[NTLK] Power Adapter Developer's Guide

Jake Bordens jake at allaboutjake.com
Sat Apr 14 17:55:05 EDT 2018


I did a little more thinking and investigating.

Quote the PADG (Power Adapter  Dev Guide) "To inform the N2 system, the device must present a 10 ma current limited voltage at the same level as unFusedPSVin to the signal PsAdptIn. This can be accomplished by placing a 100K Ohm resistor to this PsAdptIn and the voltage source. If this current limited voltage is not presented on this signal, damage to the N2 device will occur."

A 100k Ohm resistor is a pretty big resistance. With a 7.5v adapter would be 0.075mA nowhere near the 10ma current limited voltage that they mention in the same paragraph.  So I tried a 1K resistor instead.  That puts it more in the 7.5mA range... and bingo.  The OS does recognize the power adapter.  It goes from "Waiting" to "Charging" in the UI  and even "Adapter" if you eject the battery and drop it onto the floor, breaking your battery into pieces.

Side note, there are other errors in the PADG, including the pinout of the internal serial slot.  Pin 16 is not a ground, is another unfused battery pin. (oops)

So it does work.  I'm not sure what happens when the battery gets to full charge, but I'd hope that it gracefully stops charging the battery.

More experiments when I can find an alternative battery pack.  I have a few packs that can be re-celled somewhere, but it’s a project I wasn't expecting to have to do just now.

Jake
 
On 4/14/18, 10:07 AM, "Jake Bordens" <jake at allaboutjake.com> wrote:

    Elias,
    
    Thanks for the reply.  Here are my concerns with figuring this out.
    
    1.  The power pins act like a voltage source when no adapter is connected and a sink when no adapter is present.  This means any internal circuit board can't just simply have the recommended 100k resistor between the power pins and the "adapter present" sense pin because then the battery voltage would be proving the sense voltage even when no adapter is present.  I'm currently playing with a circuit with 2 diodes such that the adapter can power the power pins but when disconnected, the power pins cannot send voltage back through the sense resistor.  However, after diode and resistor voltage drops, the sense pin comes out a handful of mA lower than the power pin.  Another idea I had was 2 mosfets to switch both the power pins and the sense pins so that they go active at the same time, but I'm not sure if that’s overkill.
    
    2.  In my early testing, the UI hasn't shown any difference when power is connected along with a 100k resistor to the sense pin. I had hoped that the UI would show "charging" just like an external adapter.  I worry that I already burned out my sense pin because a briefly had it hooked up to 3.3v (no current limiting) thinking it was a digital input (but half of me thinks this was unlikely to do damage, but I'm not sure). (At one point I thought I had sent battery voltage to the Newton 5v rail, but the newton still works so I guess I ether didn't or I cought it before damage occurred-- oops)
    
    3.  I'm considering USB-C's power delivery ability to provide 9v 3a and then a buck regulator to get down to 7.5v at 1a - I have a USB-PD PHY breakout board coming to play with and see what's possible.  
    
    If you'd like to chat you can find me on the Newton slack at https://applenewton.slack.com
    
    Jake
    
    On 4/14/18, 9:48 AM, "newtontalk-bounces at newtontalk.net on behalf of Elias Basse" <newtontalk-bounces at newtontalk.net on behalf of kd5jfe at gmail.com> wrote:
    
        Jake,
        
        What test equipment do you have?
        
        A circuit like this is fairly simple in components and should be able to be tested with a good digital multi meter on ampere settings into a load resistor.
        
        If you have a current and voltage limited adjustable bench power supply it could be mimicked from that document without anything other than a wire and connector.
        
        For a simple usb charger, I use a buck boost board that has a current limiting function.
        
        While I have not done so with the Newton I will look over the spec sheet and see if I can get you a bit more information.  The sense pin that is limited to 10ma seems strait forward and can likely be done with a resistor of calculated value similar to a current limiting resistor for a simple LED.
        
        I like the idea of the charge and sync board via usb or usb c.  A 1 amp usb circuit would charge a bit slower but be safer as usb c is thousands of times faster than any port on the Newton could go and has a higher voltage.  At 5VDC you can buck boost to the required 7V range as long as it’s current limited into the ranges as described in the document and be a bit safer.
        
        I will throw this on the bench here and see if I can grab some info from testing that may help you out.  I do lots of electronics as I work on arcade equipment, computer equipment, and do board level repairs for work.
        
        Best Regards,
        Elias
        
        Sent from my iPhone
        
        > On Apr 13, 2018, at 5:49 AM, Jake Bordens <jake at allaboutjake.com> wrote:
        > 
        > Hi folks.  Are there any “real” electrical engineers out there that would be willing to chat with me on Slack about the N2 Power Adapter Guide?  (I’m just a hobbiest maker at best)
        > 
        > http://www.unna.org/unna/apple/documentation/n2platform/n2-padg.pdf
        > 
        > Specifically, I’m exploring the ability of some pins on this port to accept a charger input.  I’m exploring the idea of whether a board could be made that provides a USB-C port that can both provide a standard serial port as well as charge the Newton.   USB-C can provide 9v 3a, and that can be regulated down the 7.5v needed by the Newton.
        > 
        > I’ve played with it a little bit so far, but I don’t have a good way to test that this input is working properly and not in fact doing damage or overcharging the battery.  I had hoped when power was applied to this port, the newton would show “charging” in the battery meter, but it doesn’t see to do that.  Either it doesn’t work that way, or I’ve already damaged the port on my test newton ☺
        > 
        > Also they talk about using a 100k resistor to signal to the newton that there’s an charger present.  In a final design, I think you need more than just a resistor. Since unfusedPsVin provides battery voltage with no adapter is present, putting a 100k across to PsAdaptIn when no power is present seems wrong to me, as the battery voltage could trigger the PsAdaptIn signal when no adapter is present.  I think you need a diode at the minimum such that when the power supply is disconnected, no voltage appears back to the PsAdaptIn pin from unfusedPsVin.
        > 
        > It anyone has any idea on how this might work, how to test safely without destroying a Newton, let me know.
        > ----------------------------------------------------------------------
        > 
        > http://newtontalk.net/
        > http://twitter.com/newtontalk
        ----------------------------------------------------------------------
        
        http://newtontalk.net/
        http://twitter.com/newtontalk
    
    



More information about the NewtonTalk mailing list