Battery Medic - how to ID the REAL vs FAKE

[rojitor]

thanks alot! I got 4 from ebay (cheap! and turns out they have RED PCB! WIN!)
http://www.ebay.co.uk/itm/260753810568?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649
This is where i got them from!

yay! That's the one i ordered

[methods]

Ok - the results are mixed after the overnight discharge.

The Good News:
When I went to bed all the packs had already balanced to 4.1V
So that is good....

The Bad News:
Letting them sit overnight actually *imbalances* the packs due to the quiescent current draw of the units. First they balance perfectly... then the draw to keep the LCD screen etc running runs things out.

4.083V
4.084V
4.083V
4.101V

Now to put it in perspective - that is not that bad of an imbalance. 4.101V - 4.083V = 18mV
That is nothing to be pleased with - but it is certainly a reasonable number. Those of us who have been running lipo for a few years (and abusing lipo...) know that it is pretty common to have a bank of cells out by 100mV or worse - especially after a few deep discharge cycles. If you told me my pack would never be outside of 20mV per bank - I would be totally satisfied with that.

Truth is - most guys don't even have equipment accurate enough to measure that type of difference with any sort of accuracy Many may think they do, but that is just because they have not actually calibrated their equipment against something that is truly accurate to 1mV or less.

The News to make you feel a little better:
To be fair, these were little 4S 5Ah packs. In reality you would be balancing a 10Ah or 15Ah pack so the "imbalance" would be a lot less... most likely in the 10mV range

The options:
A great option here would be to have a trickle supply set to 4.1V/cell.
For instance - with an 18S pack you might have a 73.8V 200mA supply.
This would keep the packs up.
This type of supply is actually quite easy to make and often looked over. Most people would never consider a bulk charger that works at only 100mA or 200ma.... but actually it makes a ton of sense for an application like this. There are many junk supplies out there that can put out 0V to 100V at low current.

At this point my plan is to just hook them up when I get off work and unhook them when I go to bed. That is a solid 5 hours - and if you do this for 2 or 3 days your pack will absolutely come into reasonable balance.

Now....
Lets talk about getting a little more performance out of these....

-methods


P.S. None of my plastic was even close to being melted - though these were 4S packs to be fair. Those with melted plastic - try to measure the shunt power if you can.

[methods]

I am working on ordering 50 more pieces.

Now... I do enjoy making money off of you guys but I think it is more important to further the hobby - so I am not going to try and mask my source for these.
(not like you could not figure it out with a little foot work)

First let me say this: This seller is a really good guy. The English is not so great, but he is honest - and that is worth everything in this game.

Look how awesome this is - he sells two versions

$13.98 shipped (expect 1 month)
http://www.ebay.com/itm/RC-Lipo-Li-Fe-Battery-LCD-Voltage-Meter-Tester-Balancer-/180703582293?pt=LH_DefaultDomain_0&hash=item2a12c5d855#ht_2233wt_1110

$14.99 shipped (expect 1 month)
http://www.ebay.com/itm/LCD-RC-lipo-battery-Voltage-Meter-Balancer-Discharger-/180692663771?pt=LH_DefaultDomain_0&hash=item2a121f3ddb#shId

Yea - that is right
Nearly identical ads - different prices - totally different product.

IF you pay close attention you will realize that if you go to his store the "Good One" is listed under LIPO BATTERIES and the "Bad one" is buried down under "Other"

Let this be a lesson to you boys... if a Chinese vendor is selling what appears to be two identical products and one costs more... dont assume that he thinks you are stupid or that he has made a mistake. This just boils down to communication - and cultural shit - and ways of doing business. In the USA we like to bash inferior products and talk ours up... In China it seems like they just say that everything is "Very good" and it is up to the consumer to be savvy.

So anyway... I am not getting any sort of deal on them. Just trying to order larger amounts to offset faster shipping. Methods does not wait a month for ANYTHING... not even a date with Luke's mom. EMS, DHL, or Bust.

So - there you go.
For the cheap and patient... there is my source. Please treat him with patience and respect - it is very hard to find a good Chinese contact these days. We are working on importing some other stuff. I am preparing a list of the items I commonly import to see what he can do. At the moment I am more than a little pissed with the guys we commonly work with around here (eh HEM!)

-methods

[amberwolf]

Not sure if this is documented somewhere ( I think i recall reading it at oen point.. ) , but anyone got a link to details for the non-engineer on how to install a on/off switch on these so they can remain hooked up to ballance leads without being a parisite ? will cutting the negative most wire in the ballance plug do it ( using a small switch ) ?

I dunno, but when you find it, make sure you start the Wiki article on it (and include as much of all this new info as you can, too. ).

[methods]

Ok - here is the first prototype - double the balance current

IMG_20120210_125443.jpg (109.89 KiB) Viewed 183 times

Factory resistors are 100 ohms and there are 10 of them.
That puts 10ohms across your 4.2V pack for a theoretical 420mA balance current
It is then duty-cycled to 70% to give you an average of 292mA (at 4.2V)

The average power per channel is about 864mW and each of those little resistors sees 86mW

I have some 20 ohm 1W resistors
Two of those in parallel makes 10 ohms
Putting 10 ohms in parallel with the existing 10ohms will double the balance current

Of course we calculate the power....

The new overall average channel power will be 1.73W. The existing resistors will see the same 86mW they were seeing and the new resistors will see about 430mW each. The new resistors are rated at 1W each and typically you cut the rated power in half - so that would be 1/2W... then I have them mounted up on end which will double the cooling surface area.

So we did our due diligence to make sure what we are doing makes sense.
Now lets try it!

But before we do....

6 channels x 1.73W = 10W !!!!!

So we know that is going to equal fail without modifying the case or adding some sort of forced cooling.
Easiest first thing to do is to just modify the cover. Holes... more holes

I suspect this will be too much heat - so I am already working on a second prototype
These are 5W 3.9ohm resistors.
Balance current would be about (4.2V / 3.9ohm) * 0.7 = 750mA

IMG_20120210_123125.jpg (71.73 KiB) Viewed 183 times

These would poke way out of the case and burn the shit out of your hand if you touch them... so we will have to see how that works.

-methods

[methods]

P.S. We need to determine how hot the transistors/fets are getting that are turning the shunts on and off. That is an important piece of the puzzle once we start to go over 2X the current.

-methods

[methods]

Bah....
The surface mount resistors get up to about 150C. That is above where they can reliably run. Data sheet says they should be de-rated to about 5% at that temp That is only 50mW

Ok, on to try the next idea.
I suppose I should go search for all the stuff Gary did with this a few years back... but I am oh so lazy.

-methods

[FastDemise]

He just removed the surface mount stuff then just had a single 3.9ohm in its place. It is what I did and I just modified the case to hold the resistors outside the case for proper space to cool. Works great. I'll post pictures after class. It's nothing pretty but stays dramatically cooler.

[methods]

When you guys say "remove the surface mount resistors" are you guys actually doing all that work or are you just strategically cutting the traces?

Ok - doing the big ceramic mod.

-methods

[FastDemise]

I did all the work and removed them. Be curious if it runs cooler and therefore more accurately as well.

[methods]

Here is another epic fail in the making.

3.9ohm
5W

IMG_20120210_145215.jpg (76.45 KiB) Viewed 160 times

Notice how there is a nice air gap between the resistors and the board. Nearly all of the resistor body is outside the case. These would be secured with a bead of silicone. My V3.0 BMS worked on these - never had a problem

IMG_20120210_145502.jpg (68.11 KiB) Viewed 160 times

They are a just a tiny bit too wide - had to cheese the last one.

IMG_20120210_145524.jpg (83.56 KiB) Viewed 160 times

4.2V / 3.9ohms = ~1A
70% Duty cycle = ~ 750mA

I like three quarter amp shunt current....

.75^2 * 3.9 = ~2.2W

Again - about 50% of the power rating. NEVER run a power resistor anywhere near its power rating without active cooling. Always look up the de-rating chart to see how much power it can run **@ temperature**

The way a power resistor is rated -> they tell you how many watts it can dissipate IF you were to hold it at room temperature (magically). But once it heats up to 100C or 200C its rating goes way down.

hmm... thinking back... I had a huge computer fan blasting these last time I used them

-methods

[methods]

Works good - gets up too about 120C balancing a 10Ah 6S pack down to 4.0V
Yea... those would work... but now I am thinking about some axial resistors outside the case.

I don't really want more than 500mA - anything over that just gets more complicated and hotter than it needs to be. Double the performance is good.

-methods

[amberwolf]

I guess it would take (per channel!) somethign the size of the whole unit (or even the size of a 6s pack!) to balance at the same rate that a good charger could output current at. Like, 1C-2C for the pack.

[methods]

If you wanted to just radiate - yes.
But Fans are amazing... with a fan you can fit 10X the power in the same package.

Working on a radial version right now. Found some 6.8ohm resistors that I probably got for a Gary project back in the day.

-methods

[parabellum]

I don't really want more than 500mA - anything over that just gets more complicated and hotter than it needs to be. Double the performance is good.

How many current can the transistor take?

[methods]

I don't really want more than 500mA - anything over that just gets more complicated and hotter than it needs to be. Double the performance is good.

How many current can the transistor take?

Its a surprise
We will know when the smoke comes out.

-methods

[methods]

Ok - here is prototype number 3

6.8ohm axial resistors
in parallel with stock resistors
4.2ohms total impedance
700mA balance current (after duty cycle)
1W developed on the external resistors

IMG_20120210_165746.jpg (81.37 KiB) Viewed 290 times

IMG_20120210_165812.jpg (58.54 KiB) Viewed 290 times

Good air space in there - though it can only be worse than in the factory configuration.

Give er a test now...

-methods

[bigmoose]

...How many current can the transistor take?

...We will know when the smoke comes out.
-methods

I love the thought of electronics being tortured! "Turn it up," I say... "Turn it way up!!"

[Ypedal]

http://en.wikipedia.org/wiki/Up_to_eleven

[methods]

The Axial resistors resistors seem to work the best so far but they are butt ugly.
They get hot, but not so hot that it takes your skin off. 100C

Thinking about drilling the board so that we can plug them in like thru-hole real fast.

Let me know if anyone has any ideas. I will take them... and make money off of them... with no credit for you

-methods

[amberwolf]

They get hot, but not so hot that it takes your skin off. 100C

So at least I can make my tea with the thing. Just dip it in waterproofing first, then insert into my cup with a teabag, and wait. By the time balancing is done on some really big and badly abused pack, I can sit back and sip away.

[FastDemise]

While testing are you having any issues with the screen getting too hot. I would get black discoloratin from the heat after about a hour.

[methods]

Not at all - but I am keeping mine with the screen facing down so the heat can escape.

If they are set down face up (which is the most intuitive thing to do...) the heat just pools under the LCD.
I am considering drilling holes in the front cover to help let heat out as I suspect most people will set them face up even if we tell them to set it face down.

Still working on ideas over here.
I sure would like to just populate the empty space on the PCB and then improve the efficiency of the cooling design. Even just a moth flapping its wings would make all the difference here....

Awesome would = 250mA - 500mA diode to send the power away in the form of photons
Awesome would = tiny surface mount pager motor with a counterweight that could work the energy away mechanically
Awesome would = 2W fan on every channel - self cooling power dissipation
Awesome would = tiny DC-DC that would take the 4.2V from one channel and funnel it back as 25.2V to the pack

Maybe Poly fuses can play a role here... resistors are linear but the poly fuses would allow more current to flow with better cooling

Maybe we could utilize the duty cycle with tiny little transformers

Ok - back to finding a clean way to dump power the good ol'fashioned american way

-methods

[sn0wchyld]

thanks alot! I got 4 from ebay (cheap! and turns out they have RED PCB! WIN!)
http://www.ebay.co.uk/itm/260753810568?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1439.l2649
This is where i got them from!

yay! That's the one i ordered

I ordered from them too, got the yellow/green model... so no consistency with that seller...


edit

I really love the idea of a dc-dc converter to send the power back to the pack as a whole. how hard would it be to build?

Im using the axial resistors, but I have them mounted at a 90deg angle to the board, so they stand up above the board. Seems to work quite well, though I've always run 'em face down and with a fan blowing over 'em. I think ill build a new case to mount em in with a couple of fans to keep things cool.

[methods]

That seller is probably offering two different models with a small price difference.
I suggest searching his offerings.

-methods