[mdlug] Old 386 needed - Power Supply

[Aaron Kulkis]

Raymond McLaughlin wrote:
> Jay Nugent wrote:
>> Greetings,
>>
>> On Tue, 21 Apr 2009, Aaron Kulkis wrote:
>>
>>> If all else fails, AC -> DC power converters are extremely
>>> easy to make -- and you can probably make one which is
>>> significantly more robust, both electrically, and mechanically
>>> than what you buy at a computer store, for approximately
>>> the same cost from parts available at any Radio Shack.
>>>
>>>
>>>
>>> The circuit is this
>>>
>>>                   +-------+
>>>     o--)||(-------| ~   + |------+-----------+-------o  +
>>>        )||(       |AC   DC|   |  |           |
>>>        )||(       |       |   +-----+        |
>>>  AC    )||(       |       |      ^  |    ----+----     DC  
>>>  in    )||(       |       |     / \      ----+----     out
>>>        )||(       |       |    /___\         |
>>>        )||(       |AC   DC|      |           |
>>>     o--)||(-------| ~   - |------+-----------+-------o  -
>>>     Transformer   +-------+    Voltage   Capacitor 
>>>                   Rectifier   Regulator   (Large)
>>>                    ($2.50)  (Zener Diode)
>>
>>    Close, but no joy.  You *will* need a current limiting resistor in 
>> series between the transformer and the zener diode.  When the zener 
>> reaches cutover it cannot handle the full output of the transformer, so a 
>> current limiting resistor is needed.  You can *almost* think of this as a 
>> voltage divider (using two resistors) where the voltage drop across one 
>> resistor (in this case the zener) is always constant.
>>
>>    But over, Aaron is correct.  You *CAN* build a perfectly good power 
>> supply for FAR less than the cost of an off-the-shelf laptop supply.
>>
>>    Note:  Many laptops could accept a very wide range of input voltages 
>> simply because they had their own internal regulators.  I have run many 
>> old 386/486 laptops that said they needed 16-20 volts on a 12 volt supply.  
>>
>>
>>    If the voltage is too low --- the laptop simple won't work or will fail
>> when a floppy drive spins up, or something requires a little more uumph
>> causing the voltage to sag and the motherboard resets/reboots.
>>
>>    If the voltage is too high --- well, that's a different matter as the 
>> "magic smoke" may leak out and it will never run again :(
>>
>>    
>>    Enjoy!
>>       --- Jay Nugent  WB8TKL
>>
>> Train how you will Operate, and you will Operate how you were Trained.
>> +------------------------------------------------------------------------+
>> | Jay Nugent   jjn at nuge.com    (734)484-5105    (734)649-0850/Cell       |
>> |   Nugent Telecommunications  [www.nuge.com]                            |
>> |   Internet Consulting/Linux SysAdmin/Engineering & Design/ISP Reseller |
>> | ISP Monitoring [www.ispmonitor.org] ISP & Modem Performance Monitoring |
>> | Web-Pegasus    [www.webpegasus.com] Web Hosting/DNS Hosting/Shell Accts|
>> +------------------------------------------------------------------------+
>>   3:01am  up 13 days,  2:08,  3 users,  load average: 1.10, 0.26, 0.08
> 
> Thank you Aaron and Jay,
> 
> I have been, and continue to mull a more detailed reply, but will here
> quickly point out that I have, on hand, several serviceable AT and ATX
> powers supplies. Each of these provide both a +12 (yellow) and a -12
> (blue) output. Shouldn't this look like 24v to any DC device connected
> to it? I might want to drop this to 22v for the laptop. The circuits you
> describe are something I "used to know" but need some brushing up on.
> 
> Just quickly, shouldn't the 24 volts described above provide a
> satisfactory "first approximation" of the 22 that the laptop requires?
> 
> Also any comment on my supposition that "pin positive" is more common
> than the reverse?
> 

I would try a moderate below-operating-level voltage (say, 6v)
with both pin positive, and ring positive, in series with
an LED, and see which way lights up the LED.

If neither way lights up the LED, then you have the polarity
of the LED backwards.