"Zephyr" - Finally, the "v4" Fechter/Goodrum/Hecker BMS...
- Thread starter GGoodrum
- Start date
Zenid said:
The big TVS diodes go across the FET sorce-drain. These are optional for guys that want to run higher voltages, like 100v and up. It won't hurt to install them. The 2 pin header was intended to go in the EOC disable holes to allow plugging in a jumper for testing. This is only handy during testing. Optionally, it could be used in any of the other connection points to allow a plug.
chxs said:
When the pack is connected, voltage will feed back through the body diode in the charge control FET. This should enable the charger. For testing, you'll need to connect the pack long enough to fire up the charger. You can make this connection with a 1k or so resistor to protect against blowing anything (other than maybe the resistor). If there is a miswire or some problem, the 1k resistor might burn up, so don't hold onto it with your fingers. A light bulb can also be used.
Zenid
100 W
Thanks, but where is "across the FET source-drain"? I can't find any markings on the board to indicate where this might be.fechter said:
I'm getting close to be able to test this now, however I'm seeing equipment specified in the test procedure that I (and probably a lot of people) just don't have access to:
"Use a current-limited bench power supply set to around 100ma {...} slowly turn the voltage up to around 20V"
I don't have one of these. Can you offer an alternative way of testing this based on equipment we're more likely to have, such as the PSU that we intend to use with it?
hunter said:
Some meters will read differently than others on this test, so it's not exact. Try measuring both ways (reverse the probes). If one channel reads drastically different than the rest, that would be something to worry about. The other thing is to just double check that all the parts are inserted the right way. There has been some confusion on this and we're trying to get the instructions updated.
Zenid said:
Don't worry about the TVS diodes unless you're running 100v or more. There are holes on the board for them near the FETs, but not marked. As I recall, it goes over the FET body.
So, if you don't have a bench power supply, use the charger with an incandescent light bulb in series for inital testing. The idea is to limit the current during testing to avoid blowing up anything if there is a problem with construction, yet provide enough for the board to work without cells. I'd suggest a 120VAC, 100W bulb if you have one, or at least 60W if you're running 48v or more. For lower voltage systems, it might work better to find 24v bulbs or a couple of car tail light bulbs in series.
If you have a short, the bulb will light up bright. If the board is working normally, the bulb should not light up. You might have a hard time getting all the shunts on with the bulb in place, but you can jumper across some cell circuits to get the rest lit, then move the jumper to test the ones that were bypassed on the first test. Just jumper enough cells to get the cell LEDs to light up.
I'm working on revising the test instructions, as some of the tests are not generally required unless you're having problems.
Zenid
100 W
Thanks a lot! However we have 240V AC over here (UK). I'm using 24s pack, and have ordered an 87.6V, 10A charger as the power supply. Will a 100W bulb suffice for this, or should I use a couple in series?fechter said:
Zenid
100 W
By getting the 'shunts on', do you mean getting all the little LEDs lit? By "jumper across some cell circuits" do you mean shorting the cell connector terminals (the little holes where the wires for each cell go), for example bypassing circuits 9-16 by shorting a cell 9 terminal with a cell 16 terminal?fechter said:
Zenid said:
Right.
I find it easier to use alligator clips onto the ends of the power resistors. You can power it up with no jumper first and see if you can get the cell LEDs to light up. If not, try placing the jumper across one cell and see if that does it. Increase the number of cells jumpered one at a time until you see the cell circuits light up. If you're running 16s cells, I'd guess you should see something light up when you jumper about 4 of them, but it will depend on your light bulb.
If all you have is 240v bulbs, it might take several in parallel or a really high wattage one. You may also be able to use something like a 15W 24V bulb if those are available. This is a bit more risky if you make a misconnection as the current could go higher than the shunts are rated for.
Once you get some lit up, you can measure the voltages across each one that's lit to verify the shunt voltage setting. Move the jumper over after that to light up the ones that were off before.
Most of the control circuit tests can be done without using a jumper.
Gregb
100 W
Gary, could I make a suggestion for future issues: Modify the exit holes in the end plate to fit the 9 pin header plugs (males) so these can be soldered in place for rigidity and the sockets used to extend them to the battery.
Maybe even modify the other holes to suit higher amperage sockets for optional use.
I know this leaves the pins exposed but putting the socket there would be very difficult. and as long as the charger is the last thing connected , no real problem and it is always optional.
Greg
PS and is preferable to the exposed pins being on the battery side.
Maybe even modify the other holes to suit higher amperage sockets for optional use.
I know this leaves the pins exposed but putting the socket there would be very difficult. and as long as the charger is the last thing connected , no real problem and it is always optional.
Greg
PS and is preferable to the exposed pins being on the battery side.
Gregb
100 W
Has anyone noticed and posted that the schematic and board differ in the connection from the + rail to the fuse? On my schematic it shows the + going to the D3 then the fuse. On my board and the board layout diagram it goes to the fuse then the diode. Can't see it will make any difference but they should be the same.
Gregb said:
Yes, that was changed on the board but not on the schematic. It won't really make any difference, but it's good practice to have the fuse first before everything else.
Zenid
100 W
I'm a just finishing off the big resistors that go on the back, and assembling suitable wires to connect the charger, cells and main battery pack terminals. I'm reading 164KΩ resistance across cell terminals (could the discrepancy from 169KΩ be down to me using a 220Ω in place of a 210Ω for Rx08?). If I reverse my probe terminals, though, I get about 140KΩ. My other meter, strangely, gives a reading of around 3.7KΩ, just like Hunter got.
However, I'm still having problems understanding how the board actually works (specifically with respect to all the little bits I soldered on there). Is there an 'idiot's guide' explanation somewhere that explains it in terms that someone new to charger technology can understand? I'm also seeing some little holes on the board marked "LVC", "HVC", "alarm" and "EOC disable" whose exact purpose is still a mystery to me.
I'm going to read through every post in this thread to try and build up a picture inductively, but any links would be appreciated to posts that give a good description of what the electricity is doing in this circuit at each stage of the charging process.
I was also wondering if someone could recommend a suitable 24-pin connector for the system. I was going to use a 25-pin D-connector of the type used for computer parallel ports, but these connectors aren't nearly big enough to accommodate the wires that I'm planning to use. Can someone point me towards suitable connectors that aren't ridiculously priced?
However, I'm still having problems understanding how the board actually works (specifically with respect to all the little bits I soldered on there). Is there an 'idiot's guide' explanation somewhere that explains it in terms that someone new to charger technology can understand? I'm also seeing some little holes on the board marked "LVC", "HVC", "alarm" and "EOC disable" whose exact purpose is still a mystery to me.
I'm going to read through every post in this thread to try and build up a picture inductively, but any links would be appreciated to posts that give a good description of what the electricity is doing in this circuit at each stage of the charging process.
I was also wondering if someone could recommend a suitable 24-pin connector for the system. I was going to use a 25-pin D-connector of the type used for computer parallel ports, but these connectors aren't nearly big enough to accommodate the wires that I'm planning to use. Can someone point me towards suitable connectors that aren't ridiculously priced?
Zenid
100 W
Well I've started testing, and I've hit a snag. I don't seem to be getting any red or orange indication from the master LED since I installed it. I ran a multimeter test in LED mode to verify which leg was RED and which GREEN before I installed it, and both lit up OK, but after I attached it to the board, only GREEN would show when I repeated this test in situ. It couldn't be related to the fact that I used the recommended replacement for the original version which was out of stock, could it?
I'm running a 72V charger (82V output voltage) through a 100W 240V light bulb, and this is proving enough to bring the main LED to life. It goes green and checking the voltage across the HVC terminals gives a reading of just over 10V, which is as it should be according to the instructions. When I short the terminals, the LED goes out completely, indicating again that something is up with "RED mode".
[EDIT] Voltage across the LED legs read as follows:
HVC un-jumpered
LED GREEN leg: 5.4V
LED RED leg: 2V
HVC jumpered (shorted)
LED GREEN leg: Ov
LED RED leg: -10.6V (which would explain why it doesn't light)
Any suggestions?
I'm running a 72V charger (82V output voltage) through a 100W 240V light bulb, and this is proving enough to bring the main LED to life. It goes green and checking the voltage across the HVC terminals gives a reading of just over 10V, which is as it should be according to the instructions. When I short the terminals, the LED goes out completely, indicating again that something is up with "RED mode".
[EDIT] Voltage across the LED legs read as follows:
HVC un-jumpered
LED GREEN leg: 5.4V
LED RED leg: 2V
HVC jumpered (shorted)
LED GREEN leg: Ov
LED RED leg: -10.6V (which would explain why it doesn't light)
Any suggestions?
Zenid said:
It sounds like the LED is working and so is the voltage regulator.
With no cells attached and no current, only the green part should light up.
HVC activation should make the green part go out, so good there.
That's probably about as much as you can do on the control circuit until you actually hook up cells.
You should try to test the cell circuits next, and I like to take voltage readings across each one when they're lit to verify the voltage setting. It's best if they're barely lit.
If the cell circuits test out, then it's time to hook up the cells. Always make sure the charger is completely disconnected anytime you make cell connections. Also pay careful attention to the cell order to avoid miswires. I like to wire connectors first, then double check the voltages from pin to pin to make sure they're in the right order before connecting to the board. If you do manage to miswire a cell connection, it will most likely take out the zener, shorting it.
Zenid
100 W
Thanks! I was following the instructions under Charger Control Section Test on P10, where it said to just power up the charger + and - terminals, and to expect an orange light to flash briefly and red to appear on shorting the HVC terminals.fechter said:
This is what I'm about to do next, as per your instructions and previous posts about my setup. You'll probably be hearing back shortly about the results of this...fechter said:
I'm especially nervous about this part so will be taking special care over this. I plan to do it exactly how you described!fechter said:
I need to make some revisions on the instuctions for sure.
I suppose one other test you can do is to connect your charger direct to the board with no cells, keep the charger voltage down a bit, and put your light bulb across the pack + and - connections. If the pot is turned all the way down, you should get the red part to light up. It should be orange at first, then turn red if you short the HVC terminals. When it's orange, you should be able to turn the pot up and get it to turn green.
I'd suggest testing the cell circuits first though.
I suppose one other test you can do is to connect your charger direct to the board with no cells, keep the charger voltage down a bit, and put your light bulb across the pack + and - connections. If the pot is turned all the way down, you should get the red part to light up. It should be orange at first, then turn red if you short the HVC terminals. When it's orange, you should be able to turn the pot up and get it to turn green.
I'd suggest testing the cell circuits first though.
Zenid
100 W
OK. I'm about to attach the terminals of the pack to the board as per the instructions for Current Sensor Test, P11. I'm very keen not to screw this up and want to make sure that I understand exactly what I'm doing. So some queries:
1) When you say "Dial the EOC adjustment all the way down" is that turning the little variable resistor anti-clockwise? - On mine that is where the resistance is at its lowest setting.
2) When you say "connect the Pack - terminal to the first cell circuit and the Pack + terminal to the top cell circuit", what is the best way of doing this? Is this what you were referring to when you said to use alligator clips on the ends of the resistors?
If so then is it a case of connecting the + terminal to the top of the last resistor at the end (right of the board as viewed from face up) and the - terminal to the top of the first resistor in the first bank?
Like so: http://zenid10.files.wordpress.com/2011/06/rear-1600x1200.jpg
(N.B: This picture is from the back of the board so cell circuits are in reverse sequence )
Perhaps these details could be added to the next version of the instructions...
1) When you say "Dial the EOC adjustment all the way down" is that turning the little variable resistor anti-clockwise? - On mine that is where the resistance is at its lowest setting.
2) When you say "connect the Pack - terminal to the first cell circuit and the Pack + terminal to the top cell circuit", what is the best way of doing this? Is this what you were referring to when you said to use alligator clips on the ends of the resistors?
If so then is it a case of connecting the + terminal to the top of the last resistor at the end (right of the board as viewed from face up) and the - terminal to the top of the first resistor in the first bank?
Like so: http://zenid10.files.wordpress.com/2011/06/rear-1600x1200.jpg
(N.B: This picture is from the back of the board so cell circuits are in reverse sequence )
Perhaps these details could be added to the next version of the instructions...
Zenid
100 W
Do you mean powering my board as I did for my first test (across the charger +/- terminals with the 100W bulb in series) but ALSO connecting the pack to the board's top and bottom cell-circuits through my other 100W light bulb?fechter said:
OK, I just double checked on a board I have here. You should be able to get the RED to light up by shorting the HVC line, even with no load. So you do have some issue with the red side.
Try running it like before and measure the voltage on R10 (the resistor that drives the red LED) or U2 pin 5. It should be near zero to start, then go up when you short the HVC line. See what you get there. You should be getting something like 10v on the side of the resistor away from the LED.
Shorting the EOC disable terminals should also make the red side turn on, but the green will also be on, so it will go orange.
The other test I was mentioning would be with only one bulb across the pack connection and the charger straight to the board. This will test the end of charge current detector. I'd hold off on that test until you resolve what's going on with the red LED and check the cell circuits.
Yes, "all the way down" means anti-clockwise on the pot.
On testing the cell circuits, the outside end of the power resistor closest to the control circuit is the pack (-) connection. On the upper end, the last resistor will be the second to the last cell circuit positive. You need to connect to the tap wire connector for the top cell.
Try running it like before and measure the voltage on R10 (the resistor that drives the red LED) or U2 pin 5. It should be near zero to start, then go up when you short the HVC line. See what you get there. You should be getting something like 10v on the side of the resistor away from the LED.
Shorting the EOC disable terminals should also make the red side turn on, but the green will also be on, so it will go orange.
The other test I was mentioning would be with only one bulb across the pack connection and the charger straight to the board. This will test the end of charge current detector. I'd hold off on that test until you resolve what's going on with the red LED and check the cell circuits.
Yes, "all the way down" means anti-clockwise on the pot.
On testing the cell circuits, the outside end of the power resistor closest to the control circuit is the pack (-) connection. On the upper end, the last resistor will be the second to the last cell circuit positive. You need to connect to the tap wire connector for the top cell.
Zenid
100 W
Ahh! I think I see a problem. I thought the "R" and "G" on the board were referring to the bottom and middle holes of the three holes where the LED goes. Now I see that the middle hole is a square one, and meant for the negative: I got the red and the negative legs the wrong way round!fechter said:
[EDIT] Just to confirm. I've reversed the legs and it is now working as the instructions say (more or less). The LED comes on very briefly orange at power-up then it turns green. Shorting the HVC makes it go red.
In case you were wondering why it didn't seem odd to me that I had to cross two of the legs over to make the diode go where I thought it was meant to, I just assumed that this replacement diode for the recommended one (that Mouser were out) of was laid out differently, and needed fiddling in this way to make it go with the board holes.
I'll leave the next part of the test until tomorrow, as I need to get some more crocodile clips...
Zenid
100 W
Hi. I've made up a lead for the battery pack, and I'm all set to attach the crocodile clips to the relevant resistors. I see what you're saying about having to use the tap wire connector for the top cell, but now I have to rethink things as I can't connect a crocodile clip to this. presumably I should solder in a tap wire, or is there a better way offechter said:
I want to be absolutely certain that I know what I'm doing here, so I appreciate your patience. Could you confirm that:
1) My 24S bank is being attached directly to the board, and not through a light bulb or anything?
2) The negative crocodile clip (from the pack -) can go onto the top of the resistor closest to the control circuit as shown here:
http://zenid10.files.wordpress.com/2011/06/rear-1600x1200.jpg
but the positive crocodile clip (from pack +) goes to the top cell tap hole on the board?, and not the final resistor as shown in in my picture?
3) You say about measuring voltages, but this doesn't seem to be covered in the Current Sensor Test section on P.11, it just mentions about tests with and without the EOC jumper getting cell circuits to light up and checking the behaviour of the main charger LED. What voltages should I be checking here when I have my circuits lit up and the little orange LEDs going?
4) Charger: I'm confused about the stuff about setting the charger voltage. I've ordered a charger set to deliver 87.6V (24 x 3.65V) at up to 10A. Why does this need modifying? I thought this was the exact voltage required for a 24S LiFePO4 bank.
At the moment I'm just using an 84V output charger in series with a 100W 240V light bulb for the purposes of testing. My bank is 76.8V nominal, but is charged to 80.2V at the moment. Theoretically I should still be able to do a useful charging test with just this arrangement, right?
Zenid said:Ahh! I think I see a problem. I thought the "R" and "G" on the board were referring to the bottom and middle holes of the three holes where the LED goes. Now I see that the middle hole is a square one, and meant for the negative: I got the red and the negative legs the wrong way round!fechter said:
[EDIT] Just to confirm. I've reversed the legs and it is now working as the instructions say (more or less). The LED comes on very briefly orange at power-up then it turns green. Shorting the HVC makes it go red.
Good work.
I would suggest forgetting about the EOC current sensor test for now. The fact that shorting the HVC line turns the LED red indicates that all the components in that part of the circuit are working.
On cell ciruit testing, I'd strongly suggest testing the cell circuits BEFORE connecting to the actual cells.
Cell circuit testing can be done without using the control circuit. Run your supply through the light bulb and feed about half the cell circuits. The resistor closest to the control circuit is the Pack Negative connection, so you can hook the negative supply there. The positive connection can be roughly half way up the board. You should see the cell circuit LEDs light up when you apply power, at which point you can measure the voltage across each cell circuit. You might want to write down the voltages. Once you test the lower half, you can move the clips up to test the rest of the cells. The last resistor connects to the negative side of the top cell, but you need to connect to the positive side of the top cell at some point, so you may need to attach a wire or fit something in the hole for the tap wire. Note: the outside end of each power resistor goes to the negative side of the cell circuit.
Zenid
100 W
This is from the pack, right? Or can I just use my power supply/ 100W light-bulb set-up?fechter said:
I already have my pack ready with crocodile clips from the master terminals, should I just run the positive through a 100W bulb like I did with my PSU?
Excuse me being over-cautious, I'd rather be 100% sure than risk screwing things up...
[EDIT] Also, do I need to short out the tap wire connectors between banks (connectors 8-to-9, and 16-to-17), in order to 'join' them. From resistance readings It looks like there is no connection between the cells of different banks.
[EDIT 2] It's getting late here, and I have to be up early for work, so I ever so carefully ran a test. I put the pack through my 100W 240V light bulb and attached the negative to the top resistor closest to the control circuit, and carefully tapped resistors further down the board with the positive.
As I suspected, nothing happens when I touched the positive terminal to resistors on Banks 2 and 3. I touched it to the resistor at end of Bank one ever so briefly and the orange LEDs all glowed, but so did the light bulb a little. I didn't leave it connected because I didn't want to resist burning anything out.
Am I right in thinking that the banks need connecting in the way I described before?
