"Zephyr" - Finally, the "v4" Fechter/Goodrum/Hecker BMS...

I'd use the power supply/light bulb setup to power the cell circuits, but using the pack + light bulb should do the same thing. You want to power up about half the cell circuits at once, measure the voltages, then repeat the test on the remaining cell circuits. If the light bulb lights up dimly, that would be expected.

The light bulb you're using should limit the current to a safe level.
 
fechter said:
I'd use the power supply/light bulb setup to power the cell circuits, but using the pack + light bulb should do the same thing. You want to power up about half the cell circuits at once, measure the voltages, then repeat the test on the remaining cell circuits. If the light bulb lights up dimly, that would be expected.

The light bulb you're using should limit the current to a safe level.
Click to expand...
I was eager to get going so I ran with the pack-through-light-bulb setup. I went ahead and joined the banks by jumpering the tap wire connectors between banks (connectors 9-to-10, and 18-to-19), and up to 16 LEDs or so at a time would light up, with the light bulb glowing very dimly. Then I went along as you suggested, covering about 12 circuits at a time, measuring voltages. For the final cell, I just tapped the crocodile clip to the last tap-wire 'node' while I measured the voltage for the final two holes.

To jumper the holes between banks I used a bit of clear tape with a bit of solder braid stuck to it (I used a clipped off resistor leg for the first jumper) and just stuck it down so the tracks were joined :)

All of the voltages I got between the tap wire 'nodes' were very familiar: Every one was between 3.60V and 3.62V :D , with the exception of those between banks, which read 3.65V (9-to-10) and 3.69V (18-to-19) respectively

Are those all within acceptable limits?

So what comes next? Is this where it all gets wired up for a 'proper' test?
 
Those voltages sound like they're in the healthy range. It's normal to get a little higher reading on the end cells due to increased resistance in the wiring.

Yes, it sounds like it's ready for a proper test with eveything hooked up.

I'd suggest first running the charger into the board without the light bulb, but no cells hooked up to put the full charger voltage across all the banks. This is where you want to dial in the charger voltage. When you power up, you don't want all the cell LEDs to light up brightly. If they do, you want to dial down the voltage a bit if that's possible with your charger. The idea is to adjust the charger voltage until the point where the cell LEDs barely glow, or just below that point. It's helpful to disable the EOC for this step, by using the jumper on the board.

If you can't adjust the charger voltage, as long as all the cell LEDs aren't fully lit, you should be OK. If they are fully lit, watch the temps, as heating will be more. It's best to drop the charger voltage to the point where they barely light up.

Once you get that dialed in, then connect the cells. Remove the EOC disable jumper and charge away. I think it would be good to start out with VR1 turned fully anti-clockwise and monitor the charging cycle. When the cells have fully charged, then slowly dial VR1 up until the main LED goes green. It might be OK to set it a little higher. Setting it higher will terminate the charge a little sooner. You may need to play with that one to get it so it turns off at the right time.
 
fechter said:
I'd suggest first running the charger into the board without the light bulb, but no cells hooked up to put the full charger voltage across all the banks. This is where you want to dial in the charger voltage. When you power up, you don't want all the cell LEDs to light up brightly.
Click to expand...
Do you mean connecting the charger to the charger + and - terminals of the board (as I did for the control circuit test) AND attaching the pack to the end terminals of the cell circuits (as I did for the the voltage test). Except doing both of these without the light-bulb attached to restrict the current in any way?

Or am I only attaching the charger to the charger + and - terminals at this point?
 
Zenid said:
Do you mean connecting the charger to the charger + and - terminals of the board (as I did for the control circuit test) AND attaching the pack to the end terminals of the cell circuits (as I did for the the voltage test). Except doing both of these without the light-bulb attached to restrict the current in any way?

Or am I only attaching the charger to the charger + and - terminals at this point?
Click to expand...

A good diagram here would be helpful. Sorry I don't have one handy. Yes, I think you got it. You want to connect the charger to the board AND connect the ends of the cell circuits to the corresponding spots on the control section. The two positive sides just connect to each other on the board. The pack negative connection on the control board can get jumpered to the pack negative connection on the cell circuits. You can use the outside end of the first power resistor for this connection if its easier.
 
fechter said:
The two positive sides just connect to each other on the board. The pack negative connection on the control board can get jumpered to the pack negative connection on the cell circuits. You can use the outside end of the first power resistor for this connection if its easier.
Click to expand...
I'm confused about why the pack negative needs to be jumpered to the negative end of the cell circuit banks. This would imply that the control section is somehow disconnected from the pack. How can this be? Jumpering anything on a board I don't understand makes me nervous. Could you explain what this jumpering is doing?

Thanks.
 
Well I cautiously went ahead and wired up the charger again (shown here attached by ring connectors), without its 'safety-net' light bulb, and it was fine. Then I powered down and attached the battery bank in the way you described, attaching the positive corcodile clip to charger positive terminal, and the negative clip to "pack negative", and using a jumper lead (the yellow one here) with clips to crosswire the pack negative with the top of the resistor closest to the control circuit (the far left one as seen from the top of the board).

test3-640x480.jpg


When I powered it all up, the main LED came on green but it didn't do anything, even when I twiddled the adjuster, and no yellow LEDs came on at all. But when I attached the header to the EOC disable, the main LED turned orange and the charger came to life :)

The FETs become only slightly warm, and everything else was stone cold, not surprising as the charger is the weedy 2.5A one I use for my SLAs. The pack charged until it equalised with the charger, which switched itself off after a while. The voltage of the pack got raised from 80.2V to about 84V.

Presumably no orange LEDs came on because of the low voltage of my charger (84V) with respect to the ideal maximum pack voltage of 87.6 (or 86.4 at 3.6V per cell). And presumably also the charger stayed green without the EOC disable because no cell circuits were attached (is this right?)

This brings me to a point I quite urgently need to address now. I asked for the charger I ordered to be set to 87.6V, but presumably this should be more like 86.4V if it's working at 3.6V per cell max, right? But why does the charger voltage need adjusting at all as part of the callibration you describe? Isn't 86.4V more or less exactly what it should be?
 
Just a word of warning. Check each item you have ordered against what you actually received. I have been having a very worrying time trying to work out what had gone wrong with mine to eventually find I had received a BD139 in my order of BD136s!!!! My fault for not being more careful when checking against my order... :oops: :oops:

PS it wasn't mouser. for Oz people it was jaycar
 
Zenid said:
This brings me to a point I quite urgently need to address now. I asked for the charger I ordered to be set to 87.6V, but presumably this should be more like 86.4V if it's working at 3.6V per cell max, right? But why does the charger voltage need adjusting at all as part of the callibration you describe? Isn't 86.4V more or less exactly what it should be?
Click to expand...

If the charger voltage is too low, some cells may not get fully charged. If the charger voltage is too high, all the shunts will be on at the end of charge like the 2.6 boards. This can work, but there's a lot of wasted heat. The idea is to dial it in so there is minimum heat at the end of the cycle.

As the voltage is increased, the shunt current goes from zero to maximum over a narrow range of voltage. You want to get dialed in to the low side of this range if possible.

For making this adjustment, I was suggesting the setup below:
Connect the charger to the normal spots on the board, and use jumpers to connect the ends of the cell circuits to the control section. No pack attached. EOC disabled. When you reach the threshold voltage, all the cell LEDs will begin to light up. If your charger doesn't have enough voltage, none of them will light up. You can use the end resistor to attach on the negative side of the cell circuits if it's easier.
 
fechter said:
For making this adjustment, I was suggesting the setup below:
Connect the charger to the normal spots on the board, and use jumpers to connect the ends of the cell circuits to the control section. No pack attached. EOC disabled. When you reach the threshold voltage, all the cell LEDs will begin to light up. If your charger doesn't have enough voltage, none of them will light up. You can use the end resistor to attach on the negative side of the cell circuits if it's easier.
Click to expand...
Thanks very much for the helpful illustration! This greatly clarifies things. I take it then, that I should really be asking for the charger to be about 86.4V initially (24x 3.6V), to be fine-tuned via a variable resistor...
 
Zenid said:
Thanks very much for the helpful illustration! This greatly clarifies things. I take it then, that I should really be asking for the charger to be about 86.4V initially (24x 3.6V), to be fine-tuned via a variable resistor...
Click to expand...

Yes. It would be safest to start out with the voltage a little on the low side, then try turning it up slowly while watching the cell LEDs. Once you get them to light up, back off a bit and you should be good to go.

It's not that critical, but keep an eye on heating if they all light up brightly. If you go too high on the voltage, the HVC will kick in.
 
Gregb said:
Just a word of warning. Check each item you have ordered against what you actually received. I have been having a very worrying time trying to work out what had gone wrong with mine to eventually find I had received a BD139 in my order of BD136s!!!! My fault for not being more careful when checking against my order... :oops: :oops:
Click to expand...
I don't see what there is to be embarassed about. They screwed up in a way that is almost impossible for someone to notice. There are about 500 components to go through, you can't be expected to check every one. The only job they have to do is to send you what you ordered, and with the process being totally, digitally automated there really isn't any excuse to let a different component get mixed in with a bunch of others. IMO.

If I buy a bag of peas I don't expect to find any raisins in there...
 
I have been in and out of electronics since 1956. When you skip the basics of not checking everything it is embarrassing ........this includes schematics, wiring diagrams, BOMs (which I call parts lists) test instructions etc etc....I have had to rewrite more than one handbook in my time.....
You don't have to be paranoid, but it helps.... :shock: :shock: :shock: they ARE out to get you..... :mrgreen: :mrgreen:
 
Gregb said:
You don't have to be paranoid, but it helps.... :shock: :shock: :shock: they ARE out to get you..... :mrgreen: :mrgreen:
Click to expand...
I've been properly into electronics since... erm... maybe last year. Despite my cautious nature, I probably have something of a learning curve ahead of me :shock:

I'm starting to understand what you're saying, though, as when I got my parts, I counted that the right number were all there (my idea of 'checking the parts') and found out straight away that they'd miscounted and short-changed me in a couple of packs. I didn't think problems would extend to turning parts into a pic 'n mix though - that should be a shooting offence for component firms IMO :x
 
I've been fooled by the parts several times. Once I ordered some 12v zener diodes, and that's what the package said on it, but they turned out to be 4.5v. Boy, that led to some head scratching....
 
There's something I'm a little confused about. Why does the board need pack master + and - terminals as well as the individual tap wires? The top and bottom tap wires are the same as the masters. Why does it need both? :?

Also, in your setup to calibrate the charger voltage, you detail putting jump leads from the pack + and - terminals to the ends of the cell circuits. Why don't these pack master connections link up to the ends of the cell circuits anyway? What is their function on the board?
 
The reason is there is significant voltage drop in the main wires going from the charge controller to the pack. If they were combined, it would throw the voltages on the end cells off by quite a bit. As it is, the end cells still have an error due to drop in the balancing tap wires, but this current is much lower than the main charge current in most cases. I call this end-cell effect. By running separate wires to the end cells, the error is minimized. The error is not enough to cause any problems unless the tap wire resistance is really high and the current is high. It's best to use fairly heavy (18ga) wires at least on the end cell tap wires. Making the wires shorter helps too.
 
Thanks for the explanation!

At the moment I'm mulling over what power wires to use for the charger/pack + and - tap points on the board. Since the tap holes are the same size as the holes that go through the end-plate, the nominal 4mm diameter wires I have are just the right size for the end-plate, but too narrow for the tap holes on the board (it will fit right through the holes in the board sleeve and all), and yet the 6mm nominal diameter wires I have will fit well with the tap hole on the board, but cannot fit through the hole in the end plate.

What did you and others use for these wires - narrower ones that fit the end plate holes, or thicker ones that fit the tap holes on the board? :?
 
fechter said:
The reason is there is significant voltage drop in the main wires going from the charge controller to the pack. If they were combined, it would throw the voltages on the end cells off by quite a bit. As it is, the end cells still have an error due to drop in the balancing tap wires, but this current is much lower than the main charge current in most cases. I call this end-cell effect. By running separate wires to the end cells, the error is minimized. The error is not enough to cause any problems unless the tap wire resistance is really high and the current is high. It's best to use fairly heavy (18ga) wires at least on the end cell tap wires. Making the wires shorter helps too.
Click to expand...
Is there any reason why the master pack negative can't just be jumpered to the first cell negative, much like the first cell negative can be jumpered to the last cell positive of the preceding bank? Or will this still result in the error you describe?

This would make more sense, as then you only need to feed 8 tap wires for each bank, leaving out three [EDIT: for my 24-cell version] redundant, duplicated ones...
 
Zenid said:
Is there any reason why the master pack negative can't just be jumpered to the first cell negative, much like the first cell negative can be jumpered to the last cell positive of the preceding bank? Or will this still result in the error you describe?

This would make more sense, as then you only need to feed 8 tap wires for each bank, leaving out three redundant, duplicated ones...
Click to expand...

If you use a very heavy wire and it's not too long, then you could get away with that. Ideally you want to use separate wires directly to the end cells to minimize end cell effect.

Smaller wire can be fit into the huge holes on the control board. I just strip a little extra on the end and fold it over to make it fatter. Use lots of solder.

Personally, I prefer one big hole in the end plate that's a mouse hole (meets with one edge) so the end plate can be removed without disconnecting any wires. It's harder to keep track of the wires this way, but labels on each one will take care of that.
 
well that is the nastiest fault I have had for many many years since a canon mil plug developed a partial internal short between two pins..... make sure none of the resistors are touching each other, the insulation is not all it should be. on my channel 2, r101 and R102 touched each other probably when I put the FAm in. they partially shorted and while I thought everything was separated, they weren't. If I was starting from scratch I would recommend every hole that isn't through connected be drilled out. Maybe even them, too. This is not a criticism of the quality of the board and is probably a compliment on the accuracy of the holes size. But this makes them almost impossible to remove some components especially the 3 pin ones which I had to destroy to remove or I would have been cutting a lot of tracks which I prefer not to do. And even with a powerful solder sucker the holes were very difficult to clean. Depends whether you are like me or one of those people where nothing ever goes wrong..... :mrgreen:
 
You must log in or register to reply here.

Similar threads

DotScott
What Should My Settings be? Sabvoton SVMC 96120 and QS205 50H V3
Replies
7
Views
4,712
Santacruz
Santacruz
T
BMS "steam release" post
2
Replies
27
Views
3,893
999zip999
999zip999
fechter
DIY Test Fixture to test BMS Boards
Replies
0
Views
1,059
fechter
fechter
E-HP
    • Like
DIY Newb FAQ?
2
Replies
30
Views
10,082
Low-and-Slow
L
methods
Black Sheep Technology - TS90_BMS v3
2
Replies
27
Views
4,822
methods
methods
Back
Top