(Request)Charging from unreliable/variable amperage and Pali's new BME
 

The old BME supplied by Nokia would reduce its amperage draw in stages if it detected a low voltage situation, it would remain at this low draw even if the available power increased.
Is there a way to script the new replacement BME to automatically poll the available power say every five minutes to see if more power is now available?
My application is a bicycle hub-dynamo driven USB power converter, it can provide up to 1500mA at 20km/h but if I come to a stop or even slow down this obviously will drop off, the N900 will drop to nearly zero draw and will not recover unless disconnected and reconnected at speed, a difficult trick while moving.
I am sure others are using solar, or even the voltage drop when starting a car might cause the N900 to enter a low amp charge state.

Re: (Request)Unreliable charging source amperage and Pali's new BME
 

Really? I had no idea. That sounds odd, I cannot fathom what the rationale for that might be.

Now THAT is definitely odd.

That's a bummer and I agree a software solution would be preferrable but in the absence thereof, you can try a hardware alternative: a little gizmo that would cut the power off completely in a voltage drop situation, reconnecting it once the speed (voltage) picks up again, thus simulating your manual disconnecting and reconnecting the phone. A DIY solution with two transistors and a bunch of resistors would do the trick quite nicely.

Re: (Request)Unreliable charging source amperage and Pali's new BME
 

It is intended to quickly determine the max output amperage of a charger by sensing when it is overdrawn causing the voltage to drop below ~5v. Since the phone assumes a mains or automotive charger producing stable power to its maximum output this is actually a wise decision which doesn't over stress a cheap or small charger. The assumption only falls apart when it has to deal with severe brownouts, a cloud passing over a solar cell, or something like my bicycle generator charger when I reach a stop or steep uphill.

Re: (Request)Charging from unreliable/variable amperage and Pali's new BME
 

I see! Now I understand why it does not try to increase the current when the voltage rises again.

But, sir, if your converter behaves like that then I regret to say that it is cr@p. It should provide hard 5V or cut off if it can't. IMNSHO.

Re: (Request)Charging from unreliable/variable amperage and Pali's new BME
 

If a software solution is needed to secure having a maximum possible charging then there should be one.

For your use case I think the best and easiest solution would be a buffer like Busch&Müller Art. 461A or any other capable of being charged and output at the same time. That way your voltage won't drop when stopping at a crossing.

Re: (Request)Charging from unreliable/variable amperage and Pali's new BME
 

I have considered adding a 3.7v to 5v converter wired directly to the cells in a 11200mA/h battery pod, I am already using a B&M E-werk converter but in mountainous areas the relatively small cache battery would be drawn down pretty quickly and experience significant wear in my case. I would prefer having the N900 watching for undervolt but resetting maybe once every 60-300 seconds.
Like with the Qi thread there is a hardware solution but the software solution can be almost as good and probably easier.