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

[2007blueprius]

I'll find some time and get it all back together, it snowed again and a bit busy with work at the time,
I'll see about taking those measurements and confirm, judging by Jasons experiments it may be in the order of millivolts, a very low voltage/current, it's likely a very low signal of some sort, kind of like the throttle, I'll get back to you on that, I'll see about taking the cover off the converter and have a look at what those wires go to also. it will be a minute.
the other question is, I have a remote led from my charger to the outside of the car, red charging/ green CV mode, I was considering running another from the zephyr, parallel to the one on the board, and install it beside on the bumper, these don't draw much do they? should not be an issue running 2 status LEDs? or the longer wires resistance?
where do you guys get clear polycarbonate boards from? I need a piece for the zephyr in the mean time, is that the same as plexiglass as in homedepot?, I guess anything clear would do.

 

[fechter]

You could try just putting another LED in parallel with the one on the board and see if it works. Best if you use an identical LED. Otherwise, you'd have to come off the board ahead of the 2.2k resistors that feed the existing LED and use another pair of 2.2k resistors in series with the second LED. If you just parallel the LED, the power consumption won't change. If you add resistors and come off U2, this will double the power drain, which may cause some heating of the regulator transistor Q1. As long as Q1 doesn't get too hot, it will be fine.

 

[Zenid]

It's far too complex a solution for most people to attempt.

If you're careful and have basic soldering skills, a board like this is not rocket science to put together. The last project I had done before this was a speaker-phone circuit board when I was 11 years old, but I managed to get my Zephyr assembled with very little difficulty. In fact, the beauty of these boards is that they're from regular, off-the-shelf components you can get at any component suppliers, and if something goes wrong, you just replace the duff part. It's true, though, that not everyone can solder well enough, so any neophyte would be best advised to practise a lot first.

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

Yes, this is a niche market, but since I took a chance on having a batch of 20 made up 2 months ago, 13 of them have sold, so I think the demand speaks for itself. As for the power pack, if you buy some LiPo or LiFePO4 cells from a reputable supplier, and put them together in a sensible way, then you WILL have a workable power source - no 'hoping' about it. The power pack is the easiest part of it, as all you have to do is wire it up to the BMS with a suitable plug/socket arrangement, like the ATX connectors I used on mine.

I should add that I've been running mine for nearly 3 years and it's performed flawlessly once set up right. Until 6 months ago I was charging it at 9A, but now I'm up to 13.6A @ 85.6V using an EMC-1200 charger.

 

[tobyjug]

Hi, I have just taken the plunge and purchased 2 Boards from Zenid. Have managed to order all the parts on the BOM, however, the 32 off Rx04/05 "6.2ohm 3 watt resistors" are on back order from Mouser till the end of Feb. Could I substitute a 2 Watt or 5 Watt for these? If not is there a suitable alternative?
Apologies if this is a dumb question, but its about 25 years since I put my last Heathkit together and I have forgotten just about everything I ever new about Ohm's law

Tobyjug

 

[fechter]

Yes, you could change those. You might try looking for a different resistance but the same wattage. You want to make sure they physically fit the holes on the board. If you get the next higher resistance they have in stock, the balance current will be slightly less, which would result in less heating and slightly longer balancing time. See if they have 6.8 ohms.

 

[tobyjug]

Yes, you could change those. You might try looking for a different resistance but the same wattage. You want to make sure they physically fit the holes on the board. If you get the next higher resistance they have in stock, the balance current will be slightly less, which would result in less heating and slightly longer balancing time. See if they have 6.8 ohms.

Thanks Fletcher. Turns out they had a 6.8 ohm 3 Watt available which is identical in all other respects including its dimensions.
So now the order is in and the fun will begin in a few days. I came late to this party, so I will read through the full 32 pages of this thread. However, I am sure I am going to need some further advice before the project is complete.
Watch this space!

 

[fechter]

Those should work fine.
With my ~11Ahr A123 pack, the balancing time is very short, like about a minute. Larger packs generally take longer.

 

[ejonesss]

I've tried multiple browsers for multiple days and it's still not working for me (from their site).

I put it on my server. Try here:
http://www.zenid.com/docs/Zephyr LiPo-LiFePO4 BMS BOM-v4.4.2.doc

broken link

 

[ejonesss]

all the assembly instructions and no schematic.

you forgot to add the schematic ?

 

[fechter]

all the assembly instructions and no schematic.

you forgot to add the schematic ?

PM (schematic) sent.
We'll see if Zenid can fix the link.

 

[j4cbo]

Could I get a copy of the schematic too? I'm working on a build now, and planning to add an overtemperature shutoff (wired into the HVC cutout as mentioned earlier in this thread) - I'm thinking a MAX6502 and '3904. Is there a logic level power rail around, or will I need to add a 5v regulator also?

 

[fechter]

There is not a 5v supply, only a 12v.
I tested using a TC622, which seemed to work OK. These can run off the 12v supply.
A 5v regulator could be fed from the 'switched 12' line if you wanted to use a 5v device.

See attached instruction sheet. The schematics have been added at the end.

 

[j4cbo]

That instruction sheet says "For [the 47uF caps], the negative side (with the light-colored stripe) goes in the left hole." - that's incorrect for the board I have (recently bought from Zenid, labeled "v4.4a") and contradicts the posts earlier in the thread. Glad I caught that early

 

[fechter]

That instruction sheet says "For [the 47uF caps], the negative side (with the light-colored stripe) goes in the left hole." - that's incorrect for the board I have (recently bought from Zenid, labeled "v4.4a") and contradicts the posts earlier in the thread. Glad I caught that early

I guess it might depend on which way you were holding the board....?

The negative leg of the cap goes to the collector leg of the BD136.

We should edit the instructions for sure. What page was that on?
Edit: I see it. I'll try to reword that so it's less confusing.

 

[fechter]

The instuctions posted 3 posts up has been corrected.

 

[2007blueprius]

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.

so the weather finally broke enough for me to screw whit this project a bit today, I put the system back In the car stock configuration, no Zepher, took it for a testdrive and took some measurements.

with car off system off I got anywhere from 150mv to 300mv shortly after power down, as I kept the leads on the numbers were descending looked to me like some residual power from a capacitor I assume.

with car on system on I got 11.92v across those wires, I switched it to 20mA, and it essentially shorted it and successfully turned it off, the reading was 1.9mA,

I then powered everything down and got 300 mv going down as I held it.

now I lost track, been a while and I know 20ma was the limit on some of those electronics, hope this sounds right, I'll double check again in the morning see if I get the same numbers or I made a mistake reading the meter, but I would be inclined to say I may be able to wire this circuit directly to the Zephyr, maybe not what do I know, not much,70

anyway I'll go back and read through the last few posts see if I can make some sense of it while I wait for some feedback, oh and once shorted and the converter powers of it stays off, until manually turned off and back on, assuming a recharge, so the intermittent function as the Zephyr was designed is irrelevant once first lvc hits you're done, which is fine, I was planning all along not to deplete the pack completely and get better longevity.

in other news, pretty impressive so far a 20 mile hike with 1 stop in the middle 70 mpg, where stock I was getting about 48, again with this car it only gets better the more you go, the first 5 minutes usually average 25mpg and 50 after that, with the kit it went from 25 to 70 to 100, 100 for the first 10 miles 20 minutes, on the way back different route, we started in ev mode so the computer displayed as follows 100, 70, 95, 100 in 5 minutes increments, at the end about 20 miles total about 70 mpg average, now this does not count the electricity, just the gas, I should be about 2 kw into it according to the specs,

I am running it rather modest too at 9 amp output 14 is max, but on a longer stretch say over a gastank I should get close to 80mpg or that is my goal, we'll see how it goes.

let me know what is a good route to take about getting the zephyr to shut down that converter at LVC

 

[fechter]

The optocoupler output is what interfaces to the inverter, so if you measured 20mA, that should be well within the rating of the MCT6.

 

[2007blueprius]

I measured 1.9 ma, I'll double check in the morning to be safe the meter was set on the 20 ma,
12v across the terminals when running, I than moved the probe to amps and switched to 20 ma, that is when it shorted them and gave me a 1.9 ma reading which stayed on or as long as I held the probes on the wires, 1.5v after it had been shorted effectively turning off the discharge while system still on, Idle if you will, just some residual voltage after manually power down.

the optocouplers are the square 8 pin ghismos on the bottom I believe? they would be doing the shorting? what's the worst it could happen fry the first one to trip??

where on the board I would wire those leads? the LVC +- holes, are those the circuit terminals you mentioned that are normally open, and when LVC trips the positive gets pulled to ground? or should I solder them to the alarm labeled terminals? or maybe the 2 of the throttle terminals but which ones? come to think of it likely they are one and the same aren't they?

reality is as you mentioned before the Zephyr will only briefly short the wires as the first cell hits LVC, it goes back to normally open as soon as the voltage recovers, so it's not a latching short, which makes me even more confident as the converter, unlike an ebike controller with the throttle lvc hint, once those wires are shorted it's done until manually reset.

I am way out of my comfort zone here you make it easy to understand but I still get lost sometimes.

George

 

[Diamondback]

hi George.

i can't help as far as interfacing the zephyr to the DC-DC converter, but as yours (and mine for that matter) is still wired to the batteries
the same way as it was originally, even with the DC-DC tripped off by those two wires, there will always be a very small draw from the batteries.

this load is minimal though. it's only the residual current that would keep the large caps in the DC-DC charged up. it's probably a few ma at most.

by the way, my kit is still working flawlessly. im now up to 2.2L per 100km (107mpg) over 950km and still not quite half a tank yet.
i use about 3.7kwh each day. that's with 30% remaining on my SOC meter. if i let the red light go on (10%) it will take about 4.5kwh to charge back up.

Jason.

 

[fechter]

Yes, you would attach to the LVC + and - holes.

If you do not use the alarm feature (don't install jumper), the outputs of the optocouplers are isolated, so there is little risk of anything going wrong. 1.9mA through the optocoupler and 12v open circuit are well withing the ratings of the optos.

The only thing I would worry about is when the LVC trips, the dc-dc will turn off and the cell voltage will immediately rebound and clear the LVC. If the dc-dc latches off and needs manual reset, then it should work fine. If the dc-dc turns back on again after the LVC clears, then the system could go into an oscillating on/off state. This may not be a big problem, but could do some crazy stuff. Normally you wouldn't want to hit the LVC every cycle. The pack level voltage monitor should trip first.

Installing the alarm jumper will connect the LVC (-) output to the pack negative. If the input to the dc-dc has one side connected to pack negative, then you could install the jumper and have the dc-dc connected to the LVC at the same time. You might want to measure between the dc-dc signal connections and pack(-) to see if there is a voltage difference. If the dc-dc input is not zero on the negative side, then installing the alarm jumper could cause a problem.

 

[2007blueprius]

the DC - DC laches off instantly that is no secret, I just barely touched the leads on the amp setting and it tripped and stayed off, that works for me as the cells rebound the short will be removed and the DC-DC is still off, so the Zephyr does not have to hold it off stressing the components.

also keep in mind I set mine up relatively modest with a 2.7v LVC, I monitored it yesterday it draws an almost constant 1/2c, 45A give or take a few out if an 80 ah pack, that's another reason I was not too comfortable with a 2v lvc, on a bike at full throttle you may draw faster and voltage sag will be greater, so realistically even if it hits LVC all the time I am no where near the 2v LVC, the HVC is about 3.55-3.57v very close together I am very happy with that.

The zephyr does not have a total pack voltage sense does it? you are referring to the DC-DC, that is 47V I am told but very unreliable I am told also, again this is all gossip from hobbyists, the way batteries were charged previously it never stood a chance, it was relying on LVC HVC every cycle, 2.7*16 = 43.2, I may hit total pack limit first, but that is another thing, we'll see how it goes I may have a couple weaker cells, 15 and 16 were 200mv lower than the bulk most of the discharge, towards the end where they hit 2.7 the bulk of the pack was still above 3v, it's a used pack supposedly not many cycles, but bad ones with the previous management, we'll see how it goes I got spares.

Ok than off I go, I'll keep the previous BMS as a monitor it has a neat display and logging capabilities, I never tried it but I will once I got this going, I saw graph posted on other forums before you can monitor how the Zephyr is doing, because it will be doing all the work.

now where did I put that solder stuff? PS, I was so excited to get it together last fall I could not find my electronic solder paste so I got impatient and used plumbing paste, that was part of the reason for the funny tests, guy at the local electronic store had a fit with me when I went looking for a cleaner.

 

[fechter]

I think it should work OK then. The Zephyr does not have a pack level LVC. The one in the inverter is supposed to do that. Good on the latching feature.

Plumbing paste is usually pretty acidic and hard to clean off a PCB. The good electronic grade stuff is fairly non-conductive but still needs cleaning. I have some spray flux remover stuff I got from Radio Shack a long time ago. Not sure they still sell it. Pretty nasty solvent stuff, but works very well to remove flux residue.

 

[2007blueprius]

I got it cleaned up back than, it did not look that bad but I think it had something to do with the resistance checks being allover the place , tinkering with it after words it is rather precise, 3.56 - 3.58 v

I wonder if I did this right with the lvc, I installed it today and just for gigs I took a voltage reading off the zephyr lvc terminals and it had a half volt, I know you said it is normally open so I was expecting none, I ended up installing the Zephyr alone, I was hoping to be able to use the other bms in parallel just for the display features and log, but it screws with the Zephyr, in some tests I did all shunts came on while charger off, while charging it all got hot thou I had it well tuned previously, guess something about those electronics not liking each other, so It is a bit odd not having the display, where I could see real data as cell voltage amps in and out of the converter, it was well integrated in that aspect but the important battery management.

the zephyr is not interfering with the operation as the battery bank runs directly into the converter, system comes on I got a couple LEDs up front indicating its working and I can tell by the car battery monitor as it gains power as the system is on, wife logged a few miles groceries shopping all electric no fuel, and the car hybrid battery did not deplete at all just a bit nervous when the battery goes down weather the LVC will trip it or not, likely cell 15-16 will hit 2.7v before the 47v total pack as described previously I may have some work to do on the pack,

Last trip I had the original bms on and I got to about 42 miles in mixed mode before cell 15 was at 2.7 v so I got that to go by we'll see if the system turns off before 42 miles I do have a led indicating system turned off in the cab

we'll see how it goes , I did run an extra lead for the indicator LED to run outside the car on the bumper just like the charger as my charger turns green in CV mode no difference in charge off still green.

I did keep the jst9 connectors on the pack if I should get a couple cell logs?

 

[fechter]

It should be possible to use the Zephyr and the original BMS at the same time, but I think you'd need to run separate balance tap wires directly to the cell terminals. If the balance wires are shared, it can cause the kind of interaction you're describing. You might have a similar issue with CellLogs, but probably not as noticable.

 

[2007blueprius]

I tapped the Zephyr plugs into the original leads, Zephyr closer to cells, one other thing I done was removed the main Anderson pack plugs off the bms and ran jumpers from cell 1 and 16 so the bms recognizes total pack voltage, it may be drawing considerably, as I could hear a hiss that was not there before, also made me think something else vent wrong I had it all dialed in perfect and I tried to top it off just for a test with both bms, sure enough the shunts got real bright and hot, and even after disconnecting charger the shunts were still on where they shouldn't.


Anyway, honestly I don't miss it, it was a nice toy for me to look at and see how it is doing, it is confirmed the zephyr lvc rig up worked, just went through the first cycle, I am pretty sure the lvc tripped it as I mentioned cell 15 and 16 appeared significantly weaker previously, last charge I got 3.5kw at the plug, it is supposed to be a 4kw pack, granted I am working it with a significant margin too, but those 2 blocks rob me quite a bit range I think, last I monitored it most pack was still above 3 v some at 3.15v, I'll check it when she gets home I'm sure those 2 are significantly lower, could be that end cell effect too but still the old bms was not that accurate under load, I'll see how much it takes this time and keep an eye on it .


all in all first test I got 42 miles mixed range this last one 31 but wife ran some chores in electric only, range is good enough for our routine , if I feel ambitious I have the old 4kw pack still, couple blocks are bad for sure but I could rig up a 7kw pack, not sure I need it thou, this is pretty good for our routine, of anything I'll see about replacing those 2 blocks, at the limits I run should get quite a few cycles, rumor has it 3000 to 5000 cycles ( I recon about 70% DOD ) that's about 10 years? car is an 07, lifetime pretty much.

same with the converter, it is set at 9 amp, supposedly it is capable of 14 and you can push the economy versus range, however I was watching the display and it only drew when on load, coasting barely an amp so not sure I should bother with that either, most that done it were running it in EV only which we do also and it doesn't keep up, but it is only good to 35 mph not much use with our routine, those ones that maxed up the converter also have stories of it frying, it does get hot sometimes.


so enough of this I am very happy with the outcome, even as is it took a while to balance the first time , much faster the second, I think the pacific EV bms was good for 350ma balancing , I could not imagine the first cycle , it may have taken the whole week by itself with these large cells, and that was the largest one I came across, most others were 70-130ma?

and to think you guys developed this for a bike maybe a half kw pack, total overkill, if I were to rig up the 7kw pack, I wouldn't hesitate to use it, I believe early on the thread one asked if to trust the Zephyr with $7k worth batteries, I wouldn't trust anything else honestly, you know what they say if you want it done right DIY, and if you are electronically impaired ask Richard, I am a lot more confident in this than anything on the shelf, and there isn't much to go wrong either, the instructions mentioned some special care, I'm a grease monkey, just today I changed dogbones on a semi and welded a bumper, not sure what the electrostatic warning was about but as mentioned I used plumbing flux and solder, than I scrubbed it with oven cleaner and a toothbrush ( sacrilege ) followed by electronic cleaner of course, actually I tried that first but it wasn't doing anything. I dropped it a few times from about 4 feet high, unintentionally of course, and I was using some temporary jumpers for test charging and such twice I accidentally short circuit them , part of the clips are welded onto the andersons.

the dam thing is rugged, probably because it is simple, the fancy logging one it replaced, the one with real time display and sim card, did everything but manage the pack, so yeah I don't miss it at all, not to mention the experience, I am still illiterate in this field just not as bad,

PS Richard I'm still good for a couple fiths of your favorite poison, thou you never took me up on it, I am calling you out PM me your paypal, I insist, for tuition if nothing else, don't make me google you, I will find out where you live if I have to mail it to you.