CellLog 8S battery monitor and logger - any reviews?

I am interested in that. Forgive my elecronics ingorance, but i am trying envision how to hook it up. :?

So would one use wires 2 through 9 from the CellLog and connect them to one side of the "double contacts" for each of 8 posts, and then back out from the second contact of each of 8 posts?

And for the switching on and off I guess there is another set of connections for a STDP type switch to activate the relay? and that would be powered from a 14 volt line off of the pack somewhere?

Am I close? :|

Oops just saw Goodrums post... so maybe just a simple switch on the pin #1 wire?
 
Yes it appears to work , just unlatch first pin(-). Anybody tried the "equal draw current from all 8s mod". I am contemplating trying this, on 3 brand new cellogs. for a 20s "bms" project. I will engage pin 1 on the cellogs with a mini 3pst relay.

mike
 
Doctorbass said:
For those who would like to activate or not their permanently installed celllog8, Mike on this forum have found a great relay that can switch all contacts at the same time so they dont stay always ON when connected to the battery!!

8PST for 2.50$ !!!

PC Mount Sealed Relay - 8PST
(KO) 0490-1385

KV PC mount sealed relay. 8PST, normally open double contacts. 13.8 vdc coil. 7/16" x 1-3/8" x 1-3/16" high. Alternate P/N: DM723.



here is the link: http://www.surplussales.com/Relays/sealed_pc.html

ko-0490-1385.jpg





Very great find!!!


Doc
Click to expand...

Doc, Are we really sure we need a relay? I want to leave my cellLogs perminantly pluged into the pack but of course not actually drawing power. So last week I hooked up 1 cellLog to a Thunder Sky 24v 20ah pack. I charged it to 3.65v took my cell reading from the cellLog and then disconnected the #1 wire. the CellLog screen went blank. I have been checking the cell voltage twice a day since. The serface charge is down to 3.5ish on all cells except the cell with the missing connector wire. This cell is down to 3.4v. I am not sure but I think it has bee about 10 day now.

My question is; am I really getting a true reading for the cells? Or am I going to get a big surprise when I go riding and the pack is actually dead?

I was hoping this idea would work so I could install a simple SPDT switch inline.
 
Yes, I can also verify that disabling pin 1 will cut power.

I'm also doing a 24-channel CellLog-based "BMS". It will have optos for three CellLog alarm outputs, relays to control power to all three, which are driven from the +5V coming from the throttle/brake line, used by the LVC signal. This way, the CellLogs aren't on, draining power, unless the controller is on. Also included will be three 8-channel balancer sections, and an integrated Charge Controller, with the current limiter and auto-shutoff function. When the supply/charger is connected, it will also power the relays, to turn on the three CellLogs. The charge voltage will be set to the shunt turn-on point for the balancers, which will be 4.15V. The HVC value set in the CellLogs will be a bit higher, like 4.17-4.18V. The HVC signal and the throttling logic will keep any errant cell from going above this HVC set point, but under most cases, a cell won't get that high. The 1A shunts do a pretty good job of keeping the voltage for high cells down close to the balance/charge point.

The three 8-channel balancer circuits, and the Charge Controller section will be on one 3" x 12.5" board that will fit into two 3.06 x 6.30" Hammond extruded aluminum cases, that will be joined, end-to-end. I am also making a custom 12.5""lid" that will have holes for the balancer LEDs , and will have connections/mounting pads for the CellLogs. The opto/alarm logic and relays will be on the underside of the lid. I will also do custom end plates, with cell connections on one end, and supply/charger connections on the other.

My plan is to mount this unit on the top bar on my Townie, mainly so I can get an idea which cell is tripping the LVC, but it will also be a convenient spot to be able to monitor the charge process. My plan is to put a couple of MeanWells in a rack bag, along with an extension cord. The holy grail for me is to get to the point I can simply plug in one power cord, press one button, and go do other stuff. :) I can check out the progress by simply walking over to the bike. Balancing is only done when it is required. If the cells are balanced, the shunt circuits stay off.

Anyway, I'll post some pics of the layouts, etc., as soon as I finish them.

-- Gary
 
pretty much exactly what im doing. the charge relay will be latching, and high amperage. The charge will be normally closed. how will you engage the cellog power relays from the charger source? I assume your charge controller has some 5v available?

Balancing will be accomplished by a: quattro 4x6 on custom balance harness with 25 pin plug to 20s battery.

b: custom single cell charger switched to each cell idividually by a magnet and reed switches. i forget who to give credit for this idea!! this is a "top up" to go along with the meanwell(?)

mike
 
hydro-one said:
pretty much exactly what im doing. the charge relay will be latching, and high amperage. The charge will be normally closed. how will you engage the cellog power relays from the charger source? I assume your charge controller has some 5v available?

Balancing will be accomplished by a: quattro 4x6 on custom balance harness with 25 pin plug to 20s battery.

b: custom single cell charger switched to each cell idividually by a magnet and reed switches. i forget who to give credit for this idea!! this is a "top up" to go along with the meanwell(?)

mike
Click to expand...

I'm not sure why you need a high amperage relay, are you talking about switching the main charge current? I'm just using relays to turn pin 1 on for each CellLog. If I can find a small 3PST version, I'll just use one relay. Otherwise, I'll use one small SPST for each CellLog.

The Charge Controller has 12V regulator that is controlled via a relay/SCR combo. A "Start" button is pressed to get things going, which turns on an SCR, which in turn, energizes the relay. The relay provides power to the 12V regulator section. When the charge current drops below an adjustable set point, during the normal CV phase, the auto-shutoff logic cuts off the relay, which shuts down the 12V, shutting the whole Charge Controller down. Anyway, I'll use the 12V, dropped down, to engage the pin1 relays.

The Charge Controller has two main functions. First, during the CC charge phase, it will use a PWM-based circuit to limit the input current to an adjustable set point. This is mainly so that inexpensive power supplies, like the various MeanWell modes, can be used, even though they either have a peak current limiter, or a simple "hiccup" mode on the front end. The adjustment range is 2-20A. The second function is to use the HVC signal to "throttle" the input current in order to keep any one cell from going over the HVC set point.

-- Gary
 
cool, im just going to cut the charge current directly, by pulling a latching relay "off" with the opto/hvc. This will prevent overcharging of cells. and tell the owner its time to balance, as he will have to reset the charge relay with a momentary push button (energise the other coil in the latching relay). This relay can also be my /cut throttle relay, (I kind of like that he will have to press a button to go again, after a cell drops. Keep in mind I will still have controller lvc, and cycle anaylst on board. should be hard to kill!!!!

I just did the equal current mod and soldered wires to the pins exiting the cellog unit. I hate jst-xh connectors!!!

cheers!
mike
 
Here's what I'm looking to do:

View attachment 1

The first two relays turn on the CellLogs. The 3rd one controls what turns them on, either the +5V from the motor controller (via the throttle/brake interface...), or the +12V from the Charge Controller.

My plan is to integrate this with the Charge Controller and a 24-channel version of the new balancer cell circuit, which is shown below:

View attachment LiPo-LiFePO4 CellLog Balancer-v4.1.2f.png

This will end up being a "hybrid", of sorts, between a full digital and a full analog BMS, at a pretty affordable price, I think.

More later...

-- Gary
 
Looks good!! I am also thinking of implementing a lvc power down of the whole shebang. in case he leaves the lights on, key on , or something, . it will be a latching relay. that controlls power to the controller or dc/dc. This way the cellogs can shut themselves off!!!

I just replaced the tranaisots in the cellog , with a larger npn transistor. now it powers the relay directly... not that hard of a fix!!!

mike
 
When you disconnect the pack (-) pin on the CellLog, there is still some drain on the pack. It's tricky to measure because there is current flowing in more than one wire. I got about 340uA on pin 6 without the "all cells" mod. I was trying to find a spot on the board where I could disconnect all power and just be left with the drain of the input resistors, which are 75k ohms.

The thing also has a 'sleep mode' but is only actvated by a time out. It would be nice if this feature could be remotely controlled by an opto coupled connections.

The main ATMega 32 processor is pretty standard looking. I see the board has the ISP programmer attachment points.

Has anybody tried reading the chip? It may be possible to 'hack' the software.

I have an Atmel AVRISP lying around somewhere. I should try hooking it up.

Oh, and I love the startup screen on the 8M "Do not support the logging"
Maybe the programmers are tree-huggers?
 
HAHAH treehuggers......so current does flow, unevenly, even with the all cell mod and disconnect pin 1 then. so it will still need to be balanced because of this, thats too bad!!
you should look into hacking the cell log software!! I dont understand why these guys a junsi wont work with the ev community more for a x-channel box...if they wont then we'll have to do it eh?
 
If the output was isolated, it would be easy to daisy chain as many as you want. It won't be too hard to build a little isolated interface that would do this. :wink:

The current drain is a bit harder to deal with. There has to be a hack for this. Even at 500uA, you could leave it connected for a long time without serious balance issues. If you were going to store a pack for winter, for example, it would be a good idea to disconnect them, but I think for 1-2 months on a 10Ah pack might be OK. If the charging setup does even a little balancing, it should keep them in line.

I still need to measure how much it draws in the sleep mode. I don't think I'll be able to use it, though, unless there are changes to the software/hardware.

I also see that the alarm output can be configured for either normally on or normally off. This could allow for series connected outputs that would provide a fail-safe in the event a connection goes bad.
 
I tested the sleep mode and it seems to be pretty useless.

Using pins 0-1 (pack level pigtail that comes with it), I get around 20ma when running, and 14ma when 'sleeping'. It seems the sleep mode only turns off the backlight and display.
 
I'd love to help, and not least of all because of the treehugger line above. That's good stuff.

We are the largest US distributor and retailer for these products, and work with manufacturer's engineers on new product testing and development. PM or email - let's talk :).

Cheers,
David Gray
ProgressiveRC
 
write2dgray said:
I'd love to help, and not least of all because of the treehugger line above. That's good stuff.

We are the largest US distributor and retailer for these products, and work with manufacturer's engineers on new product testing and development. PM or email - let's talk :).

Cheers,
David Gray
ProgressiveRC
Click to expand...

David,

Thanks for the offer and I am certain many of us will follow up with you...

I know several of us have been in touch with Junsi (or someone claiming to be Junsi) whom I believe is the beginning manufacturer for these... I do believe the lower cost no logging memory version was a result of some of our contact in the past (though having the ability to output even TTL RS232 would have been nice)..

Hopefully communications with you can lead to more "fruitful" results :)

Regards,
mwkeefer
 
write2dgray said:
I'd love to help, and not least of all because of the treehugger line above. That's good stuff.

We are the largest US distributor and retailer for these products, and work with manufacturer's engineers on new product testing and development. PM or email - let's talk :).

Cheers,
David Gray
ProgressiveRC
Click to expand...

Awesome!

I'll start working on a wish list.
You can see some of my technical comments in another thread: http://endless-sphere.com/forums/viewtopic.php?f=14&t=20142

So far, the list includes:

A way to power the thing off remotely without disconnecting it from the pack (70uA drain from the input resistors is fine). Ideally, this remote power switch should be triggerable from an opto coupler.

Make the drain even on all cells. Maybe the LM324 can be referenced to the 3v supply like the 358 to avoid over voltage if the D4-D5 diodes are jumpered together.

Isolated alarm output to allow multiple units to be easily connected. This is easy.

Isolated USB output. This one may be a bit expensive/cumbersome, so is not the highest priority.
 
any issues when having the CellLog 8S monitor connected to a 8 cell series battery pack when charging using a AC regulated switching power supply?
is this okay to do, or should the CellLog be disconnected when charging ?
 
Thanks for the feedback folks and certainly feel free to contact us with any needs you may have - lots of innovative products are in the works and feedback is always appreciated.
myzter said:
any issues when having the CellLog 8S monitor connected to a 8 cell series battery pack when charging using a AC regulated switching power supply?
is this okay to do, or should the CellLog be disconnected when charging ?
Click to expand...
No problem, as long as you are using the included USB or one with the ground removed from the "shield" to prevent any possible common ground with your computer and power supply. Also note that you can use any USB cable if you are working with a laptop running on battery or if using a battery source (i.e. Pb) to feed the charger.

- David
 
Does anyone tried to recalibrate the voltage of every of the 8 channels ? using the feature included in the OS of the celllog8 ?

I can change the value but it just dont keep them in memory when i quit that menu..! :(

Doc
 
You have to hold the enter button for more than 3 seconds to make the new value stick (I think).

I found that the factory calibration is extremely accurate so far.
 
David,

Thanks for the prompt help with my charger issue!

With regard to the CellLog8s...

1.) I can understand them not wanting to provide opto coupled outputs but if not optocoupled USB (which would raise teh price) how about just 2 test vias routed to RX and TX of the TTL level serial already occupied - most of us have USB->TTL and it's so much simpler to connect 5v, gnd and TX from each unit optocoupling the TX line so a primary unit would even then be able to receive updates from the sensor modules in an isolated fashion.

2.) Internal 3.9v reference (+- 5mv) power generator to use for calibrating each channel, there is no reason a tightly regulated 3.9 (or whatever is simplest to create with the +- 5mv tolerance) so we can recalibrate these units - I've noted a bit of drift over time but more importantly 1/2 of the units I received were up to 15-20mv off (reading higher than actual voltage so still safe if used for HVC but I had to compensate for it with higher LVC).

3.) A sleep mode would be excellent - then users could configure an external pulse generator (say each 10000ms) to send a wake signal to all connected - again - just expose a pin we can optocouple it. I believe the Atmega32M has a real shutdown mode where current draw is in essence NIL but requires a wake up line... same line could be a toggle - pulse once high 5v and go to standby... pulse once again high and back to normal operation.

4.) Having control to disable the back light with a Digital input toggle would be nice too.

5.) In regard to the TTL Serial output - that USB->TTL chip consumes power - removing it will reduce the drain on the circuit while logging.

6.) Suggest they provide these with a Male 9 pin JST-HXT (8S) -> 2 X 5 Pin JST-HXT Female (4S) each - since that is how most balance taps of 8S come configured - it will also make configuring custom harnesses (for instance 2 x 5S + 1 x 6S in series connected to 2x 8S units) much easier for those of us not planning to use with 8S

On a seperate note - the iCharger 208B needs to come with an 8S breakdown HXT-JST balance adapter, perhaps it was an error but the one I rececived had no such adapter (or I would already be using it with the CellLog8, or if they had one I would be using it with the 208B Charger) - I wound up cutting 2 wires away from the JST-HXT 11 pin conector for the iCharger 1010B+ (I have one I made for this) and trimming (with a flat pair of wire snips) one of the ridges down to make it fit (it looks bad to me).

These should also be available at a reasonable price for additional units and shoudl cover: 8S (to prevent teh need to remove the adapter and maintain viability of the connector), 7S (as 4S and 3S), 6S, 5S, 3S and 2S.

David - thanks again for the prompt service and hopefully some of our feedback can be taken into account. I think I speak for many of us when I say many would be willing to assist in the development of a robust unit with useful purpose to the RC crowd but also to the EV and LVC crowd - to add features (inputs, outputs, balanced and low discharge rate, sleep modes, perhaps lower still priced models without LCDs?) and create a much better product with a wider demographic - you can always just come up with a 10S and always be playing catch up or... you could get them to make these share a bus so they can pass high and low cell voltage info to each other and thus truly know the difference across an entire pack in series (10 and 12S are becomming more and more prevalent).

Sorry it's late and I'm rambling :)

-Mike
 
You must log in or register to reply here.

Similar threads

P
CellLog 8S alternatives? Coulometers?
Replies
8
Views
1,776
dogman dan
dogman dan
Doctorbass
After the CELLLOG8 now THE POWERLOG6 !
Replies
5
Views
1,624
liveforphysics
liveforphysics
Alan B
Designs for Reliable (low V) Cell Monitoring electronics
2 3 4 5 6
Replies
126
Views
14,204
CamLight
CamLight
Back
Top