Boat Design Forums  |  Boat Design Directory  |  Boat Design Gallery  |  Boat Design Book Store  |  Thanks to Our Site Sponsors

 Boat Design Forums Batteries and New Battery Technologies

#1051
10-28-2011, 03:53 AM
 BertKu Senior Member Join Date: May 2009 Rep: 153 Posts: 1,309 Location: South Africa Little Brak River
Quote:
 Originally Posted by brian eiland A new material developed by Japanese company Sumitomo Electric could help allay such fears by potentially improving the capacity of lithium-ion batteries by 1.5 to three times, and therefore extending the range of EVs by an extra 50 to 200 percent. That would give a Nissan LEAF a range of up to 109 to 219 miles (175 to 352 km) or a Tesla Roadster a range of up to 366 to 732 miles (589 to 1,178 km) - enough to assuage the range anxiety of the most fretful drivers.
Although I am reasonable happy with the lithium battery pipe system, the charging of those batteries are a nightmare for me. At the present I made one battery consisting of 40 LiFeP04 batteries, 10 x serial/4 x parallel with Schottky diodes.
The idée was to use some 1000 of those batteries to make 2 battery banks up, each consisting of 500 batteries. What I have now done is to put 3,55 Volt selected 5 watt zener-diodes over each battery of the 10 batteries in serial placed/4 parallel columns . Thus I cannot overcharge a battery no longer during charging. Porta solution of magnets between the batteries works like a dream, but what I have not been able to figure out is, will the batteries be charged 90% or 98% when using zener-diodes. I selected the zener-diodes for temperature at 20 degrees C and for the voltage at 1 Ampere (0.2C charging current)
My algorithm is a) measure 10 batteries , b) measure 3 batteries and c) measure 5 batteries, which will tell me whether I have a bad battery and where about in a pipe. That saves me already lots of balancing wires.

What would be the consequence, if the LiFeP04 Lithium batteries are never fully charged? A lithium battery start charging already at a lower than 3.3 Volt voltage.
The combination of magnets and zener-diodes works like a dream, but before I spent lots of money on additional Lithium batteries for my project, are there any long term consequences. Will the current flowing through the rare earth magnets change the magnetic performance of the magnets ? What about, that the batteries are never fully charged to 3.65 Volt as per manufacturer specifications?.
p.s. the Chinese work with translators, have not been able to get a satisfactory answer from them.

Bert
#1052
10-28-2011, 11:56 AM
 portacruise Senior Member Join Date: Jun 2009 Rep: 205 Posts: 753 Location: USA
Hi, Bert. Some comments below. Sorry for the caps, but I couldn't figure out how to get bold quickly, so you could see my comments easily.

Porta

Quote:
#1053
10-28-2011, 01:55 PM
 BertKu Senior Member Join Date: May 2009 Rep: 153 Posts: 1,309 Location: South Africa Little Brak River
Quote:
 Originally Posted by portacruise Hi, Bert. Some comments below. Sorry for the caps, but I couldn't figure out how to get bold quickly, so you could see my comments easily. Porta
Thank you so much, always appreciated if somebody has solutions for my problems. Porta, what do you think about the use of zener diodes? I don't have to worry about overcharging , nor balancing. I could even only check which half is not comparing with the other half of a series of batteries in one pipe. It would even save me more wires. I need then only 3 wires. Earth, full set of batteries + and then a wire in the middle. If there is an imbalance, I just lift that lot out of the pipe, check and replace a battery with a near fully charged one, without having to switch the running motors off. The bad battery I could then charge in a separate, "one only" battery charger. By using schottky diodes to each pipe, I can allow the current to spread over the other batteries.
Bert
#1054
10-28-2011, 04:29 PM
 portacruise Senior Member Join Date: Jun 2009 Rep: 205 Posts: 753 Location: USA
Don't quite understand your circuit description without a schematic. In theory, if the zeners are doing their job, why would you get a bad battery? It still requires that you check several batteries to find if there is one bad one or several bad, or the string is all proportionally weak due to age. Again temperature is the enemy since all cells will not be at the same exact temperature, some may be over or under charged. I think that is why they are going to the block shape exchangable packs in cordless drills, without one cell projecting into the handle. The one cell in the handle would fail because of the considerable temperature difference, but then the whole pak would have to be discarded...

P.

Quote:
 Originally Posted by BertKu Thank you so much, always appreciated if somebody has solutions for my problems. Porta, what do you think about the use of zener diodes? I don't have to worry about overcharging , nor balancing. I could even only check which half is not comparing with the other half of a series of batteries in one pipe. It would even save me more wires. I need then only 3 wires. Earth, full set of batteries + and then a wire in the middle. If there is an imbalance, I just lift that lot out of the pipe, check and replace a battery with a near fully charged one, without having to switch the running motors off. The bad battery I could then charge in a separate, "one only" battery charger. By using schottky diodes to each pipe, I can allow the current to spread over the other batteries. Bert
#1055
10-29-2011, 02:29 PM
 BertKu Senior Member Join Date: May 2009 Rep: 153 Posts: 1,309 Location: South Africa Little Brak River
Quote:
 Originally Posted by portacruise Don't quite understand your circuit description without a schematic. In theory, if the zeners are doing their job, why would you get a bad battery? It still requires that you check several batteries to find if there is one bad one or several bad, or the string is all proportionally weak due to age. Again temperature is the enemy since all cells will not be at the same exact temperature, some may be over or under charged. I think that is why they are going to the block shape exchangeable packs in cordless drills, without one cell projecting into the handle. The one cell in the handle would fail because of the considerable temperature difference, but then the whole pack would have to be discarded... P.
I had some batteries which quicker reached the 3.65Volt level then the others in the string and as the manufacturer specified 3.65 Volt, I had to find a way to limit the risk of damaging the battery , thus I started to experiment with zener-diodes. I still had a handful from the Dixon factory of Scottsdale, before Siemens took it over, from some 25 years ago, thus I used them.
Porta, I made a decision to have a critical look whether LeFePo4 is a suitable battery for using me to go onto the water. Too many wires, too many this and that. After reading the information of the website, you referred me to, I wonder whether the old fashioned spiral lead acid sealed batteries are not a better option for me. I bought a couple of them and are also experimenting with them. Or alternative to find a good solid larger NiMH battery, to put in series and parallel to make up a large power source.
This time I have to apologize, I made the drawing on a too small piece of paper, thus instead of 10, only 8 batteries are shown in series. The idée is to ensure that if one battery is reaching the cut-off stage, that the zener-diode limit the charging voltage to that battery, before any damage could be done. But I don’t know whether it is absolute necessary to have a constant Voltage to round the charging off. Can’t I just charge with a constant current and then stop, when the Voltage has reached 3.65 Volt (i.e. 90+ %)
Drawing attached.
bert
Attached Thumbnails

#1056
10-30-2011, 04:21 AM
 CDK retired engineer Join Date: Aug 2007 Rep: 1653 Posts: 2,674 Location: Adriatic sea
Bert, I just unpacked my Makita cordless power drill with lithium packs. The old one fell overboard and drowned, it had NiCd or NiMh batteries with a built in thermal switch. Charging was with brute force until the switch opened and a solenoid was released.

The new one comes with a microprocessor controlled charger which does the job in 15 minutes. There are 2 large charging contacts and a 7 prong connector. Each of the 6 cells is monitored and so is the temperature, there is also a blower that starts as soon as a battery is inserted. Only the temperature affects the charging process: it is postponed if the battery is over 40 C (104 F).
If a newly inserted battery is colder than 10 C or the initial voltage is below 12.5 volts the charger doesn't start, if the cell voltages show unbalance the charging stops and a LED signals that the battery is to be discarded.

The toll for charging in 15 minutes is an unexpected complexity in what I thought to be a simple hand tool. I do not quite see how you could implement that in your 10 cell battery tubes because of all the wiring and connectors involved. Each tube would need an 11 pin connector if you were to measure the temperature in a single point, which may not be enough.

The zener diodes are a clever idea to prevent blowing up a bad cell, but I see 2 issues:
How you you pack these in the tube and which current can safely flow when the diodes cannot get rid of the dissipated energy?
__________________
cogito, ergo sum (Descartes' credo)
#1057
10-30-2011, 09:02 AM
 BertKu Senior Member Join Date: May 2009 Rep: 153 Posts: 1,309 Location: South Africa Little Brak River
Quote:
 Originally Posted by CDK Bert, I just unpacked my Makita cordless power drill with lithium packs. The old one fell overboard and drowned, it had NiCd or NiMh batteries with a built in thermal switch. Charging was with brute force until the switch opened and a solenoid was released. The new one comes with a microprocessor controlled charger which does the job in 15 minutes. There are 2 large charging contacts and a 7 prong connector. Each of the 6 cells is monitored and so is the temperature, there is also a blower that starts as soon as a battery is inserted. Only the temperature affects the charging process: it is postponed if the battery is over 40 C (104 F). If a newly inserted battery is colder than 10 C or the initial voltage is below 12.5 volts the charger doesn't start, if the cell voltages show unbalance the charging stops and a LED signals that the battery is to be discarded. The toll for charging in 15 minutes is an unexpected complexity in what I thought to be a simple hand tool. I do not quite see how you could implement that in your 10 cell battery tubes because of all the wiring and connectors involved. Each tube would need an 11 pin connector if you were to measure the temperature in a single point, which may not be enough. The zener diodes are a clever idea to prevent blowing up a bad cell, but I see 2 issues: How you you pack these in the tube and which current can safely flow when the diodes cannot get rid of the dissipated energy?
Thank you for the compliment, I also thought it was a good move. One should look at the concept I have in mind. I cannot play around on the water with lots of little cooling blowers or million of wires measuring temperatures.
1) If batteries can be recharged 1500 times and I use the boat once per week. I will be 140 years old, by the time it will be reaching the 1500 mark applicable. Thus the chemical life time will be my problem. Like Porta cruise said, the battery will dry out.
2) It was planned ultimately to have 100 strings of 10 batteries in series i.e. 33 Volt.
1000 batteries of 15Ah gives me 15 KWh for 2 motors. The motors run at 1 KW, with a peak of 3 KW per motor ( i.e. total HP = 8 HP for the boat.) Absolute maximum with oil cooling 6,5 KW per motor, but I need to increase the battery voltage to 48 Volt.
I haven’t tested it at that Voltage and heat dissipation.
3) The Brushless motors are running on a PWM (Pulse Width Modulation) but at approx 30 %. i.e. at 100 % = 100 Ampere , 30% = 30Ampere x 33 Volt = approx 1 KWper motor. I need for calm water approx 1 KW only, thus 2 motors = 2 KW. That should be fine.
4) Because I would sail with 100 strings, a number of strings would be disconnected and connected to solar charging at 1 Ampere. The balance would only marginal increase my current to the other batteries. Never to such an extent to be worried about heat and therefore having to balance and measure the temperature of the batteries. One worry less.
5) I have thus all the time peak torque for the propeller.
6) The batteries are 5C max and each battery is 1 C = 5 Ampere.
7) Each battery can handle max 25 Ampere, and at 1 C only 5 Ampere, but I would use it at 100 strings / 100 Ampere = 1 Ampere each string. Well below heating up of the batteries.
8) I use IR Schottky diodes 2 x 40 Ampere/100Volt = 80 Ampere. I could use smaller diodes, but they are cheap enough for me. Each string of batteries is via the Schottky diode connected to the plus bar. Therefore the current is spread out over all those Schottky diodes and batteries. Maybe the one will be 1.05 Ampere and the other 0.95 Ampere. Heat is not an issue.
9) I will have a loss of 100 Ampere x 0.4 Volt = 40 watt at 3 KW usage = 1.3% . For me not an issue. (at maximum power usage of 3 KW, cruising speed is estimated at 0.75 - 1 KW)
10) The cabling is red flexible welding cable at 114 mm2 and at 100 Ampere = 1 Ampere/mm2. For me also not an issue.
11) The charging will be done at about 1 ampere per string of 10 batteries, thus heat is not an issue for my batteries. Therefore I don’t need to measure the temperature. But over a period of time, the one battery is charged quicker full then the other.
12) Thus therefore the zener-diodes to protect each battery.
13) The charging is done with 1 ampere constant current, thus if a zener-diode zenert, it stays within the 5 watt allowable heat dissipation. 1 Ampere x 3.6 Volt = 3.6 watt.
14) My only problem is, should a number of zener-diodes get hot. How would it affect the batteries. I haven’t tested it yet.
15) Second worry, if a zener pops, it normally shortens and then the battery gets short circuit. I may have to try to find SMD (surface mounted diodes) which can handle 2 ampere instead of 1 ampere. The battery will not explode by a short circuit, but I may have to use thinner wires to the SMD component.
In conclusion:
I don’t bother about measuring and balancing each battery. Much to complex and clumsy on a boat. With those zenerdiodes, I don’t think its needed in anyway.
I did have 4 wires, which I will reduce to 2 wires, halfway Voltage (16,5 Volt) and full Voltage (33 Volt) and compare all string via a micro with each other. Should a half string be way out, I would disconnect that string on the water, without switching the motors off and check the string and replace a battery. My original prototype microprocessor was with 2, 5 and 10 batteries measuring points. I will drop one wire.
But I am so confused about the Lithium batteries reading all the problems everybody, like yourself have with those batteries, that I may have to resort to simpler batteries like NiMH or NiCad batteries. I have here 40 years old NiCad batteries, they get discharged to zero and charged an half year later and I am back to normal. They can also be charged with the same 1 Ampere constant current charging.
Weight? I have here one NiCad battery made in Mexico D type, it weighs only 70 gram versus the Varta 140 gram. But I cannot get hold of those batteries. Millennium was the make.
My problem is, the number of batteries would then be have to be 3 fold. Not something I look forward to. I am therefore testing my Spiral Sealed Lead Acid Batteries out.
Bert
P.S. CDK I would have to change the concept from round pipes to square and mount the SMD component in that corner space. I would have to have a warning system if more than 2 or 3 batteries are full, or switch it off when a certain charging Voltage is reached. I have to test those theoretical issues.

Last edited by BertKu : 10-30-2011 at 10:44 AM. Reason: Ps. CDK I forgot to answer your concern. and per 2 motors 15KWh
#1058
10-30-2011, 12:44 PM
 CDK retired engineer Join Date: Aug 2007 Rep: 1653 Posts: 2,674 Location: Adriatic sea
Re:15.
Don't worry about the zener across a lithium cell. If it fails, the cell will physically remove the diode by evaporating it.

Re: Mexican D cells.
They weigh only half because there is a C cell inside with a plastic cylinder around it. The positive terminal has an elongated cap to match the D standard. The capacity was 1800 mA/h instead of 4000, although I've seen several with misleading data on the sleeve or without any capacity information at all.
__________________
cogito, ergo sum (Descartes' credo)
#1059
10-30-2011, 02:50 PM
 BertKu Senior Member Join Date: May 2009 Rep: 153 Posts: 1,309 Location: South Africa Little Brak River
Quote:
 Originally Posted by CDK Re:15. Don't worry about the zener across a lithium cell. If it fails, the cell will physically remove the diode by evaporating it.
That is what is worrying me, I rather have a thinner pcb track to the SMD. Let rather melt the track at >10A/mm2. Occasionally it has to be lifted in anyway out to for checking or replacing a battery.

Quote:
 Re: Mexican D cells. They weigh only half because there is a C cell inside with a plastic cylinder around it. The positive terminal has an elongated cap to match the D standard. The capacity was 1800 mA/h instead of 4000, although I've seen several with misleading data on the sleeve or without any capacity information at all.
That is a dirty trick, thanks for telling me. Mine has no data/capacity information on them.
Bert

Last edited by BertKu : 11-03-2011 at 02:17 AM.
#1060
11-16-2011, 09:30 AM
 BertKu Senior Member Join Date: May 2009 Rep: 153 Posts: 1,309 Location: South Africa Little Brak River
New type solar panel

Well, this could make the difference in the future;

http://www.gizmag.com/nanoantennas-f...m_medium=email

My estimate is 26% efficiency or basically the maximum one could get out of a square meter maybe 250 - 260 watt. Time will learn, 5 years, 10 years. ??
Bert

Sorry it was an e-mail here, copyright with gzimag.com

Radio waves are a type of electromagnetic energy, and when they're picked up by traditional metallic antennas, the electrons that are generated can be converted into an electrical current. Given that optical waves are also a type of electromagnetic energy, a team of scientists from Tel Aviv University wondered if these could also be converted into electricity, via an antenna. It turns out that they can - if the antenna is very, very short. These "nanoantennas" could replace the silicon semiconductors in special solar panels, which could harvest more energy from a wider spectrum of sunlight than is currently possible.

The nanoantennas are constructed out of small amounts of aluminum and gold, and are each less than a micron in length - because light has such a short wavelength (as compared to radio waves), short antennas provide the optimal absorption. After being created, the nanoantennas were then exposed to light, to determine how well they could receive and transmit light energy. According to the initial tests, 95 percent of the wattage being absorbed by the antennas was passed along, with only 5 percent being wasted.

Not only are the nanoantennas efficient, but when their length is varied, the wavelength that they can absorb changes. Therefore, the researchers believe that one panel containing a variety of lengths of otherwise-identical nanoantennas could harvest energy from a much broader solar spectrum than is presently allowed by semiconductor technology.

To that end, the Tel Aviv team is now in the process of creating experimental plastic solar panels, nano-imprinted with varying lengths and shapes of nanoantennas. They are also looking into the electromagnetic-energy-to-electrical-current conversion process, with hopes of improving it.

Although silicon is not a particularly expensive material, the scientists believe that the superior efficiency of their panels could allow them to be smaller than present photovoltaic panels, and thus more cost-effective.

Similar research is also under way at the Idaho National Laboratory, where researchers have been developing plastic sheet solar panels stamped with nanoantennas.
#1061
01-02-2012, 02:18 PM
 wolfenzee Junior Member Join Date: Nov 2011 Rep: 10 Posts: 18 Location: Port Townsend, WA (NE corner of Puget Sound)
Quote:
 Originally Posted by BertKu Well, this could make the difference in the future; http://www.gizmag.com/nanoantennas-f...m_medium=email My estimate is 26% efficiency or basically the maximum one could get out of a square meter maybe 250 - 260 watt. Time will learn, 5 years, 10 years. ?? Bert Sorry it was an e-mail here, copyright with gzimag.com Radio waves are a type of electromagnetic energy, and when they're picked up by traditional metallic antennas, the electrons that are generated can be converted into an electrical current. Given that optical waves are also a type of electromagnetic energy, a team of scientists from Tel Aviv University wondered if these could also be converted into electricity, via an antenna. It turns out that they can - if the antenna is very, very short. These "nanoantennas" could replace the silicon semiconductors in special solar panels, which could harvest more energy from a wider spectrum of sunlight than is currently possible. The nanoantennas are constructed out of small amounts of aluminum and gold, and are each less than a micron in length - because light has such a short wavelength (as compared to radio waves), short antennas provide the optimal absorption. After being created, the nanoantennas were then exposed to light, to determine how well they could receive and transmit light energy. According to the initial tests, 95 percent of the wattage being absorbed by the antennas was passed along, with only 5 percent being wasted. Not only are the nanoantennas efficient, but when their length is varied, the wavelength that they can absorb changes. Therefore, the researchers believe that one panel containing a variety of lengths of otherwise-identical nanoantennas could harvest energy from a much broader solar spectrum than is presently allowed by semiconductor technology. To that end, the Tel Aviv team is now in the process of creating experimental plastic solar panels, nano-imprinted with varying lengths and shapes of nanoantennas. They are also looking into the electromagnetic-energy-to-electrical-current conversion process, with hopes of improving it. Although silicon is not a particularly expensive material, the scientists believe that the superior efficiency of their panels could allow them to be smaller than present photovoltaic panels, and thus more cost-effective. Similar research is also under way at the Idaho National Laboratory, where researchers have been developing plastic sheet solar panels stamped with nanoantennas.
This is why I am waiting on solar, wind technology has pretty much gone as far as it can go so first I will get a wind generator, later will add solar.
__________________
Scars remind us of mistakes we have made in the past, It's up to you keep from making them again. (my thumb, transposed by me)
It is better to die living than to live dieing (Tolstoy para phased by Jimmey Buffet)
Those who think they know everything piss off those of use who do.
#1062
01-02-2012, 02:30 PM
 BertKu Senior Member Join Date: May 2009 Rep: 153 Posts: 1,309 Location: South Africa Little Brak River
Thread 1055, we had an idee to make charging of Lithium batteries easier, without lots of wires. However after a number of e-mails with the zener manufacturer, they cannot make a 3.65 Volt power zener diode flat enough to make it a solution. The only solution, which CDK suggested and most likely could work, is having a very small zenerdiode of 400 mW with a sharp bend and a power chip transistor and a resistor. I will one day try it out.
Bert
#1063
01-03-2012, 04:34 AM
 Jeremy Harris Senior Member Join Date: Jun 2009 Rep: 689 Posts: 852 Location: Salisbury, UK
Bert,

Below is the 4 lithium cell charge controller that I use. The charge voltage for each cell is limited by the cell circuit on the left (there is one of these circuits for each cell sub-pack - 4 off for my 8p, 4S battery pack). It can be trimmed to the correct cell cut-off voltage using the 100k trimmer. The cut-off is sharp and once that voltage is reached the TIP105 turns on and shunts current around the cell, preventing it from exceeding the set voltage. These need to be on a heatsink, as they will dissipate a fair bit of heat.

The charge controller as shown is for a four cell battery pack, but it can easily be extended to as many cells as you wanted.

When the first cell in the battery reaches it's charge voltage limit it also sends a signal via an opto isolator to a current limiting circuit. This simply switches a power resistor in series with the battery charger to limit the maximum charge current during the cell balancing phase of charging. The reason for this is to limit the current that could flow through the TIP105 shunt transistors - without this crude current limiter they would get too hot.

During the cell balancing phase the charge rate slows down, as the current gradually tapers off as each cell comes up to voltage. The battery can be left on charge for long periods in this state, as each cell will be clamped at the limiting cell voltage. Full charge is indicated by all the cell limited LEDs being lit, showing that all the cell shunt circuits are active.

This is the circuit on my boat battery pack and it works well.
Attached Thumbnails

#1064
01-03-2012, 05:40 AM
 CDK retired engineer Join Date: Aug 2007 Rep: 1653 Posts: 2,674 Location: Adriatic sea
Nice circuit Jeremy, but what if your battery pack has a 1000 cells?
__________________
cogito, ergo sum (Descartes' credo)
#1065
01-03-2012, 05:56 AM
 Jeremy Harris Senior Member Join Date: Jun 2009 Rep: 689 Posts: 852 Location: Salisbury, UK
If it has 1000 cells in series, then you need 1000 of the circuits on the left. However, a 1000 cell circuit using LiFePO4 cells would be 3,650 V when fully charged, or about 3,200 V nominal, so not very practical on a small boat.

My battery pack has 32 cells, but as they are wired 8P, 4S, I only need 4 shunt circuits for the whole pack.

If the pack was, say, 48 V nominal, using LiFePO4 cells, then it would be around 16 cells groups in series (with probably several paralleled cells in each cell group) and would need 16 cell shunt circuits plus the right hand charge current limit circuit.

 Similar Threads Thread Thread Starter Forum Replies Last Post Batteries in the bilge? Squidly-Diddly Boat Design 5 10-25-2007 08:11 PM Running Three Batteries... JonM81 Electrical Systems 3 10-09-2006 06:57 PM Custom Batteries ??? bjl_sailor Electrical Systems 4 03-05-2005 05:39 PM New Technologies? Wynand N Boat Design 2 02-25-2005 02:56 PM Shipbuilding Technologies 2001 Scott Education 0 08-01-2001 12:40 PM

All times are GMT -4. The time now is 12:09 AM.

 -- Boat Design Net (Gray & Blue) -- Boat Design Net (Blue & White)