CellLog 8S battery monitor and logger - any reviews?

[Doctorbass]

NEW FIRMWARE AVAILABLE 2.06 :wink:

Released 19 nov 2009

It seem that Junsi is an active member of the RC group forum!

http://www.rcgroups.com/forums/showthread.php?t=1055820&page=11


A friend of mine sent it to me.. probably got from the RCcgroup guys...

(See attached file) You must have installed the driver with little cd BEFORE the upgrade!

Release notes:

CellLog 8S

New Features
============
*Add "Power Management" <Thanks speedebikes>
*Add sampling time display on "Inf. Screen", and the shown rest time is related to sampling rate <Thanks Dan >
*Modify the logging data structure, and adding 18% data storage capacity

Bug Fixes
=========
101: Advise the problem of LP low displayed voltage when select Fast Startup <Thanks Torsten>
102: Advise the display problem when rolling-over sometimes <Thanks Torsten>

Note
====

////////////////////////////////////////////////////////////////

////////////////////////////////////////////////////////////////
// Version: 2.05 //
// Date of Release: 2009/10/24 //
// Owner: CWF //
////////////////////////////////////////////////////////////////

Part support
============
CellLog 8S


Bug Fixes
=========
100: Solve the bug of vertical strips on LCD. <Thanks speedebikes>

Note
====


////////////////////////////////////////////////////////////////

////////////////////////////////////////////////////////////////
// Version: 2.04 //
// Date of Release: 2009/06/25 //
// Owner: CWF //
////////////////////////////////////////////////////////////////

Part support
============
CellLog 8S

New Features
============
*Add the sampling rate setting. <Thanks JohnMuchow>
*Start Quick Select battery type. <Thanks Julez>
*Add Start section Select. <Thanks Julez>

Note
====

Doc

 

[Doctorbass]

For your pleasure...

Doc

Doc

 

[mwkeefer]

Hello all,

Ive discovered an issue with the CellLog8 units.... I was just trying to use these connected to my custom 3 port monitor (outputs LogView format for x number of cells in a pack)... I had been working with this as a method for monitoring via PC and controlling (precursor to microcontroller directly)...

Here's the issue... Take a 15S pack in series, connect a CellLog8 to each of the balance taps... then attempt to connect the 3 of them into your USB hub... instant smoking wires.

I assume this is due to the USB power + the battery power and really I think only TX and RX are required ... if USB power is being provided by the cells.... I've done this before with a single CellLog8 but never tried all 3 at once yet.

I'll have to email junsi about this becuase it should isolate voltage when being powered so as not to draw the power from USB port... RX/TX can still be used - I think.

-Mike

 

[SpeedEBikes]

Here's a repost of something I just posted to RCGroups forum...

I considered wiring the alarm outputs to my controllers 'ebrake' cutoff to mimic the behaviour of a BMS which cuts power. But I realized that I don't actually like that behaviour. I want to know my batteries are nearing depletion but I want to voluntarily cut power not just have it cut out by surprise. But if you do go with the power cutoff approach, a point in favor of the celllog over a BMS is that a BMS switches the main battery power and that switch adds resistance, heat and is a potential failure point. To make it robust and efficient means added weight, bulk and cost.

So I find the idea of an alarm very appealing.

But the internal alarm on the celllog is weak. You can hear it from maybe 15 feet away in a quiet room, but on a bike/vehicle at speed or in city traffic you may not hear it. And if you enclose the celllog for weather protection the problem is even worse.

But I found a very easy solution. For about $5 you can get a super loud piezo buzzer from Digikey, part # 102-1643-ND. This buzzer will operate on something like 3-30 volts ac or dc. Volume and current draw go up proportionately with voltage. I run 6s lipo packs, 7s li-ion packs and 8s life packs. These buzzers work great with all of them. All one has to do is wire the buzzer in series with the alarm output of the celllog (think of it as a switch) across the entire pack. If one was using an 8s lipo or liion you probably want to wire across only 7 cells to avoid pushing too much voltage through the alarm. You might also choose to use fewer cells if you find the alarm is too loud. I don't think wiring across a subset of cells is at all an issue for pack imbalance as the alarm only draws a few milliamps and shouldn't ever be on for too long. Any solution to pack balancing, even if rarely used should suffice to correct for this.

I typically run two to four packs in series with a separate celllog and alarm on each. The per cell voltage alarms are all non latching. That means as you roll on the throttle and cell voltages sag under load you will start getting alarms tripping as you approach depletion. Ease off the throttle and the alarm stops. You can set the voltage levels that will trip the alarm to dial in the amount of warning you get before the pack is truly done.

I really like the redundancy of having an alarm on each pack. I can clearly hear each packs alarm. If one pack is a little weaker than another I can hear it trip first or find a throttle level that will trip one but not the others. I like the idea that if a celllog fails or an alarm fails there is almost no impact as I still have power and I still have the protection of the other pack(s) alarms. And I know to investigate if I can hear one alarm but not another.

There are alarm triggers for high and low cell voltage which most everyone will want to enable all the time. There is also a trigger for max cell voltage difference. I like to set this one fairly tight such that if any cell is dropping in voltage more quickly than normal I will know. I use packs made from cell blocks wired in parallel and then in series. If a single parallel connection fails then that cell block will deplete faster and sag harder than the other cell blocks and the cell difference alarm will alert me to that fact early before the pack is run down.

However if you have an old pack with some weak cells you probably want to set the difference alarm high enough to not trip with your typical operating useage. The celllog's logging feature is really handy for this. Record your typical useage, dump it to your PC and then check out the graphs. You'll easily be able to spot which cells have high IR and/or low capacity. For big packs made of lots of paralleled cells it then becomes pretty easy to tack on an extra cell or two where needed to create a better behaved pack.

I put both the alarm and the celllog inside my battery cases to protect them from weather. Also, I don't like the idea of having long balance taps with all the small delicate wires extending outside the protective confines of the case. Small delicate balance wires have the advantage of being self 'fused' in that they will blow readily in the case of a short. To run long external balance taps one will likely want to use heavier wire each individually fused adding complication and expense.

So in summary, I love using the celllog8s on EVs. Even though I sometimes run 24s I don't really want a single 24s celllog.

My biggest issue right now is that there isn't good auxiliary parts support for these units. I can't find prewired JST XH-8 pigtails, extensions or Y adapters. You can buy the connectors and terminals for these from Digikey, which is what I've done, but wiring them up is challenging and tedious. Also, the alarm wires are exceedingly delicate and I have broken one of the alarm wires off from a connector and I don't know how to get more alarm wire pigtails either.

Someone had asked about latching alarms. The pack voltage difference alarm is for the delta between the pack's minimal and maximal recorded voltage for the duration the celllog has been powered on. Thus once this trips it stays on until you unplug the celllog.

Finally, in my wishlist of features for the celllog it would be nice to have a way to query what tripped past alarms such that when one looks at a celllog after the alarm is over, one can readily see what it was about. Perhaps one of the 3 different view screens could be used for this purpose or a 4th view screen could be added that would show past alarms. It would also be good to be able to see info about alarms in logview when looking at exported logs.

 

[webfootguy]

I don't know if this helps, but the only 8s balance pigtail I have seen is on ebay. Check out item 270470253072.
Spendy at $13 but if it is what you need, saves lots of time.

 

[mwkeefer]

I'm still doing it the old fashioned way... making adapters from 10cm extension cords, so I adapt 3x5S into 6x4S at the taps... I haven't found 8s plugs either.

With regards to extending the celllogs to work better in EV solution, I'm waiting to hear back from Junsi but they were pretty enthusiastic about it (once they were convinced of the market size)

I've tested my CellLog8's and it seems just the USB connectors themselves (the minijacks) are toasted... they still work fine once I bypassed the SMT jacks. I've placed an order with Mouser for replacements (I think it was mouser, maybe digikey... too many orders being placed now adays!)

-Mike

 

[patrickza]

Doc did you ever look into setting up the group buy for some 24s Celllogs? I would love to have this on a bike, and I think every lithium electric car should them too. Junsi seemed quite keen when he heard that there could be a purchase for 1000 of them. Did Junsi indicate what price he's be willing to sell 1000 24s monitors at? If the price scaled up along with the number of cells then it could be $100 for a 24s. I'd take one for sure and maybe more at those prices.

I think this would be the next must have instrument for any serious ebiker after the cycle analyst of course. With one of these the cell LVC and cell voltage differences are taken care of. All the BMS needs to do then is balance the pack.

 

[Doctorbass]

I never received any news from Junsi

My friend tried to join directly the compagny and they are willing to make a 45s version!!

They also sent us the schematic of what it would look like !!!

Here is what look like the emails ( I replaced the Email adress of my friend with "X") :

Date: Fri, 25 Dec 2009 17:20:45 +0800

Hello X,

This is the hand drawing of the 45s CellLog. ( CONTACT ME FOR DETAILS Doc.)

Regards,
Cindy


--------------------------------------------------------------------------------
Date: Thu, 24 Dec 2009 13:27:34 +0000

Hi Mrs Cindy,

Thank you very much... for all your help. Now it will be look after by the appropriate purchase department.

Best wishes!...

X

--------------------------------------------------------------------------------
Date: Thu, 24 Dec 2009 09:41:29 +0800
Hello X,

First, pls pay 10% of the total 1000 pcs as deposit.
When we have a rough design of it, we will send you to have a test.

If OK, pls pay 30% of the total money,
then we will open a model and get the materials ready.

When all are done, pls pay the rest 60%, then we will arrange delivering the products to you.

Regards,
Cindy


--------------------------------------------------------------------------------
Date: Wed, 23 Dec 2009 22:50:24 +0000

Hi Mrs Cindy,

One last question... what would be the deposit for start and how would you like to manage the balance amount billed for this project…
Thank you very much for your devotion... (and Merry Xmas!)

X

--------------------------------------------------------------------------------

Date: Wed, 23 Dec 2009 17:44:40 +0800

Hello X,

The CellLog 8S's LCD is 128*64, and maybe CellLog 45S 's LCD is 160*160,
or bigger than this: 320*240.

As for the design, it will send you if you confirm to make it.
Moreover, you have paid the deposit for it.
Then we will provide the design pictures to you.

Merry Christmas and Happy New Year in advance!

Regards,
Cindy

Shenzhen Junsi Electronic Co.,Ltd

Room 578 & 588, Building 9A
Anhua Industrial Zone
8th Tairan Rd, Che Kung Temple
Futian, Shenzhen, China, 518040

Tel: 086-0755-88845648
Fax : 086-0755-88845648-12
Email: szjunsi@163.com, harrief111@hotmail.com

website: http://www.jun-si.com
http://www.hillrc.com

--------------------------------------------------------------------------------
Date: Tue, 22 Dec 2009 17:10:48 +0000

Hi Mrs Cindy,

Here are the batteries type: ( Lithium based battery)

-LiFePO4 ( 2.5V up to 3.7V)

-Lithium manganese (3.0 to 4.2V)

-Lithium Cobalt (3.0 to 4.2V)

-Lithium Polymer(3.0-4.2V)

-Ideally, for the LiFePO4, the low voltage limit must be adjustable between 2.0 and 2.8V instead of just 2.5V and for the rest the low voltage limit must be adjustable from 2.5 to 3.1V instead of just 3.0V.

-The number of cell that can be used with shall be between 1 and 45 cells.

Would it be possible to see what would look like the display to show the 45 cells in sequence.... how many rows could you display at the same time of monitoring the cell...(even a hand drawing will do it) and also it would be appreciate to keep the same features as the CELLOG8 version.
Thank you very much... wishing you a Happy Xmas and Happy New Year 2010....!

X

 

[morph999]

awesome, if it's not too expensive, I might be interested in that. How hard would it be to hook it up to say 12 Thundersky cells ?

 

[morph999]

Wired mine and made an abs case, I will still have to make an acrylic clear cover to make it water-resistant. I then mounted them to the case using 3M Dual Lock tape. The 3M also acts as a shock dampener to the CellLogs. Also shown is the miniature 3pdt switch mounted at the bottom of the case for on/off operation. I coated the contacts with several coatings of liquid electrical tape, hopefully this protects it from the elements.

I've done about 80 miles riding with the CellLogs and it seems to be performing as it should. I tested the built-in alarm sound and its loud enough to be heard when riding. I might use an external piezo-buzzer when the clear cover is mounted in case the sound is muted. I haven't done any data logging and pc connectivity yet, which I'll eventually do in the future. As recumbent has mentioned, this replaces an onboard bms to monitor low voltage. I'm thinking of just using an external "Even newer" bms with a 1A shunt for balancing while charging since my MKS 3-in-1 have been boringly slow to balance.

wow, Shinyballs, can you tell me how you did that? I only have 12 cells so I wanted to just buy two of them and do it like you have. I'd be hooking them on Thundersky cells.

 

[patrickza]

My friend tried to join directly the compagny and they are willing to make a 45s version!!

So I take it your next bike is going to be a 45s beast now

Please keep us informed on the celllog.

 

[j3tch1u]

Wow!

I am currently working on interfacing a single MCU (Atmega168 or 48 will work fine) to these CellLog8 before the USB chip... this is TTL level serial IO before the USB chip.

hey mike were you able to get the cell logs to talk to arduino thru usb? would it be possible to turn the arduino into a host? keep up the good work.

 

[mwkeefer]

Not via USB but by isolating the TTL level serial IO (on the far side of the USB IC) using 2 channel optocouplers per celllog8s I have been able to monitor multiple units from a single Atmega / Arduino.

-Mike

 

[webfootguy]

Details please. I'm very interested.

 

[Frank]

I'd be interested in a 24-cell version too.

 

[monster]

i would be interested in a 24s version.

i take it you could plug just 16s with it? can we re-cap what applications these things have? would it allow you to spot if a weak cell was limiting your pack?

 

[pm_dawn]

Hi !

If the 45s version would have a current sensor and could calculate wh that would be pretty much all you need.
my cents.

Best regards
/Per Eklund

 

[Doctorbass]

For those who are searching for these famous 8s ( 9 pins) connectors

The are called JST connectors.. serie XH

You can find them .. at Digikey:

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=455-2217-ND

And the crimp connector for AWG 22 to 28:

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=455-1135-1-ND

Doc

 

[cell_man]

I'm also very interested in finding a way to run several Cell Log8s on a bigger pack and monitor the voltage on a single PC. If anyone knows the answer to the below I would be really grateful

1. Can several Cell Log 8S be used to monitor several 8S packs (not tied together into a bigger pack) on a single PC simultaneously. Can Logview support this. I'm pretty sure this is no problem but would rather a confirmation before frying anything. As far as I know you can only monitor 1 devise at 1 time on the Loigview, but you can open a 2nd Logview and monitor another device on that.
2. Is there no way to isolate the power from the USB to prevent the issue with running multiple Cell Log8s on the same PC? Surely you could fit a DC-DC convertor or just remove the power althogether from the USB as the device can be powered from the pack and just transfer data without the power. I will have to look into the circuit for a typical USB device to try and understand it a bit better. Any good docs or resources online for USB info?

Those connector are readily available in China and with leads already connected. If anyone wants a reasonable quantity of them I'll buy a bag of them and send them. There is a an area where i get all my tools and parts where I've seen them for sale and will check the price when I'm over there at the weekend. I might even buy a quantity of the cell logs 8s if there's some interest in them and I can get them at a good enough price to be competitive. Could throw in as many tails as required too. I spoke with Junsi about this issue and they do not have any solution. I've also chatted to them before about getting a bigger Cell Log 8S. I don't think the required numbers would be quite as high as earlier mentioned. How about a 16S unit that could also be connected to a bigger pack and monitord simultaneously? Or even just stick with the 8S but modify it for series connection. I plan to go down to Shenzhen in the not too distant future and was planning a meeting with them anyway.

 

[tailwind]

Paul,

Just be aware that the original cellLog 8S only draws its power from cells 1 through 6.
see speedebikes post #118, #121
http://www.rcgroups.com/forums/showthread.php?t=1055820&page=8

Junsi posted a hardware mod to fix the problem. see post #123, #137

Hello Sir,

Any value ranges from 0 to 100 ohms is OK, as there is no strict request.

It's related to voltage regulation power consumption issue. Now it's used for 8s, but before this it is used for 6s*4.2V, then it short down 3V. If add the resistor, it will be 8s*4.2V. However, we have not thought about the battery non-balance problem. Later, we will improve it in our next version.

Regards,
Junsi

Greg

 

[cell_man]

Paul,

Just be aware that the original cellLog 8S only draws its power from cells 1 through 6.
see speedebikes post #118, #121
http://www.rcgroups.com/forums/showthread.php?t=1055820&page=8

Junsi posted a hardware mod to fix the problem. see post #123, #137

Hello Sir,

Any value ranges from 0 to 100 ohms is OK, as there is no strict request.

It's related to voltage regulation power consumption issue. Now it's used for 8s, but before this it is used for 6s*4.2V, then it short down 3V. If add the resistor, it will be 8s*4.2V. However, we have not thought about the battery non-balance problem. Later, we will improve it in our next version.

Regards,
Junsi

Greg

That's useful info and explains some things I've seen with a pack I've been testing. I looked through that post though and didn't see any solutionthing about the USB issue. Hopefully 1 of the gurus can come up with a solution for isolating the USB. I want to make some test rigs to check packs under load and ensure good matching of the internal resistance. They are really neat little devices and if I could log several of them simultaneously it would be great.

 

[pm_dawn]

Hi !

Can't you just run the logging in them end then transfer the logs to the PC one by one ??

Or do you want to look at it in realtime ?
Or do you want the PC to log all the Cellogs data in realtime?

Best Regards
/Per

 

[cell_man]

Hi !

Can't you just run the logging in them end then transfer the logs to the PC one by one ??

Or do you want to look at it in realtime ?
Or do you want the PC to log all the Cellogs data in realtime?

Best Regards
/Per

Yeah I can transfer the data 1 by 1 as you suggest but it would be so nice if it could be logged and monitored in realtime. How about the monitoring function of the Icharger 208B. Surely that must be floating as long as you used seperate floating PSUs? I'd like to have a few of these running simulaneously to monitor packs during charge and discharge cycles.

 

[Doctorbass]

something that everibody should know about the celllog8


(FROM rcgroup)

Quote:
Originally Posted by speedebikes
I've got 5 of them mounted on 7s lithium ion packs and they all exhibit an annoying 'feature'. They aren't pulling any or much power from the last cells in my packs. So over time cells 1-6 come down pretty much together but cell 7 ends up high.

The reason is only 1-6S cells supply power to CellLog, but 7-8S cells not. you can add a resistance to the PCB as the picture shows. then all 1-8S can supply power. thanks.

http://static.rcgroups.com/forums/attachments/1/4/3/9/4/1/a2833845-102-CellLog7-8R.jpg?d=1256359601

Hello Sir,

Any value ranges from 0 to 100 ohms is OK, as there is no strict request.

It's related to voltage regulation power consumption issue. Now it's used for 8s, but before this it is used for 6s*4.2V, then it short down 3V. If add the resistor, it will be 8s*4.2V. However, we have not thought about the battery non-balance problem. Later, we will improve it in our next version.

Regards,
Junsi

http://www.rcgroups.com/forums/showthread.php?t=1055820&page=10

Doc

 

[Doctorbass]

A simple software for the celllog developped by AJH and shared on the RCgroup:

http://www.rcgroups.com/forums/showatt.php?attachmentid=2906158&d=1259663935

source:
http://www.rcgroups.com/forums/showthread.php?t=1055820&page=12


I tried it.. and it work well.. simple.. not perfect but simple!

Doc