A simple and fool proof discharger/tester

auraslip

10 MW
Joined
Mar 5, 2010
Messages
3,535
For my new build I'm going to be using 12 6s5ah hk lipo packs. Since there has been some really fun fires because of defects, I've decided that I need to do a couple testing cycles before I consider the batteries safe. Rather than use my imax charger which is dreadfully slow at discharging and takes the cells lower than I'd like, I've decided to roll my own.



This is the basic idea. Six 12v 50w mr16 bulbs in any series/parallel config I need. (easily accomplished with sexless connects like andersons!)

The amp meter will be replaced with a watts up meter for testing purposes. The cell log will be used with a relay to end the discharge cycle. This is the fool proof part!

Right now I'm having trouble figuring out how to mount the bulbs. I tried hot glue, but it melted. Epoxy would work, but I'd like to be able to replace the bulbs when they burn out.



I'm thinking it'd be much cooler to mount the bulbs facing down, so that it's easier to reconfig them and so light polluting isn't a problem. However, this poses a fire/overheating risk to what ever surface the discharger sits on. Two 12v fans in series would fit perfectly at the end though. I wonder if this violation of KISS is worth the effort?
 
Have you seen the suspended 12v lighting "strings they have in some stores ??
Just string a pair of wires 3" apart across and hang the lamp connectors from them.
spotlight-cable-lighting-kits.jpg
 
I like your setup. I put together something similar using an Arduino to display and log charge and discharge data for larger (100Ahr) LiFePO4 cells. I initially used a bunch of light bulbs kluged together but was not able to get a big enough load. So instead, for a load, I coiled some 14 awg Romex around a 4" section of abs pipe and submerged that in water. I added some ice to keep the current draw high as the resistance of the coil increased with temperature. This gave me a relatively constant current load (65A) that was useful in measuring the capacity of the 4-cell modules. I also used a celllog to terminate the charge manually - never got around to installing a relay.
 
Nice, this is the same route i went actually. You should see about 8-9 amps of draw, which will drain a 5ah pack in around 30 mins.

I found that they would get so hot that they'd burn through everything. I ended up hanging mine off a pole facing downward.
The floor didn't get hot, which was a surprise.

Great winter heating and lighting :mrgreen:
 
I still prefer the one I posted a couple places previously, about "Danger- High Voltage. Not only will this kill you, but it will hurt the whole time you are dying"
or something like that. :)
 
auraslip" I tried hot glue said:
No? :shock: really ...would never have guessed :lol:

Seriously..how about a sheet metal plate instead of wood? Maybe something like a 6mm sheet of checker plate aluminium, ...
 


$3 worth of l bracket and hardware and 30 minutes and I made a lil holder.
The bottom of the bulbs are totally flat, so I hot glued some heat shrink to the aluminum l brackets to get more grip.
Should be fun to watch the heatshrink get hot and shrink and the hot glue melt :?

Now I just need to figure out the wiring for the cell log and relay. Bulbs are on their way!
 
If I undetstand correctly, the cell log will energise a relay that disconnects the batteries from the load ? correct

Does the cell log have enough power to run the relay itself? or is the alarm output just an opto relay that makes a circuit when an alarm condition triggers it?

If you need extetnal power to run the relay, how are you doing that?
 
Neil - I'vestarted a second post trying to figure out the details for how I want to do it - but yes you can run a 24v relay off the alarm port -



The idea in this circuit is that the cell log alarm port is normally closed until the alarm triggers (it can be set either way in the software settings) while closed it activates the opto coupler (the part necessary to chain together multiple units since the alarm port is not isolated) the opto coupler activates the transistor which operates the normally open relay allowing the battery to be connected to the load. When the LVC alarm triggers, the alarm opens and the relay opens disconnecting the battery from the load.

I want to use the optocouplers and transistors because it'll allow me to chain as many cell logs as I want together - with just a few mods to the circuit above it can be used as an LVC for the controller or a HVC for a power supply

you can do it like this if you like though:

b6v4e.png


Just know that the alarm port has a max rating of 500 mA
 
for mounting the lights in a way that you can easaly change them out why not gram some small diamitor steel or copper tubing w/ knotches in it to let the wires out then use a cople of springs to hold the bulbs down and in place

 
Fool proof? Considering the quality of fools we have available these days... NOTHING is foolproof :evil:
 




All done. Well mostly. I need to add two more bulbs, and see if the mounting method will work without the glue melting. Also, a working watt meter would be nice for capacity testing.
I've got to say, the mr16 bulbs are not ideal for this application. Hard to mount, and the solder doesn't bond to the pins. I had to make a strong physical connection by wrapping the wire around the pins, and then solder them.
Unless you, like me, use mr16 bulbs on your bike, I'd suggest trying another type.

FYI, some of the schematics above are wrong. I burned out the alarm port on a cell log because the diode on the relay was installed backwards. Sucks :(

Correct schematic:

 
I can't find the links I had to a couple of threads, and for some reason I can't remember even the member's name that had started the threads. IIRC one thread specifically talks about using lightbulbs as loads, and demonstrates some testing using 100W floodlights?
 
amberwolf said:
I can't find the links I had to a couple of threads, and for some reason I can't remember even the member's name that had started the threads. IIRC one thread specifically talks about using lightbulbs as loads, and demonstrates some testing using 100W floodlights?
Look here?
http://www.ypedal.com/Lbd.htm
 
my LBD setup has been going for a number of years now, 100's of cycles on it, works at anything from 24v packs ( slower ) up to 100v packs( Fast and BRIGHT )

I plan to build a metal tube to cover the bulbs, when i load test high voltage packs it feels like the 4th of july in here...
 
Is a Cell logs data viewable using LogView, or does a Cell Log come withits own software?
 
So everything is going well with the tester and the new battery pack.

Well, almost everything! I had a small glitch with my setup, and I can only guess it has something to do with a buggy cell-log. Basically I walked in and the alarm was going off (silently) but the lights were still on. I say the lights flash on and off a few minutes before as the alarm triggered the relays, but I guess the alarm port stopped functioning. Doesn't bode well for my opinion of cell-logs! Anyways, the cell was only down to 2.9v, so it's not a huge deal.

But I need to figure out a way to make a latching relay that activates with a momentary switch so that the only I can turn the main relays on. This way if the cell logs hit LVC even once, the discharge is done until I manually restart it with the momentary switch.

Any ideas on how to do this? It's probably obvious, but I'm wore out and been racking my brains on it.
 
America Zettler AZ2500P2-1A-12DE This is a self latching relay, the coil come various voltages $ 5 at alliedelectric
 


In the light of day this took me all of 10 minutes to come up with. Amazing what some rest can do for the ol' meat computer.

This is going to sound like a totally retarded question, but if I have, say a 12v relay, can I drop down the voltage of a 48v pack with resistors or a resistor divider to run the coils?
 
It wouldn't be that big of a deal to just use some resistors to drop the voltage for the relay, but it would only really work well around the pack voltage you designed it for.

I would get a snazzy switching reg from TI that will be more efficient and work across a wider range of voltages. They send free samples to 'legit' companies. I'm sure you wouldn't abuse the system... ;)
 
Cool cool cool. More efficient isn't always good, especially for a battery pack discharger :)

To be honest, what I really wanted to know that for was building a ignition circuit for my bike. As in, turn a keyswitch that activates relays that power the ignition wire on the controller, dc-dc, cell-logs, temp sensor, etc. Since I can't run relays off pack voltage with out dropping the voltage and a keyswitch wouldn't last at those voltages, I need a way to get 5v or 12v. In the past I just left the 12v dc-dc connected, but there is a parasitic drain. As there is from the controller being left in stand by mode. Eventually, I'd like to use a timer circuit and bigger relays to make a precharge circuit so that all power can be completely disconnect by a simple keyswitch.

Searching mouser under these terms gets a bunch of linear regulators that can accept input up to 100v. Could use it for the discharger, and on the bike.

Anyways, the new discharger design works. I did blow out the alarm port on the cell log by accidently shorting it to the load negative, but luckily I had a dead cell-log that I could pick the replacement part from. Everything is peachy now.
 
auraslip said:
last at those voltages, I need a way to get 5v or 12v. In the past I just left the 12v dc-dc connected, but there is a parasitic drain. As there is from the controller being left in stand by mode. Eventually, I'd like to use a timer circuit and bigger relays to make a precharge circuit so that all power can be completely disconnect by a simple keyswitch.
Set it up so that the keyswitch just turns on power to the DC-DC input, and that all the relay coils are wired up to the DC-DC output. No power wastage when keyswitch is off, and no complex stuff to power the relays or whatnot.
 
Back
Top