Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

Hi, I trying this project for load battery on the road.
5V is generated without problem, the signal D- and D+ has 1.9V, when I try to attach N900 to my step up the charge not start. In dmesg I see:
I tried also with 0.7V on D- D+ as written in batt_charging_1_1_FINAL.pdf but nothing.
I have the kernel 2.6.28.10power46.
I would like start the charge with only 500mA and not 1000mA. With 200 Ohms between D- and D+ which is the maximum source current?

Thank you for the help

Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

My guess is that N900 needs more than 500mA to charge, for example I regularly see my N900 fails to charge reliably from USB ports on desktop PCs.

Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

Maybe the problem is desktop pc with windows. In my pc with Linux the charge work always good. I need about 4 hours for full charge.

I discovered somethings. With 200Ohms beetween D+ and D- the charge is 1000mA Wall Adapter.
In this post http://talk.maemo.org/showpost.php?p...&postcount=861 i read the settings of BME on the chip BQ24150. So I tried to stop BME and launch the script, the charge works well, but I don't have none graphical feedbacks.
Can I set the registers of BQ24150 while the BME is running?
Can I tell to BME with socket unix start the charge?

Thank you

EDIT:
In the script I changed the value of register 0x01 with 0x48 (500mA)

Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

bme is not configurable.

Even if the charger can't get sufficient current it'll generally take what it can get, down to 300-400mA or so as long as the voltage is sufficiently stabilized.

Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

This may also help:
http://talk.maemo.org/showthread.php?t=75300

Re: Solution: N900 Charging from dumb power sources
 

Since somebody referred to this thread today, I like to add some late clarifications:

The spec says "SHORT D+ and D-", the 200Ohm are MAXIMUM. For any EE that's a pretty unambiguous spec that defines a valid range of 0 .. 200 Ohm for the shorting - pin to pin, including all cable resistance, contact resistance and all, which is what the 200 Ohm are meant to specify: you must not exceed this value.
Bottom line: using resistor is BAD, short the pins by simple solder jumper.

The max current with shorted (by significantly less than MAX 200Ohm, best value:0) D-pins is basically undefined, N900 will try to draw up to ~900mA from any charger that signals "fastcharger" by the shorting method.

For charging in software, there's shadowJK's http://enivax.net/jk/n900/charge21.sh.txt script now, based on my original draft quoted some posts back.
Alternative source edited by me for 500mA charge: http://maemo.cloud-7.de/maemo5/patch...t__UNTESTED.sh
Keep in mind you probably have to `stop bme`, otherwise the charging script and BME will conflict and not work correctly.

/j

Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

I tried the data pins short trick today, i've heard of it before but never had to use it until now.

I was trying to charge my n900 of a solar battery/charger controller with a usb out, and wasn't getting anything. I tore the usb module out snice it has nice leads attached to it on a pcb, and shorted the data pins with a quick joint.

Hooked it up to a sepic tuned for 5v running off 8x nimh aa's in two sets of four (6v, ~4ah), soldered the leads to the output on the sepic and it charged now.

So what exactly does this mean in terms of mah's being put into the battery with the pins shorted, or time it takes to charge? More/less than the stock 1000ma 5v input?

I assume even if i get the full 2a my sepic is rated for, it should be fine, as i've hooked my n900 alone up to a 2a usb hub and nothing bad happened. I doubt its like a led where it takes all the current it can get, and all that really matters is the voltage (up to a point, super high current could short circut your phone).

Anyway, female usb cables are dirt cheap, this works just fine with an aftermarket usb cable as well. Get a cheap sepic, a lipo pack, and a female usb port and you got a nice little usb backup battery thats not rediculously overpriced, diy, and easy to upgrade parts/batteries.

Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

N900 insructs its charger chip to charge at max 950mA. How much that translates to in the input depends on battery voltage and the efficiency of the charging chip.

Re: Solution: N900 Charging from dumb power sources
 

@joerg_rw: Any chance we can have a fix integrated on power kernel?

Re: Solution: N900 Charging from dumb power sources *UPDATED*
 

https://www.facebook.com/photo.php?f...51890228794497

probably I should have tried the software solutions before demolishing the cable, but I had quite a few spare ones and I enjoy soldering. this one works fine.