[Jimdear2]

Hi all,

Hope you will all help out here. Our batteries and other stuff will be arriving on the 4th of January.

I'm going to describe the proposed battery pack/monitering system/charging system layout for the tractor and ask you for opinions.

Don't be afraid to lay it on me if you think somethin is a bad in the set up. I know in advance there are potential holes and problems. Some are irrellivant becaust this is a competition machine and will only run for 15 to 20 seconds at a time. Some I will live with because of cost andweight considerations.

I need to find any problems that I have missed in planning. LAY IT ON ME

Initially we will be using 150 38120P Batteries in a 50S/3P layout. Eventually we will be adding another layer if we find we need it and when we can afford it.

Each set of 3 batteries (four later) will be buddy paralleled (similar to what JACKBAUER is doing).

Each buddy pack of three will be monitored by a Cell Log 8M. Eight buddy packs of three batteries to each Cell Log. This should keep us pretty aware of condition of each buddy pack. If we see a odd or out of spec pack we can look for a dud battery. The cell log will also give us warning if we exceed a low voltage we will decide on when we are pulling.

Each buddy pack of three batteries will be charged by a DC/DC converter, 5 volt in and each converter adjusted to 3.65 volts out. The DC to DC converters will be powered by a regulated 5 volt 350 +- 50 amp power supply. (I have found a number of these on ebay).

This way, no battery can be charged above 3.65 volts and if a buddy pack will not come up to spec. we can check the individual cells.

We never expect to do more then two consecutive runs within 15 to 20 minutes. The charger should restore the pack well enough between hooks.

20 inches wide x 6 inches thick x 50 inches long, about 100 pounds, a killer battery. 172 volts (3.45 voltsper cell) and 24ah nominal and up to 600 amps at I hope 2.6 to 3 volts per cell times 50 cells under load. Up to 90 kW. Oh Boy.

Y'all be well and have a great New Year.
Jim

 

[rwaudio]

Which dc/dc converters are you going to use? I'm with you on that setup though! I would do some testing as to the current draw etc when you start to charge from various SOC levels. Otherwise your setup sounds good, I'm sure others will find the holes
Good luck!

 

[Jimdear2]

My partner whom is the electronics wiz has them. All I know right now is they are 5 volts in and adjustible 3 to 4 volts out and I think he said 9 amps. I'll get names and numbers from him tomorrow if I can.

He has been using a jury rig with them to charge our test batteries. Seem to work quite well
Jim

 

[frodus]

here are some 20A vicor's. They need a 48V front end.

http://cgi.ebay.com/VICOR-DC-DC-CONVERTER-48V-3-3V-OUT-75-WATTS-NEW-/300464816889?pt=LH_DefaultDomain_0&hash=item45f51942f9

$10 each ain't bad. Seller might sell for less in higher quantities.

 

[Jimdear2]

here are some 20A vicor's. They need a 48V front end.

http://cgi.ebay.com/VICOR-DC-DC-CONVERTER-48V-3-3V-OUT-75-WATTS-NEW-/300464816889?pt=LH_DefaultDomain_0&hash=item45f51942f9

$10 each ain't bad. Seller might sell for less in higher quantities.

Frodus,

Thanks for the tip.

Darin has located something that run $3.50 in quantity. Just a board so we will need to box them, probably with the panel that will hold the Cell Logs. They are 5 volt input and adjustable output.

I plan on using something like this to power the DC to DC converters.

http://cgi.ebay.com/ACDC-Power-Supply-JF201A-20000-9009-5-Volts-400-Amps-/370376888582?pt=LH_DefaultDomain_0&hash=item563c2ed106.

I've found a bunch of these ranging from $80.00 to $300.00 with outputs up to 450 amps and inputs from 100 to 240 volts, 50 and 60 hz.

Thanks and Happy New Year,
Jim

 

[rwaudio]

here are some 20A vicor's. They need a 48V front end.

http://cgi.ebay.com/VICOR-DC-DC-CONVERTER-48V-3-3V-OUT-75-WATTS-NEW-/300464816889?pt=LH_DefaultDomain_0&hash=item45f51942f9

$10 each ain't bad. Seller might sell for less in higher quantities.

Here's the ones I'm using ($4.95 or less...), and they seem to perform pretty well:
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=380195053685&ssPageName=STRK:MEWAX:IT

I have a temporary setup that works really well, I assembled the pack quite unbalanced to see how it would bring it in line, which it did very nicely.

 

[Jimdear2]

RW,

Your set up works well? Great that is encouraging.

I have a question for you. It looks like you are using aluminium for your connecting buss. Are you having any heating problems.

I was going to order the copper to make up my buss connectors tomorrow, but just the raw material was over $200.00. Aluminium should be quite a bit cheaper. I know aluminium will carry a lot of current and you need to size up at least 50% or more. Do you know of any rule on sizing?

Are you noticing any problems? It looks like you are powering audio AMPs and I know they will pull a LOT of current.

Thanks,
Jim

 

[rwaudio]

RW,

Your set up works well? Great that is encouraging.

I have a question for you. It looks like you are using aluminium for your connecting buss. Are you having any heating problems.

I was going to order the copper to make up my buss connectors tomorrow, but just the raw material was over $200.00. Aluminium should be quite a bit cheaper. I know aluminium will carry a lot of current and you need to size up at least 50% or more. Do you know of any rule on sizing?

Are you noticing any problems? It looks like you are powering audio AMPs and I know they will pull a LOT of current.

Thanks,
Jim

So far I have 1/2 a dozen cycles at different SOC and the dc/dc converters charge things nicely, I'm charging to 3.5V now. I started out at 3.65V but at over about 50% discharge it was drawing huge currents from the dc/dc converters (mine will put out 27-28A at slight overload) however dialing the voltage back a bit put me in a nicer current output even when the cells are quite discharged.

As for the aluminum, I'm using 1/8" and there is zero heat but I can only draw about 100A right now, I did use it to shart my Porsche as well but that was a very short burst. Like you said a 50% oversize isn't bad, the technical equivalent is more like +33ish% but 50% makes the numbers easy and actually increases the current capacity slightly over copper. The weight/size/cost makes it a decent alternative, just be aware of the properties of aluminum oxide. However if you start to compare what some little 90C lipo's are using (10 or 12awg copper) to carry their big currents, 1/8" aluminum starts to look huge!

 

[Yabert]

I have a question for you. It looks like you are using aluminium for your connecting buss. Are you having any heating problems.

For my motorcycle I used 11 Ga (0.091") aluminium bar to pass around 600A peak to the battery without problems. I use NO-OX-ID compound at terminal (contact area).
But my buss bar was large! 0.091" x 2" give a great area.

My understanding is: AWG 00 (2-0) copper wire have a 0.105" square of conductor area.
So 0.125 aluminium bar need to be 1-1/4" tall for have same resistance. (0.105" / 0.125" / 0.66 = 1.27")
And in your case, because you use 3P configuration, 1/8 x 1/2 bar will be electrically good but 1/8 x 3/4 will be structurally better!

 

[Jimdear2]

Yabert,

You may find this information supprising, I sure did.

I do admit that my original method of selection of the copper was a bit of research and a bit of gut. Probably I could get by with a bit smaller stock but Since most of the work this battery will do is full on have no mercy 20 to 25C, I wanted to err on the side of caution.

I went to my metal supplier's catalog and looked up aluminium plate and bars with approx 50% increase in size. I was shocked to see the price of the aluminium was quite a bit higher then the smaller copper stock. To go with 50% bigger aluminium would have cost about 15% more.

Live and learn.

I still say the stuff you build is first quality.

Have a gooooooooooooood New Year,
Jim

 

[rwaudio]

Yabert,

You may find this information supprising, I sure did.

I do admit that my original method of selection of the copper was a bit of research and a bit of gut. Probably I could get by with a bit smaller stock but Since most of the work this battery will do is full on have no mercy 20 to 25C, I wanted to err on the side of caution.

I went to my metal supplier's catalog and looked up aluminium plate and bars with approx 50% increase in size. I was shocked to see the price of the aluminium was quite a bit higher then the smaller copper stock. To go with 50% bigger aluminium would have cost about 15% more.

Live and learn.

I still say the stuff you build is first quality.

Have a gooooooooooooood New Year,
Jim

That's interesting, based on the two metal suppliers that I use copper flat bar is roughly 7-9 times more expensive than the same size aluminum. What size copper/aluminum were you comparing? And if you don't mind where from?

 

[Jimdear2]

So far I have 1/2 a dozen cycles at different SOC and the dc/dc converters charge things nicely, I'm charging to 3.5V now. I started out at 3.65V but at over about 50% discharge it was drawing huge currents from the dc/dc converters (mine will put out 27-28A at slight overload) however dialing the voltage back a bit put me in a nicer current output even when the cells are quite discharged.

RW,
Please consider me a complete idiot when it comes to electronics, please keep that in mind regarding my questions. I probably know enough to just confuse everyone.

Help me if you can.

You say you are using 48 volt in, 3.3 volt out DC to DC converters but you mention charging at 3.65 volt and also 3.5 volts. How? Are you putting 2 units in series and then using resistors? Some other method?

When I look at your picture I can't quite figure what is going on. Can you sketch a diagram?

HERE IS MY PROBLEM

What I want to do is charge 50 individual 3 cell parallel packs with 3.5 volts to each pack. since the parallel packs are also connected in series and because I want to use a common power supply to power them, I'm told that the DC to DC converters MUST be isolated.

The advantage for me for not applying more then 3.5 volts to an individual 3 cell pack is safe charging that does not require monitoring. The 3 cell parallel packs will never recieve more then the DC to DC output of 3.5 volts. The current output can be low because normally the tractor will have at least a few days between competitions.

I have a couple of options open for monitored bulk charging I won't need to go into here.

My main concerns are:
1.) Plugging in the charging system and walking away, then returning to a balanced pack.
2.) Reasonable chance that I will not return to a burnt to the ground tractor and storage unit.
3.) Not being required to spend more then the cost of the pack for a BMS system and LiPoFe charger and then still have to worry about burning to the ground with unattended charging.

I will be monitoring each 3 cell parallel pack with Cell Log 8m units so I can set a high and low limit to shut down charging units but that requires monitoring. I want to be able to plug in the power supply and walk away knowing that no battery is going to get more then the output voltage of the DC to DC unit.

Do you know if the 3.3 volt and 3.5 volt DC to DC units he sells are isolated. I do not know where to go look for this information. Can you help?

 

[Jimdear2]

That's interesting, based on the two metal suppliers that I use copper flat bar is roughly 7-9 times more expensive than the same size aluminum. What size copper/aluminum were you comparing? And if you don't mind where from?

McMasters-Carr is the supplier.

http://www.mcmaster.com/

Multi purpose copper 99.9% pure (Alloy 110)
3/4 inch wide strip, 1/8 in thick and 3 feet long. $14.59 each
2 inch wide strip, 1/8 inch thick and 3 feet long. $34.75 each
1 inch wide strip, 1/4 inch thick and 1 foot long. $11.00 each

Super-Corrosion-Resistant Easy-to-Weld Aluminum (Alloy 5052)
3/4 inch wide strip, 3/16 inch thick and 3 feet long. 21.38 each
2 inch wide strip, 3/16 inch thick and 3 feet long. $45.29 each
1 inch wide strip, 3/8 inch thick and 1 foot long. $14.93 each

Improved-Strength Basic Aluminum (Alloy 3003)
1 inch wide strip, 1/8 inch thick and 3 feet long. 21.38 each (no 3/4 width available in this length)
2 inch wide strip, 3/16 inch thick and 3 feet long. $44.77 each
1 inch wide strip, 3/8 inch thick and 1 foot long. $14.93 each


Total for 1 piece of each in:

Multi purpose copper 99.9% pure (Alloy 110) ............................$60.34
Super Corrosion Resistant Easy to Weld Aluminum (Alloy 5052)....$81.60
Improved Strength Basic Aluminum (Alloy 3003).........................$81.08

I'm sure I could do better, but McMasters-Carr has been my supplier of choice for years.

24/7 access
Super service, usually order to door in less then 36 hours at standard UPS rates
High quality material
They usually have what I need.
Their search functions are easy to use and fast.

So it was copper for me
Jim

 

[toddshotrods]

I love McMaster for the amazing variety of hard to find stuff, but they're kind of high on metal - especially the small sizes. I buy most of my steel and aluminum supplies form Metals Depot.

Aluminum
1/8" x 1" x 4' 6061 - $4.52

3/16" x 2" x 4' 6061 - $11.88

3/8" x 1" x 4' 6061 - $10.56

Online Metals is another good one, but I haven't ordered from them yet. CFreeman54 has - that's where I got the link. They have copper for better prices than McMaster, and you'll notice a big discount for buying longer lengths. I like their site because it has a lot of info on the supplies, like weight per linear foot.

 

[EVfun]

Since we are discussing sources I thought I would toss out a couple more. Metals Depot is a source I hadn't previously heard about, thanks Todd.

I often order from Online Metals. They are local to me so standard shipping is usually ends up being 2 days to door. I can also will call and pick it up in Seattle.

I have ordered from Speedy Metals a number of times. I found this company before finding Online Metals. Since they are half way across the country from me it always takes a few extra days to get my order.

 

[Jimdear2]

Thanks to you all for the links, I have them in my Metals Suppliers folder of my Favorites Bar now.

Copper prices from them all seem about right with McMasters-Carr, Aluminium is quite a bit less. I wonder why?

Jim

 

[rwaudio]

RW,
Please consider me a complete idiot when it comes to electronics, please keep that in mind regarding my questions. I probably know enough to just confuse everyone.

Help me if you can.

You say you are using 48 volt in, 3.3 volt out DC to DC converters but you mention charging at 3.65 volt and also 3.5 volts. How? Are you putting 2 units in series and then using resistors? Some other method?

Many of the dc/dc converters are isolated, but not all. If you have a part number for the dc/dc I would be happy to check it out. Also most dc/dc converters have an adjustable output (or trim) of about +/- 10% and some have a wider range. 3.3V + 10% is 3.63V (3.65 is safe) I dialed it back to 3.5v to stay closer to the rated output current. The voltage is set by the resistor you can see in the picture, when I design the board it will be a 20 turn pot to dial the voltage in exactly.

When I look at your picture I can't quite figure what is going on. Can you sketch a diagram?

I believe my setup is exactly what you are trying to do, I have 3 parallel cells per group, with 4 groups in series (you'll have 50 but same idea) there is a dc/dc converter per group, now the wiring might look a bit odd though. The when you put all of the dc/dc converters in series to charge a series pack, almost all of the current flows through the large wires in the picture, so you are basically building a 175v power supply (50 x 3.5v dc/dc converters in series) the current will flow in one big circle, not 50 small circles. The only time current will flow through the small red wires (in my picture) is when the cells are out of balance. So when you build the charger it's better to use heavy connections between the + and - connections of the dc/dc converters and then run a wire to the cell (use heavy wire, not the small stuff I used) and fuse every wire at the cell end (VERY CLOSE TO THE CELL).

HERE IS MY PROBLEM

What I want to do is charge 50 individual 3 cell parallel packs with 3.5 volts to each pack. since the parallel packs are also connected in series and because I want to use a common power supply to power them, I'm told that the DC to DC converters MUST be isolated.

That is 100% true, if the dc/dc converters are not isolated it won't work and things will blow up.

The advantage for me for not applying more then 3.5 volts to an individual 3 cell pack is safe charging that does not require monitoring. The 3 cell parallel packs will never recieve more then the DC to DC output of 3.5 volts. The current output can be low because normally the tractor will have at least a few days between competitions.

That's the exact reason I'm designing my system, there is minimal chance of overcharging a cell. The chances of a dc/dc converter failing and putting out higher than 3.5v is very slim. There is the chance of a dc/dc converter failing and a group of cells not getting charged, if that happens you will probably destroy them if it isn't caught before you make a run.

I have a couple of options open for monitored bulk charging I won't need to go into here.

My main concerns are:
1.) Plugging in the charging system and walking away, then returning to a balanced pack.
2.) Reasonable chance that I will not return to a burnt to the ground tractor and storage unit.
3.) Not being required to spend more then the cost of the pack for a BMS system and LiPoFe charger and then still have to worry about burning to the ground with unattended charging.

1.) That is the same ideal I have, and providing a dc/dc converter doesn't fail it should be just that simple.
2.) There is minimal chance that a cell could be over charged (which is when they burn)
3.) Also the same reason I'm designing my system, a few grand for a charger, few grand for BMS, and few hundred for gauges isn't what I wanted to do. So my system will cover all of those for much less.

There are two negatives though, you are top balancing, so be careful not to go to 100% discharge or you could damage cells, and if a dc/dc converter fails you could easily destroy the cells that didn't get charged, so you need some way to monitor and confirm all cells got charged. (This could be as simple as an LED that turns on when the dc/dc converter turns on, with a blocking diode so that the cell can't turn it on) I will be doing individual cell voltage monitoring, but I'm thinking that some type of double check might not be a bad thing.

I will be monitoring each 3 cell parallel pack with Cell Log 8m units so I can set a high and low limit to shut down charging units but that requires monitoring. I want to be able to plug in the power supply and walk away knowing that no battery is going to get more then the output voltage of the DC to DC unit.

I'm not convinced that some other system turning off the charger is a good idea. (meaning reliable, and failsafe) Especially when monitoring voltage, if I do go with a system that will turn off the charger it will be based on current draw on the 48v line that powers the dc/dc converters. (too high or too low and it will turn it off) But probably a simple timer or something like that as the failsafe.

Do you know if the 3.3 volt and 3.5 volt DC to DC units he sells are isolated. I do not know where to go look for this information. Can you help?

The dc/dc converter I use is isolated and they make it pretty clear in the description, it's usually not something you have to go digging for. (depends where you are getting them of course) if you give me a part number I'd be happy to check it out for you.


On a side note from your later posts, I've used onlinemetals for years and the service is exceptional. I would recommend them to anyone, I believe speedymetals is a bit cheaper but they aren't so friendly shipping to Canada.

 

[Jimdear2]

RW,

And any others willing to comment,

Attached is a (very) poor partial sketch of the wiring diagram of my proposed combined monitoring/charging system/battery unit.

I want to inclose the 150 batteries, the 50 Dc to Dc units and the power supply wireing to the DC to DC units into a ventilated plastic case. Also there will be wiring for the 7 Cell Log 8m units.

The size of the battery would be 6.5 inches high to include the eventual 4th layer of batteries. It would be 20 inches wide and about 35 inches long with the cooling fans. It should weigh in the area of 110 - 120 pounds in its 3 cell pack version. It is designed to fit between the frame rails and under the seat of the tractor.

As I understand you, you are using the SynQor PQ48033QNA25NKS 48 volt in 3.3 and around 12 amps out converter. This is an isolated unit and through the use of trim resistors (or trim pot for accurate setting) allows the output voltage to be adjusted up to 10%+-.

What information I have been able to find on this board suggests that the input voltage can be from 36 to 72 volts DC. Could you confirm that?

I believe that is exactly what I want to build the battery system I show in the sketch. Do you agree.

If the above is true then I am golden and all of my wishes in the previous post are covered.

Thanks a HUGE bunch for your help,
Jim

 

[rwaudio]

All of the above is correct, the dc/dc converters I'm using will put out 25A, voltage and all of that is correct.

From the sketch, I would change:
tie dc/dc converter + to the next converter - and use a single wire to run to the cell, same with the monitoring side, you are using 2 wires for every connection, the + of cell 1 is connected to - of cell 2, no need to use two wires they are already connected. In my system I will be using one wire (per cell) for monitoring/charging there is no need to add more spaghetti. Keep it as clean and simple as possible and it should work for a long time.

 

[Jimdear2]

All of the above is correct, the dc/dc converters I'm using will put out 25A, voltage and all of that is correct.

From the sketch, I would change:
tie dc/dc converter + to the next converter - and use a single wire to run to the cell, same with the monitoring side, you are using 2 wires for every connection, the + of cell 1 is connected to - of cell 2, no need to use two wires they are already connected. In my system I will be using one wire (per cell) for monitoring/charging there is no need to add more spaghetti. Keep it as clean and simple as possible and it should work for a long time.

RW,

Thanks, I had already planned on the one wire setup I just drew it this way for clarity.

That is as clear as my sketches are, does the phrase "Clear as Mud" come to mind.

Jim