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

I didn't buy the enclosure yet so no, also didn't buy the tool for the connectors since it was very expensive
 
I did not bother with the tool nor the sugested connectors either, I like the outside connector ideea better, the pins are supposed to be crimp, likely I'll solder them, I think my connectors were 25c each, the anderson plugs $4 ea, my first mouser order was about $95, second $20, shipping included, I did get the extra FET and the box alone was $20either way if you need a bms is penies on the dollar, for what I am doing the next best thing was $550 and the specs are no where close
 
I think I got it, I'll be dammed, didn't think I had it in me, this stuff is a bit too delicate for me, just this morning I changed an airbag on a semi trailer in the middle of a snowstorm, and we got hit good in michigan, and I make that stuff look easy, bit of a change for me to go from 1 inch drive 180 psi impact guns to this huh!
I did start with the wrong foot thou, only one thing bothered me, the first test 16v 0.5a max PS through a dual filament car turn signal bulb, filaments in series, charger + pack -, LED came on red, but no current, so moving on everything else checked out I am super happy, just not sure about the red LED so what do you think, obviously I am a bit out of my league here.
on the other hand shunts measured 3.56, 3.57, only 2 were at 3.58, I was aiming for 3.55 good enough, I am a bit of a perfectionis but I am not going to make a big deal over 0.02v, all the other test checked out just like the book said it will.
so now I have to set the charger up, but I do need a hand there, I have started a thread about it, I have found out what pots are for volt and amp adjustment, there are 3 more pots on it, no ideea what they do, it is supposed to be a CC/CV type, my question is if I hook it to the board which stage is the charger likely to be in ( if it even powers up without cells ) the adjustmens are to be made in CV stage, it is supposed to be set at about 55v, 3.57*16=57.12, if I were on my own I say hook it up as is let it do it's thing and adjust it towards the end this way I know for a fact it is in CV stage, I can monitor the curent and set the eoc than too I recon start with Eoc about half way see what happens, either way it should be a little on the low side now so it shouldn't cook nothing.
got to get back to reading, guess I got some wireing to do tomorow after work, any feedback much apreciated.
George
 
so I got it all together today and put it on a charge, I do have a few questions so far
#1 and #16 led were dimly lit since I plugged in the discharged batteryes, as it charge 16 got lit brightly and I started measuring a voltage across the shunt resistors rather prematurely, I did not make a big deal out of it main led was orange while charging, but I do wonder of that circuit may require attention
#15 reached the top first, and it was all oktill about 55.5v total, that is where my charger was set as factory.
it was apearing to go into CV mode as for the most part amps off the wall were 10.5 and it was slowly coming down to about 9a towards the end there.
now I believe it is in that cyclyng mode LED is red an blinking briefly, 15 and 16 are brightly lit, most are still at 3.4v, quite a difference and I did manually balance it some before plugging it in, I bumped the charger voltage some and not it is blinking on and off not sure if it is the chargers doing or the zephyr.
So far I learned that my charger will only output when it senses a load, the original bms would disconect the batteries and hence terminate charge, it would not reinitiate it untill unplugged, I wonder if the zephyr cycling is causing the charger to cycle, I do not think it is smart enough.
any thoughts are apreciated
sorry for the missspell too excited, I decided to stop the experiment and drain down those 2 high cells granted the Zephyr was handleing it just fine but I did not want to rag on it, cell 16 apears to come on prematurely, they all tested at about 3.57v, but it works, once unplugged from charge the main is still on and orrange but that is it, I tryed to charge it with a turn signal hooked to those 2 cells but it still does not hold it back and cycles, which would normally be fine but I have yet to tune it, end voltage an EOC so I figured best I drain them manually
Update: it actually seems to be doing fine, I may have actually pulled this off, go figure, couple circuits were acting kindof odd but within spec, and the fuss described earlier was just the Zephyr throttleing down the charge, couple of them were way ahead of the pack, I have been helping it along with some turn signal bulbs, give the shunts a rest while charging still, just tuning it now.
question, can the Zephyr drive a relay at the EOC? my charger pulls dam near an amp even Idleing, it has it's own Eoc so to speak terminates charge below a certain draw, but it would be nice to have the Zephyr unplug it rather than draw 1 amp till morning, just a thought, if possible let me know where to hack, i am no where near smart enough
 
It is normal for the end cells to behave a little differently due to voltage drop in the sense wires.
I can't think of an easy way to cut off the AC to the charger at the end. You could possibly use a timer switch like I have on my Christmas lights.
 
you are a genious you know that right? it actually only draws 0.16 amp at rest, took me a while to get familiar with it, no manual was ever available for this charger, I had some inside help as to which pot to turn for the voltage but that is it, turns out the light on the charger turns green once on CV mode but it was still charging. and it stays green even after the zepher shut it down and I disconnected everything else.
I just got done testing and tuning and completed my first charge, gosh what an undertaking, and I am happy to anounce it works as advertised, couple of cell circuits acted up along the way it apeared that the shunts got stuck on, the first one I cleaned the board a bit more had some flux residue still, thought that may have had something to do with it, the second, I just unplugged everything and it reset itself.
boy did it work hard to get that first balanced charge, and I did have it manually balanced first on the bottom end but I don't think that is verry acurate, right of the get go some got ahead, by the time the first shunt hit ( oh by the way that #1 and #16 fluke is gone ) most were still at 3.4v, I ended up bleeding some charging others, and I had to tinker with it all day today, and graduately increse the voltage, it got cooking pretty good at some point, the zepher was handleing it cycling and all just prefered t ohelp it a bit, the results are priceless, it has been about 10 minutes since the green light, they have not sagged much, 3.56v across the board give or take a hundreth, it took a while to balance with this 80 ah pack but my understanding is that next charge wont be so bad and over time it will get more and more consistent. honestly even with all the help I had my doubts I had it in me and I did fudge up a few times, sparks flying and droped it, plumbing flux ( too lazy to go buy the proper stuff ) you name it, this thing is pretty rugged all the warning stickers, static free, sounds like these components are fragile but not really.
over all I am really happy with it I can see how the DIY and custom tuning is not for everybody, probably what shut this horse in the face, but there is just something about a hand built custom tuned item in general, you just can not get it off the shelf counting my time maybe the savings are'nt that significant but otherwise it cost me a third of the next best thing that thing I would need probably 10 of them to keep up with this thing.
Richard, PM me your pay pal email, I owe you a fith of something to kill the headacke I caused, I'm serious about it, it's called RESPECT and gratitude for your initiative persuation, and even after all this you still suport the builds, even roughnecks like myself, I can only hope that someday I get to grow up in some aspect to be just like you.
what can I say I am impressed,
 
OK, good to see you got it working.

The first cycle I ran on my Frankenpack of recycled A123 cells caused so much heat I had to pause it a few times and use a fan to keep it reasonable. Once it got through that, it's been very quick to balance since, about 1 minute or less with minimal heating.

The last versions of the Zephyr boards had a temperature sensor that would automatically pause charging if it got too hot. They were sort of a pain to implement and only come in surface mount. A simple thermostatic switch tied into the HVC line could work the same way.
 
I haven't had a chance to put it through a discharge yet, I live in Michigan and if you haven't gotten the clue yet, I am a grease monkey always have been, but times are changing and I try to keep up, the prius is my nicer car and here in Michigan they are pretty liberal with road salt to melt the snow which we got quite a bit lately, one of my biggest things, is running my older cars in such weather, they are allready rotten from road salt so I am saving my newer ones from it, I went from 50mpg stock prius gas mileage to my old cavalier which I used to think was economical but realisticaly maybe gets half that on a good day, compared to the potential 80-90mpg that this setup has to offer, I have traveled a lot and if you are from the bay area you have no ideea how dirty Detroit is this time of the year, and for anybody like me putting your car ( anything metal for that mater ) through such enviroment, you might as well beat your child, I have worked too hard for this and I might as well take an oxi torch to my rockers, it would do the same as the road salt , so the prius is parked batteries charged and all actually I have the batteries in the house, I know they don't perform well in the cold, not entirely sure how good of an ideea it was to leave them in the cat in freezing temps, go figure took me this long to get a used pack and figure out a way to charge it, and now I can't use it for at leat 3 months, or I will be posting a rocker repair thread soon ( off topic )
http://priuschat.com/threads/just-installed-my-2nd-user-enginer-4-kwh-kit.119495/page-7 this is a link to a thread between me and a couple other stubborned riggers determined to make something happen with this system, the original BMS has a nice display feature/logger, even that is funny, pretty accurate at rest funny under load or charge, I took a picture of the display shortly after the chage cell 19 and 13 I believe reached shunt last and they sagged a bit but the rest were well saturated and pretty dead on.
I was banking that after a few uses they will get int oa pattern like you all spoke of before, it was a trip getting that first ballance, I worked on it alone for a full day, especially with those large 80 ah cells, granted the zephyr was doing it's part, but it was getting hot, so I am sitting there with the wifes hair dryer, cooling it, hooking up carbulbs to the high cells, I interupted the charge several times to let it cool off, hooked up bulbs to the high cells, my rigged up PS to the low ones, mind you I had those bottom balanced manually first like most users of this system do to get by from time to time, but that was not acurate at all couple sells got ahead real fast as soon as I plugged it in,
also I hardly knew anything about my charger ( top secret ), had to dial it in while at it, wont put out without a load ( hooker :lol: ), the EOC pot was the easy one, by the time I got all the shunts to bearly come on I was barely pulling 0.3 amp off the wall, at some point it turned green itself,
batteries have been at rest for a day now, I was debating either throw a charge on those last two cells only, plug it into the zephyr or both, just to equilize them a bit better since I have nothing to do,
either way I am much smarter today, that is the biggest acomplishment here, and I was serious about that fith ( resistors, transistors whatever your cup of tea is ), I don't see a PM yet, I am pretty resoursefull don't make me look you up and mail it, ask Garry he got a call from me last spring when I first got my hands on this system, kindof caught him off side, I had read about the Zephyr years ago by the time I needed one I was a day and a dollar short storry of my life, either way, I look up to all of you's for thaking the initiative and making something happen, I dare anybody to point out a better bms, like all good things this one had to be unsuccessfull, I am no investor but if they ever get out of stock again let me know I may just take one for the team like Zenid just to keep them stocked.
speaking of which, I needed that full 1 amp bleed for this project, but that last one I read of with the resistors arrays, looked rather interesting, you guys had it working before the wifes got involved didn't you? if anything a thought worth considering that may be a bit more feasable these builds are not for everybody, bit too delicate for me, it's a miracle I did not manage to fry one of those components yet, and I still have no Ideea what most of them do.
George
 
hey George...

i note that you say these batteries are charged now, and that you most likely wont be using them for about 3 months... ?

just a tip for you. lithium does not do well if stored fully charged. lithium likes to be stored long term at about 60-70% charged.
storing lithium long term fully charged or discharged is bad for long term health.

Jason.
 
good point, thanks for pointing that out, I have been reading about this technology for a long time this is my first successfull project and allready forgot most the stuff, I may just do a testdrive shortly, I kindof want to anyhow, my other question was and I know I read this somewhere before, what is the story with cold climate operation, at the moment the car is parked outside and I do not see myself cleaning the garage soon which is not heated anyway, the guy I bought this from had a heat pad rigged up, I know the cells generate some heat during the cycle, and the pouches are incased in stainless steel boxex, I am not too concerned about them while charging or discharging but more so while sitting dorment over night Michigan is not the coldest it's mostly in the teens at night, but still I am not sure how those extremes between cold nights and however hot they may get during operation may affect lifespan.
I built the zephyr rather moderate about 0.5v below the normal HVC for lifespan purpose, I am about to throw it on the charger again now that they settled for a day just to even them out some more, and to experiment see how long it takes for all of them to reach shunt level
 
I have followed the threads in this and other topics about the origins and current state of Zephyr, and I have become discouraged that it will ever be useful to more than a few of us looking for a power solution for our bikes.
It's far too complex a solution for most people to attempt. Very few people will be willing to spend the money on a bare circuit board and a basket of loose parts, then spend $500.00 on Lion packs in HOPE of a workable power source after it's all put together many many hours later. That's a leap of faith that I, and many others, are not willing to take. If I can't get it working, I'm stuck with $700.00 worth of bad judgment.
In spite of all the dozens and dozens of topic messages devoted to Zephyr, no practical, bulletproof simple "Heathkit" set of instructions exists. For example, at the sales site, http://www.zenid.com/goodrumfechter.htm, and the documentation at that site, contains not a single illustration of what the finished product with packs installed looks like.
What needs to happen for Zephyr to take off is a complete or near complete solution; either the product including batteries ready to connect to our bikes, or the whole product less batteries ready to accept our (recommended by Zephyr) packs. We also need to know the exact products needed to support the packs, hopefully just a recommended charger.
I won't argue with the diehard DIYers that this attitude is the lazy man's loss. I simply don't have time to learn all the technical details. I want a working end product out of the box.
Barring that, I want to hear a hundred glowing success stories from guys who have taken the plunge. Until the good news stories start rolling in, the risks of failure loom far too high for most of us.
UpRider
 
I have to agree to a point, it is not for everybody, on the other hand at least in my case there is not another BMS out there that even comes close to these specs, in that aspect I am gratefull that those guys too the initiative to get it this far, than again I never backed down from a chalange nor do I expect any guarantees in life, on the cost aspect, I may have reached 1/3 the cost of the next best thing, which I looked up the other day just out of curiosity and it is an assembly of hobby king components, look up paciffic ev, I don't know enough to properly compare them I know this much the zephyrs shunts are rated 3 times as much, I measured even higher, not counting labour ( more like tuition the way I look at it ), also thou it was built for a ebike pack, I am using mine on a 4kw PHEV pack, and I have 3 more KWH worth of cells I may combine with, they overengineered it, I haven't used mine enough to confidently say this but from what I learned so far if I get arround to building a full EV something of the order of 20kwh pack guess what I'll be using for a bms?
to me there is nothing like a hand built any item, I am not form arround here and one thing I noticed, there is this sick mentality arround, that everything should work perfect or otherwise be insured, and majically rebuild itself otherwise, it is rare that one takes a chance in hope of a better solution, let alone initiative these days, and honestly if there is a bms out there that is plug and play for my system, and remotely compares to the zephyr, and it has been used by thousands, and it has 99.9% positive reviews, please point it out, with such mentality there will never be progress, this is still a relatively new field and even mass produced components are somewhat experimental.

George
 
I read somewhere that the lvc alarm can either drive a buzzer or a relay, I don't remember seeing a test procedure for that, this feature was mostly unused as the lvc strategy was to limit the throttle.
long storry short I figured out which wires to short to terminate the discharge on my setup, I can wire them to a 12v NO relay which I was going to use the alarm output to drive, I have a circuit board relay which I can canibalize off the original BMS that used to shut the converter off in an lvc event, or an automotive type such as used for headlights, question is,:
how does the LVC alarm function on the zephyr? I know it has a 12v output, does it come on and stay on in an lvc event, or much like the throttle function it disables as the cell voltage bounce back, is the output strong enough to drive an automotive relay or do I need something with a smaller coil?
and how would I go about testing it?
figured with my PS in a similar maner as I tested the cell circuits and the LEDs I can dial down the voltage till the cell circuits reach LVC and test the output of the alarm, am I on the right track? or I havent learned much after all
 
The LVC line is normally open and gets switched to ground when LVC triggers. There is a second pad on the board that has the regulated 12v supply which can be used to drive the alarm or relay coil. If you attach a piezo buzzer across the two spots, it will beep when LVC hits.

In most setups, if you stop the discharge as soon as LVC hits, the cells will recover and clear the LVC alarm almost immediately. If you continue discharge, the LVC alarm will stay on.

A small piezo buzzer can run off the board power supply. If you want to drive a relay, the relay coil must have a diode across it to catch the inductive spike when it opens. The diode needs to be reverse biased (non-conducting direction) across the coil. Diode can be 1N4001 or similar. Some relays have the diode built in.

The LVC output has a maximum current capacity of 50mA. If you exceed this, one of the TC54 parts will fry. The 12v regulator on the board is also limited. Total dissipation of the regulator is limited to about 1.5W. The board takes about 10mA, so at 50V, you can do about 20mA. It can handle quite a bit more for short periods of time. If your relay coil takes 20mA or less, it should be safe.

If you have a 12v supply somewhere else (car battery), it could be used to power a relay with no limitation on dissipation, just the 50mA max of the TC54s driving the LVC line.
 
so it is doable, and yes the 12v battery is nearby, as usually you lost me a bit in the technicalities, I need a minute to digest it, than again you gave me a hint I may be going about it the wrong way.
All I am trying to acomplish there is a 2 wire comming from my DC-dc converter to the original bms, a remote power down, black and red wire pretty self explanatory probly a low voltage signal, all the original BMS did is short the 2 via a relay, http://priuschat.com/threads/just-installed-my-2nd-user-enginer-4-kwh-kit.119495/page-8#post-1945287 a link a felow tinkerer posted, we were a bit afraid to try it since none are specialists and no diagrams are available and the converter is absolete
reading your post, I am starting to think one of the wires is ground and the other a signal pulled to ground just like you explain it I wonder if a relay is even necessary, I am obviously not smart enough to do the thinking, I'll find some time to put it back together and take some voltage readings, pictures, and go from there
 
Yes, the board already has a relay, so maybe you can drive it directly.

Measure the voltage across the two pins going to the relay.
I found some information on the relay here: http://www.futurlec.com/Relays/JRC-23F-03.shtml

Coil Power: 200mW
Nominal Voltage: 3Vdc
Pick-Up Voltage: 2.4Vdc
Drop-Out Voltage: 0.3Vdc
Maximum Voltage: 4.5Vdc
Coil Resistance: 45ohm

Hmmm... 3v/45ohm=66mA. Unfortunately that's a little too much for a TC54.
Let me think about it....
 
fechter said:
Yes, the board already has a relay, so maybe you can drive it directly.

Measure the voltage across the two pins going to the relay.
I found some information on the relay here: http://www.futurlec.com/Relays/JRC-23F-03.shtml

Coil Power: 200mW
Nominal Voltage: 3Vdc
Pick-Up Voltage: 2.4Vdc
Drop-Out Voltage: 0.3Vdc
Maximum Voltage: 4.5Vdc
Coil Resistance: 45ohm

Hmmm... 3v/45ohm=66mA. Unfortunately that's a little too much for a TC54.
Let me think about it....
Click to expand...
take your time I am a bit bussy myself at the time and the prius is buried in the snow at the moment, I need to find some time to put the kit back together and with the system powered up take a voltage reading between those 2 wires, my initial thought was to scavage the old bms for the relay and the 2 pin conector and wire it to the Zephyrs alarm, that is probably why I should not do the thinking, the other thing sugested was to leave that junk alone since it all is known to be cheap and unreliable and use some other relay, I got tons of automotive type laying arround, it than dawned on me those beeing considerably larger likely they may draw considerably so that's why I am here asking, too bad we could not get our hands on any info on these converters, rumor has it they are modified solar panel inverters
your first statement, "the board" are you refering to the zephyr? that board you see in the picture is the old bms not the converter, the old BMS employed that relay to short the 2 wires
"pins going to the relay " as in the wires from the converter across the contacts, not the coil that should be 5v, I read through the datasheet.
either way with this last mod ( if you will ) the Zephyr would be a direct replacement to the kit, I kept the original jst9 conentors on the battery packs for the original BMS, it has a cool live display and logging capabilities and I thought that might still be cool to keep thou I noticed when dialing in the zephyr even that was not acurate, bit funny too, I am almoast thinking to do away with it all together at this point.
I probably wont use the prius for a couple more months as long as I get it dialed in by than, maybe I should try it a bit sooner in case I fry something I got time to repair it before spring. :lol:

to jason have you confirmed shortin those wires shuts down the converter? without frying anything else? it only makes sense it's got to, but what do I know I'm just a grease monkey, totaly out of my realm here.
George
 
hi george, yep, shorting those two wires from the DC-DC converter will shut it off.
it's still connected to the pack, but it's only drawing milliamps or less. just he leakage current from the large caps inside it.

i have updated the thread on priuschat now and the first full charge and discharge cycle has been completed also.
have to see how it goes over the next few days i guess, though it seems good to go now :)

Jason.
 
I am kind of anxious to get mine together now, thing is I don't use the car at the moment roads are pretty ugly at the moment too much salt and I park the car outside so not sure how good below freezing temps does to lithium, I know they don't perform well so I am storing the pack in the house at the moment, lets see what Richard comes up with as far as shorting those wires, I'll get it wired up tested and ready for spring.
We're using our old car now since it's salt rotten already, used to think it is economical, went from 50mpg to 22, and I am sitting on a 70-80mpg pack, next winter I may just buy a beater prius, this one is still nice.
 
2007blueprius said:
I read somewhere that the lvc alarm can either drive a buzzer or a relay, I don't remember seeing a test procedure for that, this feature was mostly unused as the lvc strategy was to limit the throttle.
Click to expand...
I'm using the lvc alarm AND the lvc throttle passthrough as detailed in my blog entry:
http://zenid10.wordpress.com/catego...-wiring-in-an-lvc-system-goodrum-fechter-bms/

The lvc gives a signal that rises to 12V as the lowest cell reaches a discharge threshold. I routed mine through 3 parallel LEDs on the instrument panel. Now, when the pack starts to get low, the LEDs will start to flicker, at first under high load, but progressively more until they're lit all the time. Then, the throttle starts to fade, gradually getting softer until you're brourght to a halt just above the point where you'll damage any cells. I found this to be an excellent system that's really useful and intuitive.
 
I didn't realize the picture of the board was the old BMS. I thought it was the inverter.

If you're using the MCT6 optocouplers, they're rated for 30mA max, and something like 75V. What would be nice to know is how much current/voltage it takes to trigger the inverter off. If it was running, you could measure the voltage across the 2 wires that went to the relay, and try to measure the current, which you could do by using a multimeter in the milliamp mode and put the probes across the same wires. This should be just like shorting them, and you can see what the current is.
 
the little relay we used is linked above, it's working fine.

there was zero voltage on these lines with the DC-DC on.
Vincent, the installer measured this.

Jason.
 
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
3,523
Santacruz
Santacruz
T
BMS "steam release" post
2
Replies
27
Views
2,359
999zip999
999zip999
fechter
DIY Test Fixture to test BMS Boards
Replies
0
Views
637
fechter
fechter
methods
Black Sheep Technology - TS90_BMS v3
2
Replies
27
Views
3,494
methods
methods
amberwolf
The HI-Lebowski: a Lebowski SMD brain running a zombified Honda IMA Inverter: *a HOW-TO guide*
2 3
Replies
64
Views
15,446
amberwolf
amberwolf
Back
Top