[sdiy] problems running circuit off of DC TO DC converter-HELP!

[ASSI]

Dan,

On Wednesday 09 December 2009, dan snazelle wrote:
> I am trying to make a decent sized +/- 15 volt circuit work on a 9v
>  battery/9v dc adapter
> 
> the circuit has got about 10 opamps (all tl062's), a few ota's,
>  some resistors, and a tiny bit of cmos,
> 
> a second possible circuit i am working which would also use a DC TO
>  DC  has about 10 cmos chips and 4 tl072's.
> 
> I am trying to use a CHICON  EC3AE05  +/- 15volt DC TO DC converter
>  as the main part for these designs which is stated as having
>  around 100ma per rail. (these would also work with a +/- 12 volt
>  setup)

Even before going into details: these are 3W parts with a stated 61% 
efficiency.  This is a rather bad value for a DC/DC, so I assume they 
do some linear post-regulation to fend off ripple.  So the input power 
at full load is going to be close to 5W or an input current of 1A for 
the 5V part.  You definitely don't want to use a linear regulator to 
go down from 9V to 5V and burn another 4W in the process, which the 
battery would also need to supply.  And yes, burning 6W in a small 
space is going to rise the temperature quite a bit and no, you can't 
avoid that.  A 9V alkaline block has probably around 300mAh of 
capacity at this high load, so using it up in 20 minutes sounds about 
right.  It will also not put out anywhere near the rated 9V at these 
high currents after about 10% into it's capacity, so it is also no 
wonder that the voltages are sagging when you leave the rated DC input 
range for the regulator.

Depending on how the DC/DC conversion is done, the input current might 
also not be a nice 1A DC, but rather something like 5A impulses at 
some kHz repetition rate that have an RMS value of 1A.  This would 
definitely run afoul with a lot of linear regulators and your battery 
won't like this very much either.

A 9V alkaline battery can have almost 11V when new and goes down to 
5.4V if you really want to wring the last bit of juice out of it (if 
you do this: remove them immediately when empty).  If you later decide 
to use something like lithium polymer or NiMH accumulators for model 
kit, you might even look at something in the 12-16V range.  So better 
use a wide input-range part like this one:

http://www2.mouser.com/ProductDetail/ETA-
USA/OBQ23WC0512/?qs=XjKCEEBoJ0dr5q7vp2LthA%3d%3d

http://www.eta-usa.com/pdffiles/dcdc/OBQ_SC_WC_series.pdf

This also eases the requirements on the DC adaptor.

If you can redesign your circuit to work from +-12V or even lower 
and/or scale up the impedances a bit, this would certainly help to 
ease the power requirements and save some battery life.  A different 
choice of op-amps could be advised to drop the power even further.  
The quiescent current of the TL062 alone will eat around 60mA, so I'm 
suspecting that your circuit takes a bit more than the 100mA you're 
trying to design for and depending on what the circuit looks like, you 
may be loading one of the rails quite a bit more than the other.  Put 
your circuit(s) on a regulated supply and measure the current in each 
rail, then enter the values into the parametric search form and see 
what comes out.  A DC/DC convertor should not be over-dimensioned as 
the efficiency is best at typically 80%-90% load (check the 
datasheet).


Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk]>+

Factory and User Sound Singles for Waldorf Blofeld:
http://Synth.Stromeko.net/Downloads.html#WaldorfSounds