ZXD 3 stage question

[jefferso]

I have a ZXD charger. I have it set to 28.2v and 11a and set on 3 stage charge. It seems like it's skipping the CC and CV stage and going right to float. I'm using an XLR plug on MK gel batteries on my old Permobil C500 and also tried it on my F5 with the "new" Permobil batteries.

The last time I charged, the battery was fairly full, but it never got to 28.2v, just did 6 amps or so to start at 26.7v then the amps gradually went down. Anyone have an idea why that might happen?

[martin007]

3 things can happen.

1- The ZXD is malfunctioning.
2- The batteries are fully charged.
3- The batteries are deteriorated.

[Burgerman]

More likely you have not configured it correctly.

What is the CURRENT you have set for it to transition from 28.1 to 27.0V? This needs to be around 0.3A on a cyclic overnight charge. Set this any higher like 1A or above and it will go straight to float if the battery is nearly full. And then take another 16 hours to complete the charge bare minimum.

[jefferso]

More likely you have not configured it correctly.

What is the CURRENT you have set for it to transition from 28.1 to 27.0V? This needs to be around 0.3A on a cyclic overnight charge. Set this any higher like 1A or above and it will go straight to float if the battery is nearly full. And then take another 16 hours to complete the charge bare minimum.

Maybe that's it. I think I had it set to something like that, but I changed the main amps setting from 10 to 11 and I think it automatically resets that value to 30% of the current limit value. I'll check that.

[Burgerman]

Once set corectly it will do this.

Note this is a full battery! Well it was half an hour ago.
Set to transition at .2A here. And to float at 27.2.

https://www.wheelchairdriver.com/gopro/3stage.mp4

[jefferso]

Once set corectly it will do this.

Note this is a full battery! Well it was half an hour ago.
Set to transition at .2A here. And to float at 27.2.

https://www.wheelchairdriver.com/gopro/3stage.mp4

Thanks. That was it. Changing the amps changed the jump current to the 30% value. Setting it to 0.3 amps makes it work as expected.

I might have to adjust this because my ventilator battery pack constantly draws around .2 amps even when the vent is fully charged and not running. So maybe I should set it to 0.4 amps. Or maybe have to unplug that with each charge, but that's one more thing to forget.

Charging 28.2v and 2a

[Burgerman]

Set it so that it turns to float after it has sat at the high CV voltage of 14.1V (28.2) at room 20C temperature for between 7 to 8 hours. (after a normal days use/charge). Dont include the CC stage before it reaches its CV voltage.

That will give the fastest charge, in cyclic use while making the battery live the longest.

[shirley_hkg]

the CURRENT you have set for it to transition from 28.1 to 27.0V? This needs to be around 0.3A on a cyclic overnight charge. Set this any higher like 1A or above and it will go straight to float if the battery is nearly full.

Preset value of resIdual current to trigger to float
" is set at 300/1000. "
If your charge current is 11a, it shifts to float at 3.3a.

If you want BM's 0.3A, that's 0.3/11 = 0.02727

To achieve this, you have to change the 74th EEPOM value to 27 (0.2727 * 1000).

#73 is the float voltage.

[Burgerman]

You could do that. But you can also just set as normal and it and it remembers it for next time anyway.

Setting its default to 1/3rd the charge current is rediculous. That will work to top up a car, if left connected for 24 to 36 hours. It cant do a fast recharge in cyclic deep discharge use over the short nightime we have avalable. Will end up with a sulfated battery pretty fast.

Its SAFE default for general usage should be say, .7A to 1A as this will still terminate even with a battery still connected in a car, with small drain. And should have a 8 hour CV timer before going to float too. Whichever occurs first.

Setting it as a percentage of charge current is stupid. If charging a battery at 50A as I do all the time it will go straight to fload while charging at 17A ... At which point the battery is nowhere even close to charged. Maybe half?
Altrnatively if charging at 3A it will go to float after any many hours, at 1A which is still too soon but will be 99% charged. Its a stupid way to do it.

Better to set a 5 hour CV then float for top offs.
And a 8 hour CV the float for cyclic charging.

[shirley_hkg]

Yes, it can be set directly , but it won't stay there. It will be erased and return to the EEPROM % , IF the charge current were touched.

The same applies to float volt too.

[Burgerman]

It would be better if it did "stick" to whatever current you chose independently. But as long as you change nothing after setting it to whatever you want, it does "stick".

[jefferso]

I am using my old Permobil C500 now so I could try to fully charge the newer F5. It stayed at 0.44 amps at 28.2 for about 12 hours. When the main power switch was turned off, it went immediately to 0.1 amps then to float at 27v. But I'm not sure if it's actually still charging when the main power switch is off. I don't know if that disconnects the joystick power input from the battery.

So it looks like even with the joystick power switch off, the F5 uses 300 - 400 amps. It has internal bluetooth, GPS, and some other wireless so it's constantly using more power than the C500 uses, unless the main power switch is turned off.

I used the "new" Permobil charger for the new chair for a while, and I think it's probably not done well for the batteries.

[Burgerman]

I am using my old Permobil C500 now so I could try to fully charge the newer F5. It stayed at 0.44 amps at 28.2 for about 12 hours. When the main power switch was turned off, it went immediately to 0.1 amps then to float at 27v. But I'm not sure if it's actually still charging when the main power switch is off. I don't know if that disconnects the joystick power input from the battery.

What main power switch is this?

Unless you are talking about the chairs ressetable breaker?
That switch disconnects the battery from the whole chair. At which point the charger if in use will be powering the chair... And whatever current it reads is then what the chair takes to run its electronics.

[jefferso]

What main power switch is this?

Unless you are talking about the chairs ressetable breaker?
That switch disconnects the battery from the whole chair. At which point the charger if in use will be powering the chair... And whatever current it reads is then what the chair takes to run its electronics.

Yes, the power circuit breaker. After switching that off, I think it fully disconnects the joystick power port as well because it went to 0.01 (not 0.1 as I wrote earlier). So it's like the charger is connected to nothing.

My C500 has an SB50 wired directly to the battery, so even when the main power is off, it can still charge.


Here's the F5 breaker:

[Burgerman]

No it just disconnects the battery. The charger is then doing the job of the battery. DONT try moving the chair, it could hurt the thin wires in the joystick. If you turn the chair on then on a normal R-Net chair the joystick will light. And then draw a few mA like maybe 100 or 150 depending on actual modules fitted. Then when its turned off it will be taking maybe 10mA.

As for the permobil seating stuff, no idea what that will take or do.

The charge port is directly connected to the 24v wires in the bus cables that powers all modules and these are directly connected to the battery loom inside the power module. Where the breaker is.

The SB 50 chair will charge as normal, and the chair will draw nothing when off, and a couple of hundred mA when on. The charger cant read figures this low properly.

[jefferso]

No it just disconnects the battery. The charger is then doing the job of the battery. DONT try moving the chair, it could hurt the thin wires in the joystick. If you turn the chair on then on a normal R-Net chair the joystick will light. And then draw a few mA like maybe 100 or 150 depending on actual modules fitted. Then when its turned off it will be taking maybe 10mA.

Confirmed that the joystick powers on with the breaker off when the charger is plugged in. Showing 0.59a. Only goes down to 0.4 something when power turned off at the joystick. I think the resting state of this chair is quite power hungry.

So if I want it to go to float, I have to change the jump current from 300 to 450 or so. Will it get a proper charge while the chair is drawing that much sitting there while the charger supplies that power + some for the battery?

Does the ZXD have a timer associated the CV stage? Or you just try to target the right amps for it to stay on that long?

[Burgerman]

Thats quite greedy but its not an accurate way to test as the charger is set up to measure 50A. So is like using a car speedometr to measure 0.5mph.

The restibg state of R-Net here n my room is 12mA when off. Thats 0.012A
The current when on, depends on what modules you have, and mine reads 180mA. Or 0.18A.
You should not have it turned on while charging anyway.

The fact that you are seeing such high current when OFF is problematic. Its either a bad measurement, or a fault, or possibly that permobil stuff like their own seating module sucks a lot of power. That will send a battery dead fairly fast. 0.4 means nearly half an amp hour! Thats definitely not correct.
That will take out 9.6Ah per day when OFF! Thats cannot be correct.

Either way it doesent really matter. Because you should only turn off that breaker after the charging is finished if the chair is in storage. You do that because even when turned off, the control system pulls a small current from the battery. Switching the breaker off, stops that additional slow discharge. It means you only need top up the battery every 2 months. Instead of every 1 to 2 weeks (r-net).

Aim for an 8hour CV and adjust the change to float voltage current over the next few charges to achieve that. Theres no timer. There should be no need. A quick top up will take less hours at CV. A slower charge rate will take less hours at CV. A deep discharge might take 12h to complete. Thats how it should be.

[shirley_hkg]

My C500 has an SB50 wired directly to the battery, so even when the main power is off, it can still charge.

I recently squeezed 160Ah into a F5VS. I recalled the breaker got between the 2 lead bricks.
It will be in open circuit , if you flipped that breaker off, and ZXD is charging nothing.

[jefferso]

At some point the ZXD dropped to 27v and about 0.02a. I wonder if it's the Permobil internal bluetooth consuming so much power, or the wireless connection keeps trying to connect and the wireless signal isn't great in the flat so it has to keep trying. I think there's an internal switch where all the extras can be turned off. I might try that.

Something is definitely using power. Once I left the chair for about two days and it was showing 73% battery in the fancy MyPermobil app after doing nothing for two days. The app shows an estimated range along with a battery %. It used to estimate 19km range when I first activated the app early this year, now it's showing 14km.

From marketing brochure:
"How does MyPermobil estimate how far I can travel?
Permobil uses an exclusive advanced calculation method to estimate battery charge and travel distances."

It's both "exclusive" and "advanced" so it must be good!

[jefferso]

I recently squeezed 160Ah into a F5VS. I recalled the breaker got between the 2 lead bricks.
It will be in open circuit , if you flipped that breaker off, and ZXD is charging nothing.

Do you have any photos of the pack? I want to make a lithium pack for my F5VS and am trying to figure out the best configuration.

[Burgerman]

Because of the way lead batteries work, age, and peukert, temperature, and various HUGE differences in range due to load, its quite simply impossible to get closer than maybe 50% when estimating battery range or remaining capacity. Sometimes not even that good.

No matter how "advanced" permobil thinks their guesswork is. About the only thig a lead battery "guage" can really tell you is that you are suddenly out of power. Or that directly after charge you get a full set of lights. But you ALSO get a full set of lights if you give it 10 mins charge from about half full. So that too is all but useless. Battery gauges are really here to give a feelgood factor. And because the programmer who doesent understand batteries thinks you can compare voltage and charge state. And that doesent work...

So I wouldnt take a blind bit of notice of this. https://www.wheelchairdriver.com/board/ ... &mode=view

[slomobile]

How do we know that chairs are only using voltage to estimate state of charge? Disassembled firmware? Circuit analysis of a PM?

Coulomb counting battery "fuel gauges" which mathematically integrate energy in and out have been around for quite a while now and do a much better job estimating capacity, especially if they know the lead resistance(an Rnet parameter), can sense temperature, and know something about the power factor (related to motor impedance Rnet parameter).

Do we just assume they are using voltage to estimate SOC because they do such a poor job at it? Or has someone opened up a PM and checked for a coulomb counting chip?

Edit: if using the stock XLR charge port which must charge through the PM, a coulomb counter in there could theoretically measure charge going in and coming out, but if Andersons are connected directly to battery, bypassing the PM, it could only measure charge going out. So in that case it would have to use voltage I guess.

[Burgerman]

How do we know that chairs are only using voltage to estimate state of charge? Disassembled firmware? Circuit analysis of a PM?

They dont.
They use some mathematical gymnastics, a lot of averaging, ignore the loaded volts as you drive, and take into account when it was last above the 13.2V level so they know its been charged. Explained by a PG drives engineer. But that still results in complete nonsense.

[ex-Gooserider]

To me the best evidence that they aren't looking at anything but voltage is that if you run around for a bit until some of the lights go out, and then stop and sit for a while, they turn back on again... I've had times when I've been down 3 LED's while running to the can in order to return coffee, and was fully charged by the time I was done....

I wish my van would do the same thing, it would be nice if I had a gas tank that refilled itself while sitting at traffic lights

ex-Gooserider

[Burgerman]

Or when you connect a system on the bench. I do that to program power modules I keep as spares just in case I need to swap one. You can make the battery lights go up and down by setting the power supply volts to different levels. Theres a delay, sometimes a few minutes. But you can do this pretty easily. Also in programming you can tell it the voltage it cuts off power, or calibrate for different battery types. Its not only voltage, but its the main signal used. Theres a lot of mathematical gymnastics and guesswork to estimate things too.

Likewise I can take a fully charged fresh chair and drive it hard up a steep slope for 2 mins (a 5 floor multi story car park in this case) and it will lose at least 2 lights and sometimes 3. As goose says, sit in the resturant on that top floor and they all com back.

The opposite is true. I can take a chair with 2 lights left, on its last legs, and connect to my van as I go home. After about 1 mile its reading full... Its NOT full but the voltage goes up fast at 50 to 70A charge from the alternator. As soon as you leave the van, withing 1 hour its back down to a couple of lights again. The guage hasnt a clue about state of charge, only voltage. And voltage is NOT state of charge on a lead battery. It can be, if discharge at constant rate over 24 hours. Otherwise its not close.

[LROBBINS]

Using LiFePO4 batteries exposed two problems in my CANbus wheelchair control system. The nearly flat voltage curve makes the voltmeter even more useless that it is with Pb, and also eliminates the scheme I had for blocking movement while charging, based on voltage and without a charge sense contact. I am now working on using another 200A current sensor as a Coulomb counter in the main battery line. The Coulomb counting (converted to Amp Hours used) should give a direct indication (to show on the Display's AHr bar) while I hope to use reverse current flow to zero out throttle and steering while charging. The latter may be tricky as I have to find a reverse-current condition that won't be misread as charging when a motor acts as a generator (while sliding down a hill or during passive deceleration). Turning the controller on for a couple minutes on near the end of charging will reset AmpHoursUsed to zero (which will be false if the charging is interrupted before this, so this should be done after the PL8 has started balancing, i.e. when current has fallen to say 1 A or less so even if interrupted the "0" AmpHoursUsed will be at least close to correct).

I'll let you know how this works out.

[shirley_hkg]

Do you have any photos of the pack? I want to make a lithium pack for my F5VS and am trying to figure out the best configuration.

He uses chin control and always turn on a dime . I have to squeeze the most it can take.

I used 2 of these semi assembled packs and stack them together.

[foghornleghorn]

I used 2 of these semi assembled packs and stack them together.

That looks interesting to me and might fit in my Invacare Storm 4.

Do you have any more photos of the pack with wiring? Curious to see how it would be with those plates on instead of bolt terminals.

[Burgerman]

8x 230Ah cells shld fit in the storm 4?
Unless its a height issue?

[shirley_hkg]

8x 230Ah cells shld fit in the storm 4 ?

No. Height and length are not enough. That's why I use semi assembled cells , instead of welded studs.

Bus bars are aluminum and are laser welded. Top side is copper plated, so balance wire can be soldered. (100W large iron)