Hello. I want to be able to have the ability to charge on the go, but I realized my single cell charger method just takes up way too much space and the weight would be nicer if it was lighter (Each single cell charger has its own relatively heavy transformer!). So, I was thinking that the best portable method would be to carry around a dead-simple small CC-CV charger that allows taper voltage adjustment somewhere in the 34 - 37 volt region (It doesn't need to be very powerful - 2A would be fine - small size and light weight is far more desired). The charger would attach to a simple circuit on the batteries that functions as a high-voltage cutoff (HVC). That is, if a cell reaches above a certain voltage, it would cut off the charging current. Does anyone know of any chargers and circuits like this? Since the HVC module would have cell-level basis, it seems also possible for it to detect for low voltage conditions and send a shut-off signal to somewhere else, like the controller. Does anyone know of any suitable circuits? I don't really need a complete BMS with balancing, as I realize that incredibly ups the part complexity and cost - just a dead simple LVC and HVC detector would be nice. I can interface with microcontrollers if it seems like that may be the best solution.
Simple charger suitable for 10s Lifepo4? LVC + HVC circuit?
- Thread starter swbluto
- Start date
dak664
1 kW
- Joined
- Aug 17, 2008
- Messages
- 317
For $1 each cell an ATTiny25v can measure the voltage to ~10 millivolts without any added components. With an infrared LED and appropriate firmware it could send an RC5 [TV remote protocol] message to turn off a store-bought power switch which the charger is plugged into. You'd need to open a window to the pack to allow the infrared to flood the room.
If the micros are on an 8 second sleep cycle, and coupled so each can reset the next up the line, then resetting the first could cause each to send its voltage in turn to a custom RC5 receiver that could log the data or send it to a computer...
If the micros are on an 8 second sleep cycle, and coupled so each can reset the next up the line, then resetting the first could cause each to send its voltage in turn to a custom RC5 receiver that could log the data or send it to a computer...
Ok, I know that my focus seems to be just the LVC + HVC circuit but I think I'll probably want to source the charger first to make it even a possible idea!
Does anybody know of a charger that has a current limit around 2 amps and has an adjustable taper voltage in the range of 34 - 37 volts? I can take a little flexibility with the amp rate, but the voltage range is fairly narrow. It has to have a Cconstant-current and then Constant-voltage operation. It's going to be used to charge lifepo4, and it doesn't really have strict requirements beyond that.
Does anybody know of a charger that has a current limit around 2 amps and has an adjustable taper voltage in the range of 34 - 37 volts? I can take a little flexibility with the amp rate, but the voltage range is fairly narrow. It has to have a Cconstant-current and then Constant-voltage operation. It's going to be used to charge lifepo4, and it doesn't really have strict requirements beyond that.
grindz145
1 MW
I was thinking about building a charger from the ground up (for onboard charging). I found a few charger ICs from Linear Tech but they really didn't do what I wanted to essentially I am just going to use one of their Switcher Regular ICs with a programmable current limit. You could probably use that as your charger and add the LVC/HVC from there and that way you have a small package to start with. You would still need a small transformer of course. Let me know if you're interested in doing something like that maybe we can do an open source charger, I think that would be fun
-Troy
-Troy
grindz145 said:I was thinking about building a charger from the ground up (for onboard charging). I found a few charger ICs from Linear Tech but they really didn't do what I wanted to essentially I am just going to use one of their Switcher Regular ICs with a programmable current limit. You could probably use that as your charger and add the LVC/HVC from there and that way you have a small package to start with. You would still need a small transformer of course. Let me know if you're interested in doing something like that maybe we can do an open source charger, I think that would be fun
-Troy
Sure, an open source charger would be fun! It doesn't need to be "smart", but it'd definitely be wise to design it to be fail-safe.
I think there's probably a switcher IC out there with an adjustable current limit (Could be biased using a resistive divider), and then the taper voltage should be pretty easy to implement using a standard negative feedback scheme on the PWM rate and hopefully the switcher IC's PWM rate or, via some means, the output voltage can be controlled. I think I'm mostly interested in finding the right IC, but it needs to be a topology that offers isolation so a regular buck converter doesn't seem ideal when it comes to minimizing the transformer weight and size (It would be used primarily for isolation and a buck's would operate directly on the 60 Hz AC).
Through the help of those at the dutch-force forums, it seems that the likely candidates are going to be the flyback or a full-bridge design. The current thought goes the flyback topology offers the lowest cost for moderate power levels, but yet the full bridge can be twice as small but the components cost would something like 4x at much.
Now I just wonder if flyback ICs exist...
I've since found "forward" converter technologies which apparently can have a smaller transformer compared to a flyback converter, but it needs an extra inductor and requires resetting the magnetic flux to zero after every cycle which apparently would make it more costly. So, I'm leaning towards a forward converter although a flyback is sounding more n00b-friendly.
I found the flyback/forward converter ICs at linear. I just now have to figure out which ICs might be applicable...
http://www.linear.com/pc/viewCategory.jsp?navId=H0,C1,C1003,C1042,C1113
I think I should check national and maxim for their products.
http://www.linear.com/pc/viewCategory.jsp?navId=H0,C1,C1003,C1042,C1113
I think I should check national and maxim for their products.
I'd just hack an off the shelf power supply. You just need to reverse engineer it enough to find the feedback for the PWM and tap into it. It will probably be best to start with something designed to be a battery charger like a SLA charger, since it will have nice current limiting to start with.
The HVC/LVC can be done with analog parts like my BMS circuit. You can just leave out the shunts. Actually the LM431 is rated for 100ma, so you could have something like 50-75ma shunts for no extra parts. I think Gary was working on something like this for Lipo packs, but it's just a resistor change for LiFePO4.
The HVC signal can lower the current limit on the charger or simply turn it off. Toward end of charge, cells would start hittnig the HVC and the charger would switch off long enough for the wimpy little 75ma shunts to bleed off the excess, then it would turn on again. This would make it start blinking. You could also simply terminate the charge as soon as the first one hits, since at this point you'd probably have 90% of a full charge if your cells are healthy.
I'm still looking at using an ATTiny to do the cell circuits. Programming it to do the HVC and LVC functions should be fairly easy. From other reports I've seen, it looks like it may be necessary to calibrate each one to a known voltage when you set it up to get accurate set points.
You can run the HVC and LVC signals on the same optocoupler for the cell circuit and have a buffer to split out the signals. You would not normally be charging and discharging at the same time. In a system that uses regen, the HVC signal could be used to inhibit regen.
The HVC/LVC can be done with analog parts like my BMS circuit. You can just leave out the shunts. Actually the LM431 is rated for 100ma, so you could have something like 50-75ma shunts for no extra parts. I think Gary was working on something like this for Lipo packs, but it's just a resistor change for LiFePO4.
The HVC signal can lower the current limit on the charger or simply turn it off. Toward end of charge, cells would start hittnig the HVC and the charger would switch off long enough for the wimpy little 75ma shunts to bleed off the excess, then it would turn on again. This would make it start blinking. You could also simply terminate the charge as soon as the first one hits, since at this point you'd probably have 90% of a full charge if your cells are healthy.
I'm still looking at using an ATTiny to do the cell circuits. Programming it to do the HVC and LVC functions should be fairly easy. From other reports I've seen, it looks like it may be necessary to calibrate each one to a known voltage when you set it up to get accurate set points.
You can run the HVC and LVC signals on the same optocoupler for the cell circuit and have a buffer to split out the signals. You would not normally be charging and discharging at the same time. In a system that uses regen, the HVC signal could be used to inhibit regen.
fechter said:
It would be possible to do an active-cutoff with a 36V 2A charger that I have, but the charger is big and the fan is noisy. Would anyone recommend a smaller charger? I have a small 24V 3A charger from ping that's small like a laptop power supply and it doesn't make any noise, so I know something of a similar size must be possible with 36V 2A but I just don't know of any chargers like that. I'm thinking it'd probably be easiest to be on the lookout for a 36V lead charger.
grindz145
1 MW
fechter said:
Are you trying to ruin my fun?
I think we could design it right to spec, and make it programmable so that we could use it for multiple chemistries / capacities.
I was thinking of just using a small transformer, full bridge rectifier, and then a buck. This would allow for alot of flexibility. I will get some schematics and stuff together and start a dedicated thread
grindz145 said:fechter said:
Are you trying to ruin my fun?
I think we could design it right to spec, and make it programmable so that we could use it for multiple chemistries / capacities.
I was thinking of just using a small transformer, full bridge rectifier, and then a buck. This would allow for alot of flexibility. I will get some schematics and stuff together and start a dedicated thread
Are you thinking of using the iron to isolate at 60 hz AC, then rectify it and then pass it to the buck?
What are your specific project goals?
My specific goals are going to be as light and small as possible and also quiet. It seems a flyback is most appropriate for my desired power levels. It's going to be rectified first, and then switched through the transformer at a rather high frequency. I got recommendation from the electronics forum to just hack apart an old working computer power supply, kind of like what Fechter suggested, for things like the high-frequency transformer.
grindz145
1 MW
Basically, I would like to have something a little more high-power and a little more flexible (voltage-range wise). I am not a switcher expert, but it seems like it would be hard to regulate sufficiently with the flyback. The computer power supplies make alot of sense though, they chop the voltage first which makes a lot of sense.
I'll simulate a couple different configurations to try to get a better idea. I know at work we tend to use alot of buck converters for similar applications and frankly, I was going to use it because I'm familiar with them, which is not the best reason...
I would like to push a bit more current though if possible, Its looking like my new pack is going to be somewhere around 70Ah at 48V. Of course 2A would pretty much eliminate the need for charge termination
I'll simulate a couple different configurations to try to get a better idea. I know at work we tend to use alot of buck converters for similar applications and frankly, I was going to use it because I'm familiar with them, which is not the best reason...
I would like to push a bit more current though if possible, Its looking like my new pack is going to be somewhere around 70Ah at 48V. Of course 2A would pretty much eliminate the need for charge termination
grindz145 said:Basically, I would like to have something a little more high-power and a little more flexible (voltage-range wise). I am not a switcher expert, but it seems like it would be hard to regulate sufficiently with the flyback. The computer power supplies make alot of sense though, they chop the voltage first which makes a lot of sense.
I'll simulate a couple different configurations to try to get a better idea. I know at work we tend to use alot of buck converters for similar applications and frankly, I was going to use it because I'm familiar with them, which is not the best reason...
I would like to push a bit more current though if possible, Its looking like my new pack is going to be somewhere around 70Ah at 48V. Of course 2A would pretty much eliminate the need for charge termination
Forward converters are basically an isolated form of buck converters, and 2A at 48V is well within the realm of possibility - that's another design I was looking at, and it seems like it does have better regulation and it's more reliable (Probably better for an SMPS noob like me) albeit as a slightly higher cost - I guess it doesn't matter that much if I'm getting most of the costly components from an old computer power supply. If you're considering of an "ugprade path" for such a large battery, you might want to look at full bridge converters. Figure 8-3 in the "Power supply cookbook" (Can be found online, hint hint) outlines a fairly helpful graph for what topologies are ideal for what output power levels and input voltages.
grindz145
1 MW
Power supply cookbook eh? That sounds intriguing...:-D I will take a look at that. I really have more like 10A in mind (depending on how big it will end up physically)
Oh wait, I just found the IC mentioned on endless sphere! http://www.maxim-ic.com/quick_view2.cfm/qv_pk/5524
It has integrated HVC and LVC for upto 12s which is about all I needed. Seems to be a steal at $4.
It has integrated HVC and LVC for upto 12s which is about all I needed. Seems to be a steal at $4.
grindz145
1 MW
That would work great on your pack itself or as part of a charge termination circuit. It's not the ideal solution to charge the battery up fully. Charge termination needs to be current dependent of course. Nice simple little chip. I have also seen some which have an over-current functionality without using a sense resistor (reads the voltage drop over the FET when it is on) Ill try to dig on of these up, but if you're already limiting the current before-hand it doesn't really matter.
EDIT: I think this would actually be really nice on the battery side of the Konion pack I will be building up...
EDIT: I think this would actually be really nice on the battery side of the Konion pack I will be building up...
grindz145
1 MW
I've noticed so many power management/battery related ICs now list Ebikes in their "applications" section of the datasheet. I feel like ES has alot to do with this...(nevermind that there are entire nations that ride ebikes)
rkosiorek
100 kW
i've been using regular off the shelf cheapo chinese switcher power supplies. most use a TL431 as the voltage reference. i ditch the 431 and instead use a circuit based on a $1 part by ST Micro. the TSM101 is designed as a voltage and current controller. so far i have used it for 24V 4.5A chargers.
rick
rick
grindz145
1 MW
rkosiorek said:
Rick,
That is exactly what I was looking for! With that I can program it with simple discrete parts in a really cheap way. Thanks! Of course that can be just the front end of a sophisticated charge termination circuit as well. Nice little IC.
-T
rkosiorek
100 kW
i'm starting experimenting with a variation that uses the "OVP" signal output of the ANE/PSI/LifeBatt VMS system. full charge will be at 5.0A. when the Over Voltage Protection signal turns on the charger will switch down to 200mA. this will allow the pack to balance without swamping the 200mA shunts.
rick'
rick'
Similar threads
