The Zivan NG1 revealed

Patrick

100 W
Joined
Feb 10, 2007
Messages
130
The Zivan NG1 is an excellent (albeit expensive) battery charger. 900 watts, three stage, and fully programmable. Unfortunately, the programming can only be changed with a factory supplied microcontroller. However, the end user can change the voltage and current setpoints.

Best to understand the charge profile before using in new applications. The Zivan manual http://zivanusa.com/pdf/NG1.pdf explains some things, but they have left this part out. They use the European convention of W (watts), I (current), U (voltage), a (auto off) and o (float). Two examples are explained here: http://zivanusa.com/ChargeCurve.htm. Chargers are well labeled when built:

cover.jpg

As you can see, this charger has an IUa curve (what we would call CC/CV); so, constant current (I1 = 11 A) until it reaches U1 (56.8 V), then constant voltage (again, 56.8 V) until current tapers to I2 (5.1 A), then shut off.

There is a little more to it than that. Each stage also has a timer associated with it (T1, T2, T3). These are usually just maximums, i.e., each stage will terminate at Tmax if it doesn't meet normal criteria. In the case of NiZn batteries specified on this charger, there is a pause specified of 5 minutes, and that is the second stage. So after reaching CC, it shuts off for 5 minutes, and then starts the CV stage.

There is an LED to indicate each stage. It is progressively Red, Flashing Red, Orange, Green. There is also an optional temperature input, it is nominally set at - 5 mV/C.

So let's set this charger for some lithium cells. A 16 cell pack at Vmax of 3.7 volts/cell should stop at 59.2 V. 11 amps is fine for a larger pack, but it can be adjusted as well. So let's open up the charger. Be VERY CAREFUL! Everything is live, including the heat sinks which can be at mains voltage or above. The capacitors can also hold a healthy charge.

View attachment exposed-pointer.JPG

The trim pot on the right is the voltage setpoint. Turning it clockwise will increase the voltage. Zivans require a load in order to operate - during the soft start (avoids inrush current) they test battery state before beginning charge, so you will need to connect something that will take whatever voltage you're going to set. I think it's easiest to set up your meters, and crank the voltage above where you want it to be. Then start a charge cycle, and when the battery voltage reaches desired level, lower the adjustment pot until it goes into the next stage.

View attachment 1

Of course, run a full cycle to verify your new charge curve is what you want it to be.

But wait - there's more! It might be that your existing curve doesn't match what you want (like the IUI curve), and you'll want to get your NG1 re-chipped. There is a little jumper block next to the MCU on the digital board:

View attachment adjustment.jpg

This allows the selection of one of up to 16 charge profiles. This is usually used to select the 3rd stage current, to allow for AGM finishing of different sized battery packs. However, the profiles can be completely independent, within the confines of your charger. Each charger has a maximum voltage and current, e.g., 48 V and 15 amps, or 72 V and 10 amps (these are indicated by the second two letters of the model number). So with one NG1 you could charge a wide variety of pack voltages and chemistries.

Truly a charger for all seasons.
 
Good topic. :)

I have an NG1 that I got from you guys some time ago, but the profile my version uses seems quite different. Here's my label:

NG1%20Label.jpg


I did try cranking up both the voltage set point, and the max current, but the max voltage I can get it to is 59.08V and it appears that mine has an additional CC stage. I think the way it is supposed to work is that it will only go into the full CC mode, with the max current (i.e. -- about 15A max...) if the pack voltage is below some first setpoint. In this case, the LED should be red. Once the voltage gets above this first setpoint, the current is dropped to about 5.5A, and remains at this CC limit until the voltage reaches a second set point, which appears to be the one controlled by the trim pot. During the "intermediate" CC mode, the LED is yellow, which according to the label, means the pack is at the 80% level. Once the 2nd set point is reached, when the CV mode kicks in, the rest of the profile is "normal", wih the current gradually dropping down to under about 20mA, when the charger shuts itself off.

With the jumpers, are you telling me I can select other profiles that are already loaded in this version of the chip I have, or do these still have to be programmed into the chip first? It would be great if I could at least get one that eliminates this "extra" CC mode because with my a123 and LiFeBatt-based packs, they never get low enough to not immediately go into the secondary CC mode, with the 5.5A CC limit. As it is, that means I've got a $500 5.5A charger. Because of this, I have held back recommending the NG1, unless they can get a version that doesn't have this limitation.

-- Gary
 
Hi Gary,

Yes, I've tested one of those F4EGSN models. F4 = NG1, EG = 48 volts, SN = chip major rev. As you can see on the charger, it comes programmed with the IUIa profile. This is specifically for AGM, with a pulsing current during stage 3. Here is the profile for mine:

View attachment F4EGSNprofile.JPG

It's the IUIa curve, not ideal for lithium:

IUIa-curve.gif

You can see that stage 3 uses a higher voltage than stage 1, and I expect it's that way on your charger. Notice on your label it says Um = 66 V; probably that 3rd CC stage is minimal current , and ways until load gets up to Umax. I'm not the complete expert, but my understanding is that the voltages are programmed in, and the trim pots are used to calibrate the exact voltages.

Does your charger have a little dial on the end, where the LED used to be, adjustable from 0-F (hex)? These are the settings for 3rd stage current. If so, you have different profiles enabled, but frankly, none of them will do you all that much good. You need to get yours re-chipped. Zivan (Zivanusa.com) has curves for lithium, and will know what you're talking about. They might easily charge some extra if you want a bunch of different profiles loaded (different cell counts, voltages or currents, etc.), but if you just want a 16 cell, 15-18 amp charger that should be a cake walk. You'll want an IUa curve.

The 48 volt units can go pretty high. Here's the old (original) lead charger, also Umax of 66 V. You might run one on yours to see what's happening.



Let me know if you have problems.
 
Excellent info...

Do you know if the temperature probe input can be used to throttle the charger? I was thinking of using that instead of the onboard current control on the BMS circuit. If the optocoupler that signals cell(s) are high gets triggered, I'd like to see if there's a way to use that to drop the charging current.
 
fechter said:
Excellent info...

Do you know if the temperature probe input can be used to throttle the charger? I was thinking of using that instead of the onboard current control on the BMS circuit. If the optocoupler that signals cell(s) are high gets triggered, I'd like to see if there's a way to use that to drop the charging current.

As designed, the temp probe just changes the output voltage. And, of course, it's an analog input. I've thought about this as well, but don't know if the programming software has this flexibility. It would be lovely to use it as an analog control of charge current. As a real kludge, I thought about setting a separate profile for low current, enabling it, and cycling power on the charger. Ugly! I will ask the question (may take a while to get an answer). And it might be quite expensive to get the custom programming.


GGoodrum said:
Yes, that profile looks like what I'm seeing, for sure.

So, how do I get this beast re-chipped?

I know a couple of people who are shifting from lead to lithium, and I'm going to try ordering chips for them all. They're running 16 cell, 50 Ah packs (we're waiting to get some boards from you!). I think you can pretty much choose what ever current and voltage setpoints you want, but need to verify this. Ideally, I'd like a few profiles, with CC at max current until U1 (maybe 59.5?), then a choice of settings for the CV stage - maybe 2, 3, 4, 5 amps - something like that, and tapering to 100 - 200 mA. You can PM me with what you think you want. I'll see what's possible, and report back here - I'm a big fan of the Zivans.
 
I actually think the "2-stage" CC mode currently programmed in my NG1 would be the "perfect" LiFePO4 profile if the first trip point voltage could be tweaked up higher, to around 58V. A full 15A until the total of the cells hit 58V, then throttle back the curent to 5A until the 2nd trip point is hit (59.2V, and then just hold the voltage there until the current drops to about 20-30mA, at which point the charger can shutoff.

With the BMS board, though, the current isn't really ever going to get below either 250 or 500mA (depending on the shunt resistor used...), as the shunts wil always bypass this much, so that each cell can have at least this much available in order to "catch up", if it needs to. In this case, maybe the best thing to do is have them also program in a CV mode timer of, say, 40-45 minutes, which should be more than enough balance time for healthy cells.

Anyway, I'm also a big fan of these chargers. For 900W unit it has a very small footprint, and is incredibly light. Definitely my favorite charger, even with its "current" limitation, which is due to the profile programmed in my version.

-- Gary
 
I see the dilemma here, and can't think of a good solution. A timer, sure - but it depends (to some extent) on cell size and state of balance. Probably different for 10 amp packs and 100 amp packs. Assuming charger control, you'd want an OR to limit current, and an AND to shut off current. Too bad we don't have charger control. And I'm not going to let Rube Goldberg rear his ugly head, although I am tempted to start thinking about a power disconnect . . . .

Let's see if the good folks at Zivan have an answer.
 
The DIN connector:

There is a 5 pin DIN connector near the battery connector. It is used for two things; a temperature probe and a remote LED display. First, let's get the pins defined, as the numbering is a bit bizarre (an extension of the three pin design). Looking at the business end of a FEMALE connector, with the 180 degree pins on top, they are numbered clockwise as follows:

1 - white
4 - blue
2 - green
5 - orange
3 - brown

Obviously reversed for the MALE connector. I have listed wire colors here. I think these may be standard, I do know that they correspond with the Zivan supplied remote display (well, pins 1, 2 and 3, anyway).

The remote display uses pin 2 (green) as the common negative for the LEDs. Pin 1 (white) is the green LED, and pin 3 (brown) is the red LED. I have built a cable for the remote display - it doesn't work at all. In fact, if you just use a 5 pin DIN extension cable, the remote doesn't work. My guess is that the very small signal wires won't carry enough current (although it works fine when hooked up to a lab supply). When I figure out this mystery, I will post it. I suspect if you use your own LEDs it will work fine. Note that you could use voltage sensing to determine what stage the charger was in (1, 2, 3, or off).

Pins 4 (blue) and 5 (orange) are used for temperature sensing. I checked two different probes for resistance - one measured 35.5 K and the other 38.x K. Both did respond to my heat gun, and went up about 3K when they got pretty hot (hotter than you would want a battery to be). I did connect a 100K pot to these leads - it made no difference at all that I could tell. I am testing against an old NiZn pack that charges so quickly it would be hard to measure anyway. The manual states the voltage differential is -5 mV/*C, which is pretty tiny - even 100C is only 1/2 a volt. With the leads disconnected, there is 5 volts across 'em - blue negative, orange positive.

There is an application note that describes varying the resistance between 1400 and 1700 ohms to vary the charger voltage quite a bit. This is clearly a custom version of the control software. I will continue to investigate if there is a way to make a significant difference to the charger output using the temp. probe and post if I find a way.
 
How much does this go for? Where can we purchase?

-steveo
 
Around $600-650 $USD

Found one here

http://www.electricvehiclesusa.com/product_p/bc-ng1-12-96-110.htm


May be beter economiclly to get Two Mastech 5005-2 that are good for 100volts at 5 amps or 50 volts at 10amps
http://www.amazon.com/Mastech-Dual-Linear-Power-Supply/dp/B000E14DJO
 
I know this is an older post but I thought I would bring it back up a little now that the NG1 seems to be dropping in price a little (at least on Ebay). I picked up a 72V NG1 for $250. Not bad!

I found this write-up very interesting and was wondering if there were any new discoveries regarding the charger. I did a quick test with a spare LED and found the the pinouts for red and green LED's were correct and helpful. In the manual it states that at 80% full charge a yellow LED will be illuminated,Is this yellow LED powered by the 5pin DIN connector? It could be helpful in troubleshooting to find the pinouts for the yellow LED as well as the red and green. I noticed that pin 5 (orange) produces a solid LED with no batteries connected, but I think this is just signaling for the temperature sensor. Any ideas?

Thanks for the great post.

Mick
http://www.zombiEVintage.com
 
I think you scored on that 72V charger. Hopefully the charge curve is something you can use, or perhaps alter to your needs.

The "yellow" LED is just the red and green together. While the manual and the stick-on label say this is "80%" charged, what it means is that the charger has gotten to the 3rd stage. What that means is entirely dependent on the charge profile (i.e., what the stages are as programmed).
 
Thanks Patrick. I hope I don't have to fuss with the charging curves too much. Im planning on running a 72v, 85ah Lead pack on a motorcycle I'm working on. From your experience with this charger, is it likely that it will need to be "tuned" for this application? Im still pretty new with large pack battery care.

Thanks!
Mick
http://www.zombiEVintage.com
 
Most Zivans were made for lead, so it will probably be just fine. What parameters are printed on the label? (e.g., IUIa, I1, U1, Umax, etc.)

If it is missing, a good thing is to check the maximum voltage. It will most likely be about 15 volts / 12V batt, or 90 volts for a 72 V nominal pack. If you can measure current at the same time, that would be ideal. You could then trim the voltage (if you need to) using the pots as pictured in the top of the post.
 
I have a NG-3 with an IUIa profile for 144V flooded lead acid. To convert this to a Lithium charger (IUa) it would seem that all you need to to is kill the charger after the 2nd stage.

i believe that Red led comes on on stage 1 (CC) , and blinks on stage 2 (CV). The yellow LED comes on on stage three and green after done.

A simple circuit could be created to detect the yellow LED (both Green and Red) and shut off the charger.

I can draft up something (couple logic ckts, a flip flop, and a solid state relay). but... has anyone done this?

Mark
 
You raise an interesting point, but I think it needs some elaboration. Forgive me if I belabor the obvious, but I'm writing for those who are newer to this than the experts (as this thread has been "stickified").

Lead batteries have a few variants; flooded, AGM and Gel. Very similar, but there are slightly different needs (from a charging standpoint) for each, mostly different voltage setpoints. There are also some newer strategies for the final stage, e.g., slow pulse charge. My personal experience is that these make a big difference in longevity - as an example, I've got 20 months and 2000 miles on a lead pack that is almost as good as new (Zivan NG1, or course).

But to paraphrase Nietzsche, lead is dead. The future is lithium, and these batteries are entirely different; they have to be managed on a per cell basis. So the intelligence of the (lead) charger moves to the BMS, whether simple analog or the most complex digital controller. Each cell has to be raised to its "charge point", and then held there until charge current tapers to its "minimum". The charge point, maximum current, and minimum current is different for each type of battery.

In the ideal case, the BMS and the charger are one unit, with voltage and current levels dependent on the size of the pack, individual cells, temperature, etc. So the "battery charger" as a standalone unit is old technology. The BMS unit built by Ggoodrum and Fechter is a great example (and great design). It starts with with a circuit to limit the "charge point" of each cell. But to limit the excess heat, it has to throttle down the input current. And the latest revision is to shut off the input (i.e., battery charger) when all the cells are full. But it illustrates the point, that the charger and BMS function as one unit.

Now to answer your question. You are right about the LED status indicators. Certainly a circuit (a yellow light detector) could shut off the charger. But would the lithium pack be charged? And much more elusively, would it be properly charged? Is the current right? Will the cells spend enough time at the CV? All of this depends on the integration with the BMS. And nobody knows the answer to these questions, as the projected lifetimes for these cells far exceeds our experience with them.

It's as clear as mud.
 
Electric Motorsports uses this charger to charge their lithium batteries in their motorcycle so it should be fine. Electric Motorsports said "Zivan Chargers are Unique because there can be programmed for just about any battery chemistry. This allows you to have the optimal charging Algorithm for your particular pack of batteries whether you are using AGM, Gel, Flooded SLA, Ni-Cad, NiMH, Lithium or any other rechargeable battery that may come on to the market in the future."
 
Umm, it's not quite that simple. First you need to get the Zivan set up for the particular pack you using, and that isn't necessarily easy. I know, 'cause I'm the one that had that NG1 reprogrammed when working with Todd at electric motorsport. I'm trying to get a standardized lithium profile built by the folks at Zivan, ideally with a couple of setpoints.

Second is that they are using that charger on a ThunderSky pack without a BMS. This is not recommended, and frankly they're surprised the cells are staying (relatively) balanced. They are using a different OEM charger on their new electric bikes, and it has an internal BMS.
 
I have some questions which will clear up some of my confusion. Is the NG1 a single charger which can be set for any voltage between 24-90VDC or a line of chargers which cover that range of voltages? Also are people changing these settings themselves or just getting them preprogramed by Zivan?

Right now my impression is you can take an Zivan NG1 set it to 87VDC output and use it with Gary's Open BMS yourself but I'm probably missing something here. So I'm thinking about getting a used 29.4v at 18A charger and changing it to 87v at 8.5 amp. Which I can do correct?
 
Back
Top