Topic: Lithium Ion Battery Tests and question about running 798 at low voltage

[johnsondel]

I recently use a set of 3 Lithium ion batteries to power the 798.  Each battery outputs 4.2 V when fully charged and has a internal circuit that protects the battery from discharging below 3 volts.  The batteries are rated to deliver 2.6 Ah each.
When connected in series they can deliver 12.6 V to 9 V output as they go from fully charged to the discharged cutoff voltage.

I was able to run the HB in simulation mode for 2 hours.  The batteries were still supplying current but the HB was showing 9.8 V which is below the 10V minimum voltage that the HB is specified to operate on.  I was concerned about operating the HB below it's minimum voltage so I shut it down.  Is there a problem operating the HB below it's specified minimum voltage?

When I tested the batteries after shutting down the HB they had voltages ranging from 3.57 to 3.59.  Since these batteries can deliver power until their voltages reach 3.0 V it might be possible to get another hour our more out of them by using a switching regulator that would boost the voltage to 10 V.  if the HB can safely function below 10 V one might gain additional time by letting it function until the batteries cut out at 9 V or when the HB shuts down.

Another option is to add a 4th battery in series.  This would give an voltage operating range from the batteries of 16.8 to 12 V.  The HB would then operate until the batteries cut out when they dropped below their minimum operating voltage.  This would be practical if the HB uses switching regulators to step down the supply voltage to the internal voltage(s) used by it's circuits.  If it uses linear voltage regulators the excess voltage is to a large extent wasted.  My second question is 'Which type of voltage regulator does Humminbird use in the 798'.

Once I determine the best approach to maximizing the power of these rechargeable lithium ion batteries I will then mix parallel and series connected combinations to boost the battery pack output to be comparable to the Gel battery that HB sells for ice fishing units (7 ah).  As an example, one could take 9 batteries and create 3 packs of 3 cells connected in series and then connect these packs in parallel to achieve a voltage range of 12.6 V to 9 V with an output rating of 7.8 Ah.

The batteries used for this test were 18650 UltraFire Lithium Ion Rechargeable.  They cost about 4.70 each (including free shipping from Hong Kong) and should be able to be recharged 500 to 1000 times.  A battery pack setup using 9 batteries would weigh under 3 lbs compared to the 6 lbs of the 12 V gel battery sold by HB.  Note that the test that I conducted did not have the transducer connected.  However, reports by other HB owner say that the biggest drain on the battery is the brightness settings of the unit.  Which running in simulation mode the unit's brightness setting is at maximum.

Del

[johnsondel]

I'm re-running the tests with 4 batteries in series.  I tested the voltage at the output of the battery holder.  It shows a reading 0.34 V higher than the reading on the HB.  That is consistent with other posting that were seeing a voltage drop of around 0.4 volts.  This test should give me some idea of how efficient the voltage regulator(s) are inside of the humminbird and whether the unit is wasting some of the extra power that was added with the 4th battery.  I will try to post the results this evening.

[ITGEEK]

Wouldn't it be simpler to just use a small 12 volt marine battery?
Less connections and less to go wrong.
Also made for harsh environments.
You will have ample power to spare if only used for the sonar.

They do make 12 volt lithium ion batteries now.
They are however, very expensive.

Is the goal to spend as least as possible and still have something that works?
Will you put these in a corrosion resistant container?

[LocDown]

couldn't you just get a small atv or riding mower battery and be done with it?

[kron]

I guess that the benefits of Li -Ion batteries is less weight and led/ acid free which makes you independent when tilting the batteries. There is also acid free batteries on the market. However they are more expensive.

I don't know, but guess that LI-Ion is like Li-Po and pretty dependent on the temperature. Here in Sweden, this is a pretty common problem in winter when temperature is below 0 degrees Celsius (32 degrees Fahrenheit). Under normal temperatures (guess 20 Celsius, 68 Fahrenheit) the batteries will not perform so well.

[LocDown]

If it's just for the fish finder, how long are you planning on being out before you can charge back up? Sounds like you're making it more complicated than it's worth. You could probably just wire 2 6v lantern batteries in series and be good for awhile.

[xSilmarilSx]

Ok, here is my take at the same problem...

I also had been using Li-Po 3 cell batteries for my 778C. My main use is weight saving on my kayak, or when using my sonar in a portable mode for a day or less.

I regret not to choose a 4 cell bat. But the preliminary info that I got was not consistent. 

My bat is a 5Ah 3cell Li-Po made for RC Helicopters. The biggest argument is weight saving. The second one, is that I can recharge the battery using any 12V Lead-Acid battery, or 120V. And since the recharge rate of these li-po packs are 1C, you can recharge the pack quickly in around a hour and a half. And you can pump more than half the battery capacity in about 30 min. This can't be done with lead-acid chemistry.

Biggest drawback, expensive if you don't already have a charging solution. Very picky about their end of life voltage (3.7V/cell) NOT 3V/cell... Low total capacity ( around 6Ah for already built packs).

To answers certain OP questions:

Your 798 use a switching regulator, so your 4 cell batt will be correct and not "wasted" by a linear regulator.
Your main problem will be your 16850 bats. You MUST charge each lithium cell separately. So even if you build series/parallel packs, you will need to charge each cell at a time ( time consuming).
If you really want to go this route, you are better to buy pre-made 4 cell packs on HobbyKing website and a solid(quality) charger for them.
And yes, it will be more expensive than lead-acid. And don't run your cells below 3.7V which is around 11V on your 798.

[xSilmarilSx]

For your information , your sonar should draw around 800 mAh when powered, sonar working, and full brightness...

Since you use only 80% of total capacity of a battery pack, the math should be around 1Ah of "theorical" consumption for your battery pack. So 7.8Ah will do around 7.8H at 20°C

[johnsondel]

Thanks xSilmarilSx for adding your comments on your experience with Lithiums.  I didn't know that 3.7 v is the lower limit for safely discharging the lithium batteries.  I knew that if you go below 3.0 you must apply a trickle charge at the beginning of the recharging process and that the batteries should not be recharged below, if I recall, 2.4 V.

The reasons for looking at Lithiums are:
 
1. Weight.  Don't have a snowmobile and some of the lakes that we go into are closed to motorized vehicles.  We have to walk 1 mile or more with sleds to get into some of the best lakes.
2.  The small lead acid batteries don' t seem to last very long.
3.  I like to work with electronic circuits and microprocessors so I wanted to see if new technology like lithiums will work better in a number of applications where I have used other types of rechargeables.

I don't think the HB uses switched voltage regulators.  My first test with 3 batteries in series resulted in 2 hours of operation using simulation mode.  When the voltage reported on the HB reached 9.8 V I shut it down.  As previously reported the batteries tested out at 3.57 to 3.59 V.  I figured that if I added an additional battery, the batteries would reach their cutoff voltage of 3.0 and shut down before the HB reached its minimum voltage of 10 V thereby giving me more up time.  It turned out that I got less then 1.5 hours of up time.  I had figured that I could run an errand and get back in time to record the end time when the batteries reached their cut off voltage.  I thought I would get more up time then I did using 3 batteries.  Such was not the case. 

This morning I performed another test but I ran the HB in normal mode - no transducer - full brightness.  I got 2.5 hours of up time before the HB reported 10 V.  Every minute I recorded the voltages reported by the HB and the voltage at the terminals of the battery.  The surprising find was that when the batteries were fully charged and supplying voltages of 10.5 V or more the average rate of voltage drop was 14.4mv per minute.   At 10.4 and below (down to 10 V as reported by the HB unit) the voltage drop averaged 4.7 mv per minute.  This would suggest that the unit is 3 times more efficient operating near at or near the lower operating voltage limit.  Both this test and the test when the 4th battery was added suggest that the HB is using linear regulator(s) that dissapates the extra voltage by converting it to heat energy. 

One other statistic.  During the last test the HB operated for 2.5 hours or 150 minutes.  During that time 57 minutes were spent running in the optimum voltage range of 10.4 and below.  If you could stay withing this range (10.4  to 10.0) all the time using these three lithium batteries you could nearly triple the up time of the unit.  My next attempt will be to build a switching regulator that keeps the voltage in this range using 2 or 3 batteries.  Switching regulators are about 90% efficient so if I can double ( 5 hours ) of up time I would consider this to be the ideal light weight fast rechargeable power source for the unit.  Note that 3 batteries with the battery holder only weighs 1 lb.

Del

[xSilmarilSx]

The thing that flaw your calculation, is that you base all your readings from the battery voltage, and you suspect that the discharge curve of your lithium batteries are linear.

You must meter the battery voltage AND the power consumption (amps) directly from the battery pack.
Without amperage measures, you can only guess what power drain the 798 is.

And the thing with the switching regulator, is an info that I derive directly from Humminbird tech support.
And since the units are able to function between 10-20V at a current draw around 800mAh, using a linear regulator will heat the unit pretty quickly at 14V (charging voltage) if the unit only need 9V for safe operation. (since you can't UP voltage with a linear regulator.)
So if you discharge 14V-9V (5V) at 800mAh inside a linear regulator, you loose 4W of power inside the unit.
I've drawn the power consumption curve for my 778C form 0 - 10 brightness setting, and I can assure you that particular unit can't have linear regulation since max power consumption at the battery is around 8-9W, if you account a theoretical draw of 4W for a linear regulator, it's impossible for that unit to survive with only 5W on real consumption.

If you can find a regulated variable power supply that monitor volts and amps, you will be able to better map the power consumption of your particular unit in different power settings...

All I can say, is don't run your unit a max brightness.. around 7 is the best combo for visibility/power consumption. The biggest drain is the Backlit. and the Black and white unit are champs in that department! (low consumption because thay don't require backlit for daytime viewing)

[johnsondel]

It still doesn't explain why the batteries discharged to under 3.0 V in less than 1.5 hours on the second test when I added an additional battery.  The first test ran for 2 hours and was shut down when the HB displayed 9.8 V.  The voltage on these batteries averaged 3.58 V after the first test.  Both the first and second tests were run in siimulation mode.  By adding the 4th battery in test 2 an additional 33% of power was added to the supply which the should have been efficiently used by a switched voltage regulator to extend the up time of the HB.  Also, additional power was consumed by letting the batteries drain to less than 3.0 ( actual measured voltage was 2.95 V).  All 4 batteries used were new, fully charged and tested to be at 4.24 V prior to running these tests.

Also Lithium ion batteries have a relatively flat curve
(see: http://www.ibt-power.com/Battery_packs/Li_Ion/Lithium_ion_tech.html )
The curve shows a more rapid voltage loss at the high and low voltage ends.  On the third test, where I measured voltages over time, I had the slowest rate of voltage decline at the end of the test.  All three batteries measured 3.55V after being removed from the battery holder at the end of the test.  If anything, the voltage drop should have been more rapid at the end of the test if these batteries followed a simllar power curve to the one on the website referenced above. 

I agree with you that this is not an accurate test if power consumption is not measured (requiring metering of the amps and volts together).  I didn't have the equipment to do so.  If I can get a printed circuit board from Analog Devices I will put together a switched regulator that will boost a 6V to 8.8V supply to 10.4V and perform the 3rd test over again to see if (1) I get more up-time and (2) see if the voltage drop per minute stays more even.  I used there software to design a switched boost voltage regulator that, theoretically is > 90% efficient.

The bottom line is that it is rather amazing that I can run for 2.5 hours on 3 small batteries.  I wasn't running a transducer but I was running with full brightness.  If you decrease brightness 50% and add the transducer I suspect that the up-time will remain the same or be even better (just a guess).   If the addition of a switched voltage regulater doesn't improve the circuit I will probably go with two battery packs of three batteries each (3 batteries connected in series) which will double the Ah and should give me 5 hours of up time.

One other item.  You mentioned that 3.7 V is the end life of the battery that I'm using.  Everything that I have read talks about 3 V.   If I group three batteries in series times n number of series wired groups and shut down the HB when it reports 10 V supply the lithium batteries should remain above 3.5 volts base on test 1 and test 3.  I know the batteries that I'm using are rated at 3.7 V but does that really mean that you shouldn't go below that value?

[sonar2000]

Question. Since voltage is a potential and current (amps) does the work..... are we measuring the right piece.
Just curious.....
Chuck

[Bob B]

I also think it is important that you be measuring the current draw.  It will take the guesswork out of your measurements.

At my local Harbor Freight I could by a volt/amp meter that would do this for about $10....much less than you are spending trying to do this the hard way.

[xSilmarilSx]

Till the OP can come with some current draw measurements, it will never truly know what is the power consumption of is setup. But is methodology for it's experiment is quite good!


For the question about 3.0V and 3.7V, you can see from your discharge graphs (link) that pushing your cells below 3.7V is detrimental to battery life and not giving you really more usable time (Remember to never discharge cells below 80% total capacity.)

And you posted that your cell voltage was 4.24V which is not good. You should never go higher than 4.20V
Your charger can be wrong, or your voltmeter. Just remember, when working with lithium chemistry charging, you MUST use quality charger, or else the cell can catch fire real bad.

For your information, I checked on HobbyKing website and they sells 8Ah, 3cell Li-po cells for around 50$
Which I think is a better solution for your project than multiple 18650 cells.

Oh, and don't trust the voltage shown on your HB to do your consumption math!

The power draw of the transducer, when used between 0-100ft of water is around 1W

[Bob B]

I probably should have stayed out of this, but one of my pet peve's is when something is "over engineered."

I have found that when something gets too complicated there are too many places for things to go wrong....step back and look for the simple effective solution...in this case....check the amp draw.

Sorry for the diversion.

[johnsondel]

1st sentence - I agree.  I have a power supply that reports analog Voltage and Amps but the amp labeling has come unglued and  causes the needle to hang.  Even so, it is an analog meter.  I might try to take it apart and fix it but it is still a crude reading. 

2nd sentence - I agree.  According to the charts that I referenced in the last note the voltage drop per minute drops off quickly starting at 3.6 V.  I actually measured down to 9.9 V, as displayed on the HB, before I shut down the HB.  I captured 5 extra minutes of data while the unit was at 9.9 V.  The voltage drop rate, under 10 V,  was 10 mv/minute  for those 5 measurements suggesting that it was starting an accelerated downward rate in voltage drop. As I stated before the cells measured 3.55 V after they were removed from the battery holders.  This is about where the chart, previously reference, ceased to be linear.

3rd sentence - I'm using a charger made by the same company that makes the Lithium ion battery.  It fully charged cells are only 0.24 V over the 4.20 limit.  As you say, it could an error in my meter.  Usually the chips that are used in these devices are quite accurate.

4th sentence - Thanks for the information on Hobby King.  The cells that they use do have higher aH ratings.  The cell packs that I found on the site at that rating are 3 cells in series.  Using 18650 cells would require 9 cells to be in the ball park.  The problem with battery packs is that you are charging, in this case, 3 cells in series.  You could run into problems if one of the cells deteriorates before the others.  By charging each cell individually the bad cell will be detected by a good charger and the recharge process would be stopped before there is a fire.  I would be paying 9 time $4.70 to get 7.8 aH using the 18650 cells.  I would rather charge 3 cells instead of 9 if they are capable of producing 8 aH / cell.  Do you know what the model number of the individual cells used in your battery packs?

5th Sentence: There is about a 0.25 to 0.3 V  difference between what I measured on the battery holder contacts and what was displayed on the HB monitor.  There is 6 inches of fine wire from the battery contacts to the HB standard cable connected via spad connectors.  At 9.9 V indicated on the HB monitor the battery contacts voltage showed 10.18 V.  This difference might be reasonable given the expected voltage drop in the cable and connectors.  There is a bigger discrepency in the voltage measued across the battery under load and no-load conditions.  When the tests were over (tests were stopped when HB readings reached 9.9 V)  the batteries with no load showed 3.55 V or 10.65 V connected in series.  However as the end of the test, under load, the reading across the battery holder read 10.18 V.  This difference might be accounted for when you consider the affect of battery resistence under load vs no load conditions.  The HB voltage is probably reasonably accurate give the supply line losses, the voltage drop from internal battery's resistance and the one decimal digit accuracy reported on the HB monitor.

6th Sentence.  The transducer load would represent about 0.1 aH (100 mAh) current draw or about 1/8 of the total load if you use the 800 aH total load presented by the unit that you previously mentioned.  This ratio of course depends on your settings, especially on the brightness seting but also whether the gps navigation options are set on. 

[xSilmarilSx]

The problem with battery packs is that you are charging, in this case, 3 cells in series.  You could run into problems if one of the cells deteriorates before the others.  By charging each cell individually the bad cell will be detected by a good charger and the recharge process would be stopped before there is a fire

This section is wrongly interpreted. Since these packs are 3 cells, they feature a charging port which is connected in parallel for each cell, so that the charging occur on each cells separately. The output connector is connected in series to deliver the total voltage of 3 cells..

So, with the correct charger, you charge each cell at the same time, safely.
And some powerful chargers can put 5 Ah in each cell to reduce the charging time.

[xSilmarilSx]

Here's the link for the 8Ah battery pack.

http://www.hobbyking.com/hobbyking/store/__20845__ZIPPY_Flightmax_8000mAh_3S1P_30C_USA_Warehouse_.html

[johnsondel]

The power output does appear to be the same.  I got my amp meter working on my power supply.  I took two measurements:

0.5 Amp at 15 V output
0.75 Amp at 10 V output    Volts * Amps = 7.5 Watts for both readings.

I was viewing the system screen in normal mode w/o the transponder.

xSilmarilSx, you can tell me 'I told you so'.  It does look like they are using switching regulators. 

These are analog readings from the cheap panel meters on the power supply.

[xSilmarilSx]

Even if your readings are analog, at least now you can really test your system fully.

And yes, I told you so! I already done all you are trying to do!
But It's ok to try your own setup, I like it.

Now, with theses new measurements you can proceed further and continue your test
and obtain what your after.

Good luck, and if you have other questions, keep asking, we are here to help!