coil typically draws less than 10 watts. 

A MAJOR second benefit is that if a manual energizing switch is
provided, one can parallel the two batteries so that the house
battery can boost the vehicle if one runs down the vehicle battery. 
Or, as happens to me occasionally, one can use the vehicle battery
to boost the generator after the house battery becomes too low to
crank it.  Can't do any of this with a diode isolator.

I suppose mechanical failure is possible but I've certainly never
experienced it.  If I ever do, the extra relay laying in my spare
parts box can be installed in minutes at a cost of about $13. 
Compare that to the cost of a new diode isolator after you smoke it
by accidentally hooking jumpers up backwards or overload it.  To me,
the relay option is a no-brainer.


From: John De Armond
Subject: Re: Need help with battery isolator
Date: Thu, 09 Mar 2000 05:01:22 EST
Newsgroups: rec.outdoors.rv-travel

sbourg wrote:

Since you're in face-saving mode now, I'll only comment on your
errors and let the merits of the devices stand on their own.

> Any externally sensing alternator has absolutely no problem with
> the voltage drop across the diodes - most unregulated alternators
> have hundreds of volts voltage compliance, and the alternator
> output simply goes a bit higher than before to compensate. The
> voltage drop works in your favor by allowing a larger voltage
> drop to the higher current draw side without reducing needed
> charging to the other.

Compliance voltage is not the issue.  What voltage the regulator's
sense lead sees IS the issue.  If the sense terminal sees the
voltage BEFORE the diode isolator, either because of the way it is
wired or because there is an internal resistive or diode link inside
the regulator between the output and sense terminals, then the
charge voltage applied to the batteries WILL BE reduced by the
0.7-1.0 volt drop across the isolator.  I know that this is the case
for some (all?) Bosch internal regulated alternators.  I suspect it
in many others.  I particularly suspect it in nippon-denso and
Hitachi alternators because both of these alternators will regulate
the output voltage even with the sense lead disconnected.

> True, there is a power loss in the isolator - 100 Watts or so
> when the alternator is producing 1500 Watts - much less as the
> battery charge goes up. All isolators I have experienced are
> ruggedly over-designed, so thermal runaway is no problem. I have
> never experienced nor known of anyone to have a problem.

The term "thermal runaway" has no meaning in the context of a
diode.  This is a term that refers to a transistor's increased gain
with temperature causing it to run away and destroy itself unless
externally ballasted.  Good old fashioned thermal failure in the
diodes IS the issue.  Perhaps my definition of "ruggedly
over-designed" is a bit more conservative than yours but I find the
aftermarket units to be way under-heat sunk.  Battery charging isn't
the issue.  Running an inverter or other heavy load which draws
heavy continuous current is.  I'd hate to try to count how many of
these things I've replaced on vehicles where the load is high and

> Certainly, if you jump a battery reverse polarity in any modern
> vehicle, the least of your problems will be the isolator.

Maybe, maybe not.  But not having the isolator on the vehicle is one
less thing to damage.

> Relays, on the other hand, have mechanical points - subject to
> pitting and developing a high-resistivity contact. This can allow
> a series drop of many volts, lengthening recharge times and
> preventing a fully charged state. The first you may realize this
> problem is when your batteries have been so chronically
> undercharged that they prematurely fail.

Rubbish.  You're regurgitating that crap put out by Hellroaring
Technologies here:  This is
the typical crap I'd expect from a company trying to sell a
multi-hundred dollar electronic gizmo to replace a $13 relay. It is
a strawman argument that can't stand even mild scrutiny.

The last time this issue came up, I tested the battery isolation
relays in my possession using a Biddle milli-ohm bridge.  This
bridge passes up to 10 amps through the unknown and reads the
voltage drop (aka Kelvin connection) to measure fractional ohm
devices.  Probably the worst case relay would be the one on my 82
Itasca which is, as best I can tell, the original relay.  It
certainly looks grody enough.  Yeah, the resistance bounced around a
bit but was never over 1 milli-ohm (that would be 0.1 volt drop at
100 amps).  Even when a contact closure resulted in a contact
resistance near 1 milli-ohm, invariably, the next closure was much
less.  The nature of the construction of the isolation relay, with a
loose disk of copper making the connection between fixed studs when
energized, practically guarantees that the relay won't make the
connection at the same point on the disk twice.  Thus, if there is a
bad spot on the disk that causes a relatively high connection, the
next time, a different area makes the contact and the connection is
normal.  The new relays pushed the limits of my bridge, with contact
resistance in the dozens of micro-ohms.  At this range, the
resistance of the studs and how well the Kelvin contacts are made
must be considered.

I should add that when I actually measured the voltage drop across
the relay in my MH, the drop was on the order of a few millivolts
for a 50 amp load.  What happens is the higher current burns through
any oxide or arc residue that may be on the contacts that would
distort any measurement made at lower current.

Next, let's take a look at your claim that a relay may drop "many
volts".  Let's say that "many" equals 2.  That would have 14 volts
on one battery for 12 volts on the other - reasonable.  If I drop 2
volts at 100 amps, that's 200 watts of loss - a level of dissipation
that would quickly render the relay a crispy critter.  Since we
don't see reports of flaming relays, we can safely assume that your
claim is both hyperbole and incredible.

> A relay will cost less, but it is only initially cheaper. Thus it
> is the method of choice by cost-cutting RV manufacturers In RV
> use. It will need replacing eventually and will cost you more in
> battery failures, neither of which is the manufacturer's concern.
> Bad opinions of isolators can always be traced to improper
> application, installation, or simple misunderstanding of how they
> work.

At this point, I'll allow the reader to determine the relative
credibility of either side of this debate.  It's the last I have to
say on the subject.


From: John De Armond
Subject: Re: Need help with battery isolator
Date: Sat, 11 Mar 2000 04:18:14 EST
Newsgroups: rec.outdoors.rv-travel

Wes Stewart wrote:
> John, in one of your previous posts, you mentioned measuring relays
> using Kelvin connections, so I know that you know that you have a
> "forcing" lead and a "sense" lead in that situation.
> The purpose of the sense lead it to monitor the voltage at the junction
> of the forcing current and the device under test (DUT).  In theory, the
> sense lead carries no current and therefore suffers no voltage drop, so
> it correctly measures the voltage at the junction.  This is true in
> power supply circuits as well, where the sense lead is connected to the
> load or to another point at which regulation is to be maintained.
> That said, why would you want to connect the sense lead to anywhere but
> at the battery terminal?  It is the battery voltage that needs to be
> maintained at the correct level to ensure battery longevity.  Using your
> method of connecting somewhere to a load "common" defeats the
> zero-voltage-drop-in-the-sense-lead principle.

From a manufacturer's perspective, I speculate that it is viewed to
be more important to keep the voltage constant at the loads rather
than at the battery.  I imagine that they'd get far more complaints
and warranty claims if the brightness of the lights went up and down
with engine speed along with blower speed.  Not to mention shortened
bulb life.  The battery is pretty tolerant of charge voltage so it
can just sorta come along for the ride.  Again, this is speculation
on my part but it is also the way I'd design things.

BTW, my old Fury's voltage regulator is referenced to some point
near the battery (not sure where, never run it down).  The lights do
run up and down with engine speed and with electrical load. 
Somewhat annoying.  

> In other words, the alternator should be charging the battery; the
> battery should be supplying the loads.

That can't happen, of course, because current can't flow in both
directions at the same time over the same wire.  Under stead state
conditions, the power flows from the alternator to the loads and the
battery is floating off on the side.

> With respect to the basic question of which is better, a diode "isolator" or a
> relay, there are advantages and disadvantage both ways.

I can't argue with that.  It's a shame that people like sbourg must
go to such extremes to defend his position.  Makes the issue much
more polarized that it needs to be.  I'm certainly not 100% against
diode isolators.  Heck, my Datsun Z car still has one in it to run
my radios.  Good application - external regulator, well defined
sense point, room to place the thing in a good stream of cooling air
and a relatively low current draw (60 amp alternator).  If the
diodes fail, they're gone to be replaced with a relay.  I'd
certainly never tell someone to rip out a functioning diode
isolator, especially if it is an OEM installation.  OTOH, I'd never
buy a replacement diode isolator when I can replace it with a more
suitable relay.

Hopefully, THE END of this thread :-)


Index Home About Blog