Clone imax b6 (50W) more discharge current mod

[Juha-Matti Sidorow]

I want to increase discharge current. I found this mod from other forum:

http://www.rcgroups.com/forums/showthread.php?t=2073448

This was very simple to do, but charger will not "see" actual current. Is there any way to make charger see actual current? Where is that measurement point at pcb?

[JagiChan Sir]

Look at the schematic, the relay is used to connect the battery and the bulb directly. Essentially the relay is acting like an "electronic switch". The charger just supplies the drive for the relay (mentioned as "bias" in the mod). Essentially the bulb is in parallel with the battery. Hope this helps.

[Juha-Matti Sidorow]

Yes, i understand how that circuit works, but i wondering is it possible make an circuit that shows current on charger screen right? Now it shows only charger discharge current, because extra load is parallel with battery.

Is it possible to add extra discharge fets or swap current to higher power rate one? then also firmware need changes (cheali-charger).

[JagiChan Sir]

On Tuesday, February 9, 2016 at 6:43:21 PM UTC+5:30, Juha-Matti Sidorow wrote:

Yes, i understand how that circuit works, but i wondering is it possible make an circuit that shows current on charger screen right? Now it shows only charger discharge current, because extra load is parallel with battery.

Is it possible to add extra discharge fets or swap current to higher power rate one? then also firmware need changes (cheali-charger).

Sir anything is possible the question is "how far are you willing to travel?". 

Look at the schematics that are posted in this group, you'll quickly understand that to make your wish happen, there needs a lot more to be done than just adding a relay.

What you need to do is this 

1. The electronic load monitoring the FET IRFZ44 consisting of LM2904 and its associated components needs to be modified to accommodate the extra current measurement you are looking for.
2. You also need to make sure that the max current measurement results in Vadc within the acceptable voltage range of the ADC (that means the max output from the electronic load connecting to pin 35 of the uC should not exceed 3.3V).
3. Of course the "magical software" also needs to be adjusted to read and display the actual current.

[Paweł Si]

It should also be possibile to connect a load in series with the battery (on the red wire)

you could use the  FET IRFZ44 just as a on-off switch, although:

1. the load should be the right "size"

2. you would need to use the balance port (so only LiXX batteries)

3. discharge current needs to be smaller the the maximum discharge current (as JagiChan Sir explained)

4. a small firmware modification is needed

[Juha-Matti Sidorow]

If i just replace 0,5ohm resistor with higher power rate one. IRFZ44 can handle 50W, so example 4,3v li-ion cell drains max. 8,6A @ 0,5ohm resistor = 37W.

Only need firmware to allow max current higher.

[diggit]

Charger is designed to dissipate power on discharging transistor (mostly IRFZ44) which is driven in ohmic mode. If you want to raise discharge power limits, you have to get better cooling for this transistor. 

I modified mine with additional cooler from graphic card and moved/replaced all power components (5) and attached to it (isolated). Then replaced sensing resistor with smaller one. Finally flashed FW with modified limits.

[diggit]

Tested with GC cooler fan running on 5V and 20W seems to be OK. With 12V for GC coller fan, it would be more, but it sounds like jet. 

[JagiChan Sir]

@diggit : Nice work :) But you could do better, by replacing the 0.5ohm with let's say 2ohm/10W. This way, the IRF44Z would be out of the ohmic range and not get hot but the resistor would dissipate the power and you can keep the resistor outside.

Here is the modification that I would do to avoid that fan and the noise.

1. Remove the 0.5 ohm resistor from the board.
2. Drill two holes into the IMAX case, to house 2mm banana jacks.
3. Connect the 2mm banana jacks to the two points where the 0.5ohm resistor was originally placed.
4. Close the case carefully.
5. Modify the firmware to accommodate the discharge current you want to see and flash the firmware.

Now you can connect any desired discharge load to the newly created banana leads. For example you could connect an automotive bulb of 12V 50W and discharge the battery (and you don't need a fan to cool the load)!

Remember : the max current that can be measured by the charger is still limited by the "circuit of the electronic load". Of course you can change that by making appropriate changes to the electronic load.

Regards,

JagiChan

[diggit]

With your mod, you won't be able to regulate discharge current. If you look into schematics, you can see discharge circuit. It is pretty neatly done. Input is voltage from MCU(PWM), which is filtered. Main current regulation is done by sensing voltage on sense resistor. Circuit drives transistor to set required voltage drop on this resistor. If you increase resistor, same voltage drop will be at lower current U=R*I. It is constant current load.

Yeah, there is another one resistor for current sensing for MCU but that one is used in charrging mode only.

[Juha-Matti Sidorow]

I ordered 100W 0,5ohm resistor from ebay. Is there hex for higher discharge current?

[diggit]

You have to build it by yourself. Look here https://github.com/stawel/cheali-charger/blob/master/docs/building.md

Locate file HardwareConfig.h for your charger, modify limits, and build it. One thing is software limit, the other is hardware limitation.

SW power limit protects transistor. If you won't improve cooling of load transistor (in 50W chargers, IRFZ44 near battery terminals), it will burn.

YOU HAVE BEEN WARNED!

[JagiChan Sir]

@diggit : I agree with your observation, but look at what I wrote at the end "the circuit needs to be modified to accommodate the extra current". I think I did not explain clearly, changing the 0.5 ohm to 2 ohm will bring the FET out of the "triode mode of operation", which is essentially "the ohmic range". This will keep the FET cool and the actual power being dissipated will be in the resistor, which can be left floating in the ambient air for cooling. This way, you don't need that active cooling that you have implemented.

The guys at SkyRC maybe reading this forum, and you never know they may implement this idea and sell it as a different model :)

Regards,

JagiChan

[Diogo Carvalho]

The guys at SkyRC did not implement that ideia because that will not work.

If you use the mosfet as a switch it will be only ON of OFF, so you can't define the discharge current.

With a 4.2V battery you will have only 2.1A of discharge current or zero.

You need mosfet in linear mode, so you can change his resistance to adjust the current.

It will be very hot? Do you need a good heatsink? Of course yes.

Discharging is wasting energy, and that energy needs to be dissipated. In this case in the form of heat.

[diggit]

Diogo Carvalho, I agree.

Here is some further explanation...

" changing the 0.5 ohm to 2 ohm will bring the FET out of the "triode mode of operation", which is essentially "the ohmic range" " Are we talking about R50 in this schematic which is on the left, second sub-circuit from top below FET? R50 is current feedback resistor. Whole that sub-circuit takes care of driving FET to open exactly that much, to make voltage drop on R50 multiplied by 3 (DA1:1) equal to voltage from MCU divided by 2.

default state, expecting that MCU can give 5V at max:

5/2=I*R50*3 => I=1.66A this is maximal current, that allows this circuit to flow through R50 and FET

if you change R50 to 2 ohms, you will decrease max current I=410mA.

This is, how you set HW limitation.

I would be possible to share load between some resistive load and FET if you put in between battery and FET. This way, part of heat will be dissipated on this load and theoretically higher discharge curent might be possible. This woudl quire adjusting load according to voltage and current of battery. => you need variable load and that is exacly what this FET does.

[JagiChan Sir]

Ok, now I see, you are using the original unmodified circuit, and here I have modified tons of stuff and am looking at my modified schematics... sorry, my bad...

[Igon]

 

 equal to voltage from MCU divided by 2.

by 3  - because of forgotten R and C

https://groups.google.com/d/msg/cheali-charger/WkkV64jA56Q/L2_pIwb5CAAJ
what is C19, R141 and so - https://drive.google.com/file/d/0B1RXXTatsA1cbkM2dXFxTldjTUU/edit

[diggit]

Thx for correction. I'll refer to proper schema next time. BTW gain of R50 voltage is different too. It is not important for this thread, at least now...

[Juha-Matti Sidorow]

I don't understand that hex building stuff so well. All i need is latest version with 10A max. discharge current limitation. After that i can use battery specified max. values. example: 3.7 li-po 1s = 10A, 2s = 5A, 3s = 2.5A, 4s = 1.25A and so on..

[diggit]

Tell me what charger do you use. (or link to current hex). There are 2 limits, power(W) and current(A) to set. Tell me your desired values and I'll build it for you.

[Juha-Matti Sidorow]

This is imax b6 50w atmega32 clone, but if i remember right, last time i tried flash it with clone hex and it will not work (no boot). original version hex works or bin. Why there is hex and bin? If i understand right. bin is "crypted" and hex is not crypted?

Let's say 10A and 43W limits.

[diggit]

Oh, so at first, test if cheali firmware runs on your HW. There is no point to build some modified image until you make default one running.

HEX stores data in hexadecimal format with memory addresses where data belong including checksums.

BIN is just plain blob of binary data.

[Juha-Matti Sidorow]

So which one is better to flash? bin or hex or does it matter?

[diggit]

They are equivalent, but I prefer hex. Programmer can detect if file was corrupted. (no so reliable as hash)

1 minute google search.

[Igon]

Let's say 10A and 43W limits.

It is a bad idea to use transistors near its maximum power and beyond "Maximum Safe Operating Area" - reliability is so-so.
Especially for switching FET putted in linear mode
Also it is not easy to cool TO-220 through heat-conductive rubber at such power levels - 140 and more degrees can be riched easily

[diggit]

This is meant for modified HW as I was notified few posts earlier. All warnings were said. :)

[Juha-Matti Sidorow]

I can replace IRFZ44 with higher power rate transistor. Any suggestions?

[Juha-Matti Sidorow]

I tried flash hex and bin both clone and original and now there is only ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ on screen.

[Juha-Matti Sidorow]

Ok, i have to reset fuses, now both clone and original hexs works, so which one i use?

[Juha-Matti Sidorow]

this is clone, but programming port is same layout than original. how can i check which firmware working better? original or clone?

[Juha-Matti Sidorow]

What is the difference between:

cheali-charger-imaxB6-clone_1.99-20160216_atmega32

and

cheali-charger-imaxB6-original_1.99-20160216_atmega32

I have clone, but also original hex works?

[Paweł Si]

https://github.com/stawel/cheali-charger/blob/master/docs/flashing.md#differences-between-versions

they differ only in the default calibration

[Juha-Matti Sidorow]

So i can use both if i just do calibration before use?

Is it better to do calibration, expert calibration or both, and which one first?

[diggit]

Pick any of them. Calibration is necessary anyway.

Did you even read main README?

Calibartion section is followed by link to expert calibration which is, as noted, optional.

In expert calibration is explained what it is good for.

When you read through, last sentence is:  If you do any changes in "options"->"calibrate"->"voltage" you have to repeat this process.

So common calibration of voltage must be done earlier than expert calib.

[Povl H Pedersen]

Can I run a H4 bulb or other load on the XT-60 of the battery for discharge, and run the discharge program on the cheali doing low amps discharge, and thus use the cheali to buzz when target voltage is reached ?

Any plans for a charger mode in 2.0 that will observe voltage over balance connection only, and sound alarm at a given voltage ? Can be used for discharge under load graphs, and to graph voltage drops under load if playing with new motor / prop combo

[Juha-Matti Sidorow]

Yes, i have read common readme.

If same firmware works with many chargers, why there is so many variations?

Example what is this: cheali-charger-imaxB6-clone-50W-4.5A

What does that 4.5A means? There is no readme or any info at hex folder.

[diggit]

You can compare charger specific sources in src directory. There are not many variations. Mentioned version has maximum charging current limit set to 4.5A instead commonly used 5A for 50W chargers, nothing else. (I don't know the reason...)

This is probably yours cheali-charger-imforaxB6-clone. (filename changes every new build)

[Juha-Matti Sidorow]

Thanks diggit for helping me out! I need only these changes to test my charger limits before and after modifications:

#define MAX_DISCHARGE_P         ANALOG_WATT(50.000)

#define MAX_DISCHARGE_I         ANALOG_AMP(5.000)

Can you build hex for me or help how i can do it?

[Juha-Matti Sidorow]

Now i have modified firmware (thanks for diggit), but after i calibrated. it won't let me charge or discharge. Says: "calib. error I discharge 14"

[diggit]

Cheali-charger is pretty smart! It checks, if it is even possible to measure such value. It concludes, that with such current limit and your calibration, maximal value is out of measurable range. If you want to measure so high current, you have to decrease current sensing resistor (already discussed in this thread).

BTW if you have modification from link in 1st post. You don't even need modified FW limits, because additional discharge current is not measured by charger and thus not limited...

[jusubbi]

I want cheali-charger measures current. I want test many single cells per day so i need more discharge current to do that, With 1A discharge current i can test only 1 cell per day (capacity check).

Calibration voltages are pretty close by default, but currents needs lots of adjustment.

Maybe this clone has default calibration values closer original hex than clone?

How can i make a backup from eeprom? I tried this one:

avrdude.exe -patmega32 -cusbasp -Uflash:r:flash.hex:r -Ulfuse:r:lfuse.hex:r -Uhfuse:r:hfuse.hex:r -Ueeprom:r:eeprom.hex:r

But it messed up whole charger and now there is only ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ ▓ on screen again.

If 0.5ohm resistor is too high, i have also 0.3ohm resistors, but they are only 7W

[diggit]

Damn, I am bit confused.

Do you have somehow modified hardware? (like mine)

If not, increasing limits may (and in result will) damage your charger.

If yes, give us some description, schematics, links...

Dumping eeprom...

this is all times same

avrdude.exe -p atmega32 -c usbasp

and append for eeprom dump

 -U eeprom:r:eeprom.hex:i

 this should work, eeprom.hex will contain data from eeprom

[jusubbi]

I have this one:

https://groups.google.com/forum/#!searchin/cheali-charger/ver.$201.3/cheali-charger/7FB3Xt2-tbE/KSL6EHHZlh0J

ver. 1.3 on pcb. I haven't done any modifications yet. Discharge transistor is NDP6060.

I also noticed that external temp sensor will not work with cheali-charger. With imax b6 firmware it works.

[Povl H Pedersen]

For faster discharge I do plan an experiment. A load (car bulb) across + and - controlled by relay.
The relay will be controlled by Arduino on serial port, which will monitor if any discharge/storage program is taking place and > 50mA power drawn by cheali. If so, it will turn on bulb for faster discharge.

How does this sound ? Arduino is $1.50 and 5V relay with 10A load rating is $2.50 for 5 pcs. Of course that would make any measurements of discharged power worthless. But maybe it could be calculated in the Arduino.
Serial data would be passed through. But bulb on/off could be inserted. Or the estimated mA in bulb.

[jusubbi]

I changed discharge resistor to 0,3ohm. Now it gives "calib. error I discharge 4" with 0,5ohm it was 14.

Is this 4 or 14 some code or is it calibration value? example 4000 or 14000?

[diggit]

If you want to increase discharging capabilities of your charger, here are some milestones:

1) put discharging transistor on heatsink and cool it (I use heatsink from graphic card)

2) change sensing resistor to lower values, this will allow higher discharging currents

3) change limits in firmware to use potential of your modification

One by one. If you haven't improved cooling of transistor yet, do not proceed to next setp.

You are still out of range. If original limit was 1A, and now you want 10A, you have to divide resistor 10 times... 0R5 -> 0R05 or set some sane limits...

Necessary wattage of this resistor can be calculated P=R*I^2 -> 0.05*100=5W

[jusubbi]

Now i finally get some progress. I put two 0R1 5W resistors parallel, so now i have 0R05 sensing resistor. After that i run calibration again and voila! No errors and discharge works. Next i have to put some seriously heatsink to  FET. Here is picture from quick test.

https://goo.gl/photos/wPnmefjJkPKjYMwk6

[diggit]

Good, charger behaves as expected. Find some massive heatsink with fan. (CPU, GPU,...  cooler) My modification can handle 20W with fan at 5V and blows hot air.

You should know, that increasing load current range could have negative effect on regulation precision. Test it and you'll see if it is still able to discharge at low currents like 50mA.

[jusubbi]

50mA and 300mA discharge calibration are ok. Checked with fluke multimeter.

Do you have fan constantly on? It would be nice if fan can be on only when dischargin.

[diggit]

Fan on my charger turns on when charging or discharging is active. (I moved charging circuit power parts on heatsink too.)

If you want to turn on fan on discharging only, there is pin 30(PA7) on ATmega which is LOW when discharging and HIGH if not. You can invert it and put transistor on output to switch your fan. Sth like this, but replace right resistor with your fan. (+ wire of fan could be connected to input voltage or 5v)

[jusubbi]

Is it bad idea to take fan control from FET gate? Like this:

[diggit]

It depends on FET's gate voltage. I'd prefer not to load analog part. With circuit i suggested, fan transistor will fully open every discharging. 

[jusubbi]

So is this right (components also):

[diggit]

Sorry, this circuit won't work, my fault.

Use this (search page for keyword inverter) , it suits exactly your needs.

[jusubbi]

I build that circuit and it works very nice.

Now i have to figure out how i transfer all that heat better to heat sink. I have pretty big heat sink and it doesn't even warm, so there is something wrong with heat conduct.

[Diogo Carvalho]

you can use two mosfets in parallel, so the current is half in each transistor and the heat is best transfered to the heatsink because you have more contact area.

[diggit]

If you mount that transistor directly to heatsink with thermal grease, you should get good thermal conduction.

[jusubbi]

I replaced NDP6060 FET with p0903bsg (found it from scrap mb). It has rds_on only 9.5mΩ versus older 22mΩ. I also put FET and resistors to inside of heat sink where air flows and some thermal paste. Now FET and resistors is about 60'C when dischargin. Fan is also on only when dischargin, thanks for schematic diggit!

But there is still some too hot parts. That second resistor that i haven't replaced (battery minus side FET over 120'C) and one diode (80¨C). That diode is maybe ok, but resistor needs some work.

Here is few pictures:

https://goo.gl/photos/5d1acJyxsoiMcRiQ8

https://goo.gl/photos/fGXHxLoyqWdtzqmv6

https://goo.gl/photos/Zs3uupYbRhZcyUAE7

https://goo.gl/photos/yxaTLCdDANfv2F487

https://goo.gl/photos/oxaQoUL9Py69BEj1A

[diggit]

There were no need for FET replacement. In this circuit, is is driven in mode, where FET is not fully opened (even several ohms). This is how current regulation works.

The other resistor is for current sensing in charge mode, but remains in circuit even for discharging. You should decrease it's value too to lower power loss there. (then you have to re-calibrate currents)

Could you make photo of your heatsink and FET mount to it? It looks like wrong placement of FET. 60°C is not that bad...

Diode is expected to warm up during charging only, it is part of buck/boost power supply for battery.

[jusubbi]

Ok, i broke NDP6060 FET's feets, that's why i have to replaced it. Which values i have to focused when i searching replacement transistor?

That charging current sensing resistor is 0.05ohm so it's pretty small already. Maybe i do same thing to it and put two 0.10ohm resistors to parallel and those to inside of heat sink where air flows.

That diode is hot only when charging.

Here is photos of whole device and close up from FET's mounting point.

https://goo.gl/photos/HtEt9nWAagiUtnVc6

https://goo.gl/photos/jKmsoJA6rzXNZ4X37

[diggit]

Look into FET's datasheet. Find chart with Vgs (V) on X axis and Ids (A) on Y (current depending on voltage). 

In Datasheet for p0903bsg page 3, chart called "transfer characteristics".

Maximal voltage that FET driving circuit can deliver is 5V. At this voltage, Ids must be greater than your maximum current with some margin. p0903bsg is ok, IRFZ44N should be too.

Put your FET into vertical slot in ceter center, somewhere in the middle of whole heatsink. (not on the end of heatsink like now) You can use wires between FET and PCB, only Drain and Source wires must handle your max. current, gate wire could be thin.

Those resistors are made to be cooled by air flow. Spread them apart to get bigger surface to cool.

Resistor heating to 120°C is 2W, you have R*I^2=0.05*100=5W power loss at 10A. Change it to higher power one and/or lower its value even more.

[jusubbi]

Now i have IRF640N for discharge. I placed it center of heatsink and it's getting only 40°C ~15W power without fan.

Those "chargin diodes" goes over 100°C when chargin ~15W power. Maybe there is need to do something for them?

There is also problem with voltage sensing:

When there is no load or charging voltage is measured right, but when dischargin voltage is ~200mv below of right value. Current measurement shows right value.

Is this some bug or what? I have done calibration and expert calibration and i use balance connector to sensing voltage (i use single cell only).

[jusubbi]

Now i replaced SR330 diodes with BYW80. Temperature is now below 40°C when charging.

Voltage measure error still remains. 400mV when discharging. When charging or iddle there is no error. Here is some picture:

https://goo.gl/photos/DtAnuveS69k6ELMo6

https://goo.gl/photos/JeCbGdToAXKzZX5H8

https://goo.gl/photos/q5nJ3H8PRVFsFzyS8

https://goo.gl/photos/6eCJuTwoWWcarGcz8

https://goo.gl/photos/xEwsSXAufKQthGLf8

[diggit]

On Friday, March 4, 2016 at 7:11:31 PM UTC+1, jusubbi wrote:

Now i replaced SR330 diodes with BYW80. Temperature is now below 40°C when charging.

Good job, some of the components really worth changing.

 

Voltage measure error still remains. 400mV when discharging. When charging or iddle there is no error. Here is some picture:

There is probably some voltage drop on wires/PCB traces/connectors. Try to put voltage measurement on balancer cell 3 or higher. (more advanced circuit eliminating problem with common wire)

[jusubbi]

vb1 is only one that shows wrong voltage when discharging. vb2-6 shows right.

i have some short circuit once and i make new traces to balance connector. i guess that was vb1 then.

it shows correct voltages when charging. i'm little bit confused.

if i use just one cell, there is no problem, i can use vb2-6 to measure voltage. but i also have 2s lipo batteries for my r/c plane and they needs vb1-2 to work right.

[diggit]

On Saturday, March 5, 2016 at 11:05:25 PM UTC+1, jusubbi wrote:

if i use just one cell, there is no problem, i can use vb2-6 to measure voltage. but i also have 2s lipo batteries for my r/c plane and they needs vb1-2 to work right.

You can use Vb2 and Vb3 for charging 2S. In fact, when it is possible (4S and less) I rather skip first two cell slots and use Vb3~Vb6 to measure voltages and do the balancing. FW handles these skipped slots well.

[alex alex]

Hi,

So for a 4S pack (for example) I connect main leads to the pack and the balance port to 3-6 (skipping first two) and start charging. No firmware/hardware mods are needed? Can you confirm please?

Thanks.

[diggit]

So for a 4S pack (for example) I connect main leads to the pack and the balance port to 3-6 (skipping first two) and start charging. No firmware/hardware mods are needed? Can you confirm please?

Look here. Commits enabling this feature are linked in the end of bug report. No HW mods needed.

[jusubbi]

Now i'm pretty happy to this charger. After modifications i can charge or discharge 5A or 50W (originally discharge 1A or 5W).

I also use only balance ports 3-4 for sensing voltages lipo and port 6 to 1S lipo (i don't need more than 2S). There is less voltage error than use ports 1-2.

External temp sensor also working now. There was defective diode and i replaced that. Here is some pictures:

draining 12v car battery 4.2A

draining 12v car battery 50W

discharge FET temp

discharge resistor temp

heatsink temp

charging 12v car battery 5A

charging 12v car battery 65W

charging resistor temp

charging diode temp

I noticed that charging is 65W not 50W, is this normal?

[Paweł Si]

2016-03-09 14:02 GMT+01:00 jusubbi <jus...@gmail.com>:

I noticed that charging is 65W not 50W, is this normal?

hm... could you check the output voltage with an multimeter,

and check "Vout=??" on one of the screens?

is it possible that you have:

ENABLE_DYNAMIC_MAX_POWER

enabled?

[jusubbi]

I have very strange problem after while of charging battery. Wires R shows 0 Ω and battery R looks fine. Charger still trying to charge without error, but there is no current at all.

[Paweł Si]

2016-03-15 15:26 GMT+01:00 jusubbi <jus...@gmail.com>:

I have very strange problem after while of charging battery. Wires R shows 0 Ω and battery R looks fine.

Rwire=0Ω because Iout = 0A,

we calculate Rwire using the formula:

Rwire = (Vout-Vbattery)/Iout

where Vbattery - is the balance port voltage.

if Iout == 0A we just display 0Ω to not  divide by zero.

so it's probably not related to your problem,

 

Charger still trying to charge without error, but there is no current at all.

when we reach 4.2V on at least one cell, we set Iout to 0A for a moment (to measure the battery's internal resistance),

Iout stays at 0A until all cell voltages stabilize (until voltage difference between two measurements is less than 6mV)

could you check how the cell voltage behaves after Iout goes to 0A,

 

[jusubbi]

could you check how the cell voltage behaves after Iout goes to 0A,

 

 Both Vout and Vbattery are same 3.784v. Difference between those two point are 0mV. I charge only one cell and use Bal6 to measure battery voltage.

[jusubbi]

calibration menu shows:

vb1 0

vb2 0

vb3 4mV

vb4 6mV

vb5 9mV

vb6 3783mV

v1-6 3783mV

vout 3786mV

There is no battery @ vb1-vb5 only vb6, but there is still some voltage?

[Paweł Si]

2016-03-16 7:59 GMT+01:00 jusubbi <jus...@gmail.com>:

 Both Vout and Vbattery are same 3.784v. Difference between those two point are 0mV. I charge only one cell and use Bal6 to measure battery voltage.

maybe there is a bug in the firmware,

could you make LogView log of this charge and send it to me?

2016-03-16 8:07 GMT+01:00 jusubbi <jus...@gmail.com>:

calibration menu shows:

There is no battery @ vb1-vb5 only vb6, but there is still some voltage?

that should not matter, if it does (and there is a bug) I should see it on the LogView logs.

[jusubbi]

Here is log

[Paweł Si]

2016-03-17 9:37 GMT+01:00 jusubbi <jus...@gmail.com>:

Here is log

just as I thought, we assume that the battery voltage will be stable at some point in time: (see pic)

maybe this assumption is a little bit to strong.

I will fix this issue this weekend.

[jusubbi]

Ok, good to know if that problem can solved with software. Here is another log, where charging is normal about 30min then come this problem, but this time it recovers by itself. Later same problem came back and this time it won't recover.

[jusubbi]

Here is longer log. Error occurs ~2h point. Does that R Wire looks normal to you?

[jusubbi]

Did you fix it? If not, can you tell me where i can find this value from source? Then i can try to build hex which have 5A or 50W discharge limit and this voltage assume fixed.

[jusubbi]

can someone help me? i can't use this charger, because charging stops randomly.

[Igon]

 >> i can't use this charger, because charging stops randomly.

How does it look like?
Charger says - "complete" or continues "charging" but without current?

[jusubbi]

continues "charging" but without current.

[Igon]

Try to replace src\hardware\atmega32\generic\50W\SMPS_PID.cpp
With attached one SMPS_PID.cpp
It will not fix the initial reason but can be helpful
(probably the reason is a spurious ADC readings)

Modification prevents the protection reaction at the single wrong reading

[jusubbi]

i tried that modified SMPS_PID.cpp, but same problem still occurs.

later i changed voltage sensing wires from bal 5-6 to bal 4-5 and now looks like it works without problems.

next thing that i want to do is make possible to measure temperature and use serial port same time. because i don't need all balance ports i can use one of them to temp measurement. hardware mod is easy: just need add extra 3 resistors and one diode.

i just need some help what i have to change from src to make this happen?

[Igon]

>>later i changed voltage sensing wires from bal 5-6 to bal 4-5 and now looks like it works without problems.

strange..
I did not understand why current becomes 0 and not equal to min Ic: from charger's settings at this case

>>i just need some help what i have to change from src to make this happen?

You need somebody more experienced in charger's ADC routines - "bal 5-6" readings are used at feedback, calibration, etc and must be carefully disconnected from all of it