bcwrench
1 mW
I have been toying with a circuit which uses the CellLog 8S as the main character.
I am not an electronic wizard. I do enjoy soldering stuff on a perf board or slapping something together on a breadboard. If I let the smoke out I learned something and it didn’t cost an arm and a leg.
I would appreciate those of higher schooling to look over what I present here and tell me if I am all wet.
I believe this circuit will draw less that 60ma when powered by the battery pack. In my case this is a pittance as the original scooter have incandescent lights that drew more than 4 times (4.6A) the current my LED lights use.
There have been so many discussions on the pros and cons of BMS’s that I have become numb to them.
I however am a data and gadget nut. I think the CellLogs give the best bang for the buck over any BMS because it can record data. I have yet to see a BMS for the scooter and bike crowd that does just that at any price. BMS’s seem to be good at only doing one thing, and that is saying Oh, oh!
My concerns are of LV and HV observation (I try to elude the word control) along with the data recording. AndyH and others have spent a lot of time doing tests on the bench, which has been most enjoyable and informative, if not thought provoking. I applaud them for all their efforts and time helping others. Things at times have gotten a little heated, especially when polar opposite theory and/or fact is presented. But at least we don’t throw down the gloves and start swinging as in some arenas. In the automotive performance world the phrase often quoted is; “you race cars not dynos”. Therefore test data under fire is the best data .
I have a small Motorino scooter that I have spent more time and money on than I care to calculate. Put it this way, the only parts that are untouched are the front wheel, tire and suspension, the rear shocks and tire and the swing arm. These things become addictive.
The circuit below came about from my interest in the CellLog BMS that that was around for a very short period. I felt it was not explored to its fullest.
First and foremost I wanted something that would allow me to observe trends. This is one thing that a BMS cannot do. They can only react to opposite end, outside limits. They are akin to closing the gate once the cattle get out. Another comparison here would be a tachometer in a car. Since your can see the trend you can be ready for the next gear. Ever tried shifting a car with a rev limiter only?
Some are purchasing the CelllogM rather that the CelllogS which has data record ability. This alone is worth the addition Cocos as it can reveal problems through trend plotting. Looking at live data or just pages of numbers reveals little as it overwhelms to mine. But record it, dump it off-board, graph it and a whole new world opens up.
The old adage, “a picture is worth a thousand words” could never be truer.
I have not yet finished this project as I am looking for feedback as to how it could be improved or more importantly where there may problems lurking.
The circuitry above allows me to use the CellLogs during riding and provides the LVC that reduces throttle load when initiated. It also allows the use of the CellLogs during charging as the HVC.
· The ignition switch (Ign SW) in the above circuit is drawn for clarity and is actually placed in front of the 12V converter and switches pack voltage to it. This part of the circuit controls relays 1 and 2.
· SW1 is not really a switch per se. It and the battery it connects to is actually double contact socket into which an outside power source is plugged. I have drawn the circuitry as such for clarity of operation. In the charging mode the CellLogs are brought to life with this outside power source. The mainframe of the scooter is totally isolated as it is switched off during charging. In an emergency all one has to do is pull out the connector, charging halts, period.
· SW2 emulates the output side of an Opto-Isolator. My system has 16 - 40AH TS cells and I hope to eventually go to 20. I have set up 3 CellLogs so that I can go to the next level.
· Since the isolators are not electrically common to the CellLog circuitry I have tied the 3 isolator outputs to a common circuit.
· Relay 1 is used to activate the TPS override while riding therefore it occupies the NO contacts of the relay. The NC contacts are connected to the Opto-Isolators.
· Relay 2 is a small reed relay that switches the CellLogs on via pin #1 on each CellLog unit.
· Relay 3 is a pair of BOSCH relays that switch the low voltage side of the charger. From what I have seen, many of the charge controlling systems in BMS’s switch the AC side of the charger. I find this very strange as it entails additional wiring outside of the bike and adds IMO a lot of extra “stuff”. I have found no explanation as to why it is done. Sort of takes the phrase “thinking outside the box” to a different level. Some may say that switching the LV side is a problem due to the amperage being higher. This may be a problem with trying to switch a 100A load but I am not aware of any chargers for scooters that slam out that much.
The only thing that the CellLogs cannot do is balance. Since I have personally had bad karma with distributed type balancers that have cost me cells, I choose not to use them. Neither do I have room for a centralized system. Again they are reactive rather than proactive. My cells are hidden under a large custom storage box, a floor and in battery box. Access to them is a 1.5-hour affair. The balancers can’t tell me anything if they can’t be seen. By the time my CA (great product from Justin, BTW) alerted me to some sort of goings on it was too late.
Along with this CellLog circuit I have also upgraded my system with a separate 20ga harness and cannon connector on the kick panel that allows me to hook up a 21-pin BOB. This allows me to double check the CellLog readings at a moments notice without opening or removing anything. It also allows me to attach a Voltphreak single cell charger to an offending cell(s) if it/they just need(s) a little nudge to balance up.
Inside the storage box I have installed terminal blocks with a separate 14ga harness so that I can use a lab power supply to balance things even quicker if need be. It also allows the implementation of removable balancing system if need be as I am still not convinced of the top balancing theory. With the ability to record load and charge profiles via the CellLogs it allows me to determine when I need to perform battery maintenance. This should pretty much obsolete the balancer issue for me.
There are those that say the balancers make it easier to charge as you don’t have to sit on the charging bench and wait. I find this rather ironic as many also say to NEVER hook up the charger and walk away.
Where I find balancing is a problem is when maintaining large capacity cells. Most of those I see using balancers without problems are those with small packs. The balancers can easily handle the shunting currents (<1A) that come with these small packs. Now drop in a 100AH cell and try and handle the currents they use during charging. With my 40AH cells, by the time I found out that my balancers were flakey, 3 cells had gotten over 4.3 volts. As an experiment I used a carbon pile on two of the cells during charging. I found that I had to shunt over 3 amps in order to keep the cell voltage to under 3.70V. Yes they are bad.
Now put that up against a 40 or more cell pack and it spells fire, as Neil Young found out.
Again balancers and BMS’s are reactive components. I want something better. I come from a performance background and races are never won by reacting to bad events. They are won by recording data under fire and analyzing the results. Then formulating strategy and procedure to avoid the reactive events.
View attachment 4
Notice that relays 1 & 2 have changed state. #1 has switched to throttle de-rate mode and the # 2 has switched on the CellLogs. Wiring for relays 2 & 3 has been left out for clarity and for the most part is redundant.
SW1 is in the off position (charging plug not in use) as the system is set for ride mode. The LED above SW1 and Relay 3 are inactive.
SW2 (the Opto-Isolator) pulls the throttle low to reduce engine load and hopefully save the cell(s) afflicted with low voltage.
The one thing that the CellLogs can do that no BMS that I know of can do is monitor for voltage difference between cells.
What a novel idea! Rather than wait till the LV flag is thrown, it lets you know something is happening well before the, you know what, hits the fan. You gotta give a high five to RC crowd for this one.
Since Relay 1 is unpowered it defaults to the Opto-Isolator. The remote power has been connected (depicted by SW1 being on). The LED lights and the transistor goes into saturation. This energizes Relay 3 and allows the charging process to begin.
The remote power also engages Relay 2 to switch on the CellLogs. Before plugging in the charger you can set the CellLogs for recording.
A reverse biased diode prevents Relay 1 from being energized by the remote power. It also isolates the Opto-Isolator from any spiking that Relay 1 may cause when turned off. The spike will actually dissipate by completing a circuit through Relay2 and through to ground. Relay 2 because of its high impedance does not pose a spike threat.
The second diode also prevents any possibility the charging relay from being engaged with the key on and remote power not connected
View attachment 1
Relay 1 is in default position. The Opto-Isolator is in the on position due to the CellLog sensing a HV event. The Opto pulls the voltage for the LED and the NPN transistor low. This opens Relay 3 to discontinue charging. To confirm HV condition one only has to look at the LED.
Since the remote power is keeping the CellLogs turned on and they are recording, the HV event can be captured.
With further data recording during discharge and charge cycles one could easily weed out an errant cell.
All the testing of Ri, AH and WH under steady state lab conditions does give valuable insight but is nowhere near as accurate at data gathered under actual load (accel/decel) and transient conditions.
The only thing that happens is a HV signal is not executed by the Opto-Isolator. One may get flustered at that but in reality if one is so thoughtless to switch on while charging and drive away, one deserves what one gets. I my case the main charging connector is located under the seat. Do I need any more confirmation that I can’t ride the scooter?
I believe the CellLog can outperform a BMS because it can be configured they way the operator wants it.
· Does a BMS allow one to select their LV limit? Very few if any, easily.
· Does a BMS allow one to select their HV limit? Again, very few if any, easily.
· Does a BMS allow one to select voltage difference between cells as a monitor? None that I have seen.
· Does a BMS allow one to record data to offload for analysis and trend plot? Maybe but not in my price range.
· Does a BMS allow real-time display? Some, but again not in my price range.
CellLogs coupled with a Cycle Analyst would be the cream of the crop for battery maintenance IMO.
BTW, I could not have put together this circuit and do the calculations without the aid of Circuit Simulator. It is a JAVA applet that allows you to design and test circuitry. It figures out component voltage, amperage and power by allowing you to change variables. It has a scope function for looking at waveforms as well. Not a full blown SPICE program but as someone that appeared in the past on this forum would say, “close enough for government work”.
You cannot save files but can export them as text file which can be imported back in. You can find it at
http://www.falstad.com/circuit
I am not an electronic wizard. I do enjoy soldering stuff on a perf board or slapping something together on a breadboard. If I let the smoke out I learned something and it didn’t cost an arm and a leg.
I would appreciate those of higher schooling to look over what I present here and tell me if I am all wet.
I believe this circuit will draw less that 60ma when powered by the battery pack. In my case this is a pittance as the original scooter have incandescent lights that drew more than 4 times (4.6A) the current my LED lights use.
There have been so many discussions on the pros and cons of BMS’s that I have become numb to them.
I however am a data and gadget nut. I think the CellLogs give the best bang for the buck over any BMS because it can record data. I have yet to see a BMS for the scooter and bike crowd that does just that at any price. BMS’s seem to be good at only doing one thing, and that is saying Oh, oh!
My concerns are of LV and HV observation (I try to elude the word control) along with the data recording. AndyH and others have spent a lot of time doing tests on the bench, which has been most enjoyable and informative, if not thought provoking. I applaud them for all their efforts and time helping others. Things at times have gotten a little heated, especially when polar opposite theory and/or fact is presented. But at least we don’t throw down the gloves and start swinging as in some arenas. In the automotive performance world the phrase often quoted is; “you race cars not dynos”. Therefore test data under fire is the best data .
I have a small Motorino scooter that I have spent more time and money on than I care to calculate. Put it this way, the only parts that are untouched are the front wheel, tire and suspension, the rear shocks and tire and the swing arm. These things become addictive.
The circuit below came about from my interest in the CellLog BMS that that was around for a very short period. I felt it was not explored to its fullest.
First and foremost I wanted something that would allow me to observe trends. This is one thing that a BMS cannot do. They can only react to opposite end, outside limits. They are akin to closing the gate once the cattle get out. Another comparison here would be a tachometer in a car. Since your can see the trend you can be ready for the next gear. Ever tried shifting a car with a rev limiter only?
Some are purchasing the CelllogM rather that the CelllogS which has data record ability. This alone is worth the addition Cocos as it can reveal problems through trend plotting. Looking at live data or just pages of numbers reveals little as it overwhelms to mine. But record it, dump it off-board, graph it and a whole new world opens up.
The old adage, “a picture is worth a thousand words” could never be truer.
I have not yet finished this project as I am looking for feedback as to how it could be improved or more importantly where there may problems lurking.
The circuitry above allows me to use the CellLogs during riding and provides the LVC that reduces throttle load when initiated. It also allows the use of the CellLogs during charging as the HVC.
· The ignition switch (Ign SW) in the above circuit is drawn for clarity and is actually placed in front of the 12V converter and switches pack voltage to it. This part of the circuit controls relays 1 and 2.
· SW1 is not really a switch per se. It and the battery it connects to is actually double contact socket into which an outside power source is plugged. I have drawn the circuitry as such for clarity of operation. In the charging mode the CellLogs are brought to life with this outside power source. The mainframe of the scooter is totally isolated as it is switched off during charging. In an emergency all one has to do is pull out the connector, charging halts, period.
· SW2 emulates the output side of an Opto-Isolator. My system has 16 - 40AH TS cells and I hope to eventually go to 20. I have set up 3 CellLogs so that I can go to the next level.
· Since the isolators are not electrically common to the CellLog circuitry I have tied the 3 isolator outputs to a common circuit.
· Relay 1 is used to activate the TPS override while riding therefore it occupies the NO contacts of the relay. The NC contacts are connected to the Opto-Isolators.
· Relay 2 is a small reed relay that switches the CellLogs on via pin #1 on each CellLog unit.
· Relay 3 is a pair of BOSCH relays that switch the low voltage side of the charger. From what I have seen, many of the charge controlling systems in BMS’s switch the AC side of the charger. I find this very strange as it entails additional wiring outside of the bike and adds IMO a lot of extra “stuff”. I have found no explanation as to why it is done. Sort of takes the phrase “thinking outside the box” to a different level. Some may say that switching the LV side is a problem due to the amperage being higher. This may be a problem with trying to switch a 100A load but I am not aware of any chargers for scooters that slam out that much.
The only thing that the CellLogs cannot do is balance. Since I have personally had bad karma with distributed type balancers that have cost me cells, I choose not to use them. Neither do I have room for a centralized system. Again they are reactive rather than proactive. My cells are hidden under a large custom storage box, a floor and in battery box. Access to them is a 1.5-hour affair. The balancers can’t tell me anything if they can’t be seen. By the time my CA (great product from Justin, BTW) alerted me to some sort of goings on it was too late.
Along with this CellLog circuit I have also upgraded my system with a separate 20ga harness and cannon connector on the kick panel that allows me to hook up a 21-pin BOB. This allows me to double check the CellLog readings at a moments notice without opening or removing anything. It also allows me to attach a Voltphreak single cell charger to an offending cell(s) if it/they just need(s) a little nudge to balance up.
Inside the storage box I have installed terminal blocks with a separate 14ga harness so that I can use a lab power supply to balance things even quicker if need be. It also allows the implementation of removable balancing system if need be as I am still not convinced of the top balancing theory. With the ability to record load and charge profiles via the CellLogs it allows me to determine when I need to perform battery maintenance. This should pretty much obsolete the balancer issue for me.
There are those that say the balancers make it easier to charge as you don’t have to sit on the charging bench and wait. I find this rather ironic as many also say to NEVER hook up the charger and walk away.
Where I find balancing is a problem is when maintaining large capacity cells. Most of those I see using balancers without problems are those with small packs. The balancers can easily handle the shunting currents (<1A) that come with these small packs. Now drop in a 100AH cell and try and handle the currents they use during charging. With my 40AH cells, by the time I found out that my balancers were flakey, 3 cells had gotten over 4.3 volts. As an experiment I used a carbon pile on two of the cells during charging. I found that I had to shunt over 3 amps in order to keep the cell voltage to under 3.70V. Yes they are bad.
Now put that up against a 40 or more cell pack and it spells fire, as Neil Young found out.
Again balancers and BMS’s are reactive components. I want something better. I come from a performance background and races are never won by reacting to bad events. They are won by recording data under fire and analyzing the results. Then formulating strategy and procedure to avoid the reactive events.
View attachment 4
Notice that relays 1 & 2 have changed state. #1 has switched to throttle de-rate mode and the # 2 has switched on the CellLogs. Wiring for relays 2 & 3 has been left out for clarity and for the most part is redundant.
SW1 is in the off position (charging plug not in use) as the system is set for ride mode. The LED above SW1 and Relay 3 are inactive.
SW2 (the Opto-Isolator) pulls the throttle low to reduce engine load and hopefully save the cell(s) afflicted with low voltage.
The one thing that the CellLogs can do that no BMS that I know of can do is monitor for voltage difference between cells.
What a novel idea! Rather than wait till the LV flag is thrown, it lets you know something is happening well before the, you know what, hits the fan. You gotta give a high five to RC crowd for this one.
Since Relay 1 is unpowered it defaults to the Opto-Isolator. The remote power has been connected (depicted by SW1 being on). The LED lights and the transistor goes into saturation. This energizes Relay 3 and allows the charging process to begin.
The remote power also engages Relay 2 to switch on the CellLogs. Before plugging in the charger you can set the CellLogs for recording.
A reverse biased diode prevents Relay 1 from being energized by the remote power. It also isolates the Opto-Isolator from any spiking that Relay 1 may cause when turned off. The spike will actually dissipate by completing a circuit through Relay2 and through to ground. Relay 2 because of its high impedance does not pose a spike threat.
The second diode also prevents any possibility the charging relay from being engaged with the key on and remote power not connected
View attachment 1
Relay 1 is in default position. The Opto-Isolator is in the on position due to the CellLog sensing a HV event. The Opto pulls the voltage for the LED and the NPN transistor low. This opens Relay 3 to discontinue charging. To confirm HV condition one only has to look at the LED.
Since the remote power is keeping the CellLogs turned on and they are recording, the HV event can be captured.
With further data recording during discharge and charge cycles one could easily weed out an errant cell.
All the testing of Ri, AH and WH under steady state lab conditions does give valuable insight but is nowhere near as accurate at data gathered under actual load (accel/decel) and transient conditions.
The only thing that happens is a HV signal is not executed by the Opto-Isolator. One may get flustered at that but in reality if one is so thoughtless to switch on while charging and drive away, one deserves what one gets. I my case the main charging connector is located under the seat. Do I need any more confirmation that I can’t ride the scooter?
I believe the CellLog can outperform a BMS because it can be configured they way the operator wants it.
· Does a BMS allow one to select their LV limit? Very few if any, easily.
· Does a BMS allow one to select their HV limit? Again, very few if any, easily.
· Does a BMS allow one to select voltage difference between cells as a monitor? None that I have seen.
· Does a BMS allow one to record data to offload for analysis and trend plot? Maybe but not in my price range.
· Does a BMS allow real-time display? Some, but again not in my price range.
CellLogs coupled with a Cycle Analyst would be the cream of the crop for battery maintenance IMO.
BTW, I could not have put together this circuit and do the calculations without the aid of Circuit Simulator. It is a JAVA applet that allows you to design and test circuitry. It figures out component voltage, amperage and power by allowing you to change variables. It has a scope function for looking at waveforms as well. Not a full blown SPICE program but as someone that appeared in the past on this forum would say, “close enough for government work”.
You cannot save files but can export them as text file which can be imported back in. You can find it at
http://www.falstad.com/circuit
