The way I understand it, the regulation that already occurs within the
application can only operate within a tolerance of the regulating IC's
nominal input voltage.
Not exactly. There are precious few (if any) three-terminal regulators that
won't handle a 30V or 40V input for 15V out. Your configuration is far more
likely to suffer from UNDER-voltage out of the rectifier arrangement.
The application is using 15 Volt DC regulators
(output), after the bridge rectifier, and can only operate (+/- x%) of 15
volts DC input.
No, three-terminal regs don't work that way. Get a datasheet for a 7815 or 7915
and have a good read of the input voltage range. Let's see, 7815 ...
Dropout voltage (typical): 2.0V
Vin (max): 35V
So a 7815 will regulate as long as the input voltage doesn't exceed 35V or drop
below 17V on the troughs of the input waveform.
The rectification of (nominally) 9 VAC I understood to result in 18 VDC
(ignoring any small voltage drop across the diodes).
It may be small but it isn't insignificant. And I don't like your rectification
theory either.
So even when the input voltage is nominal,
the input voltage to the regulating IC's is only within 20% of their nominal
output.
The Vin-Vout capability of (again, 3-terminal) regs is a spec figure you'll find
on the data sheet, often termed drop-out voltage. See above. It's *not* a
percentage thing. If your input drops (even on transients) to 2V above your 15V
output, or less, you will lose regulation. 20% of 15V is 3V which - if your
rectifier is really providing 18V DC - is barely enough, and ripple at any sort
of load will kill you.
Now what happens if the domestic supply voltage is not nominal?
If the domestic supply comes in at 264 VAC (+10%), the existing linear
regulator
I think you mean transformer
now supplies 9.9 VAC to the application, the bridge rectifier
doubles that to 19.8 VAC.
I don't like your rectifier model, but let's ignore that for the moment.
The input voltage to the regulating IC's is now
only within 40% of their nominal output!
They should be happier. Warmer, but happier. Realise also that the output of
your rectifier system will show significant ripple as load current increases,
and those dips threaten the very regulation those regulators are trying to
provide.
Your concept of how regulators work is more than a worry. And IMNSHO it
unperpins all your problems.
I monitor the domestic supply and regularly see voltages coming into our
house outside the range of +/- 10%.
One solution, of course, would be to redesign the power supply within the
application's case, including upgrading the regulating IC's input voltage
tolerance to at least 40% of its output, if such an IC can be found!
Read the data sheet, and understand what a 3-terminal reg does.
However space within the applications case would not allow for a toroidal
transformer, which I consider to be the optimum solution for a number of
reasons that I don't need to explain here (or maybe they will insist that I
do that as well?!!), and I have already invested considerable effort (and a
small amount of money) in upgrading the capacitors within the existing
application's power supply and really don't want to loose out on the already
considerable improvements I have made therein.
Throwing bulk cap around will obviously improve (aka mask) the regulation
situation. But it is masking the symptoms, not fixing the cause.
You mentioned in another post that this is accompanying a move from a 9VAC
wall-wart to a hopefully better transformer.
What I sincerely suggest you do is (in order):
(a) Get the data sheets for the regulators that you are using. If they aren't
7815/7915 then I feel compelled to ask why not.
(b) Measure (even with a DMM) the input voltage to the regulator(s) and
convince yourself that you have enough headroom to operate properly. Better
still if you check the ripple with a CRO too. Even better, get a CRO and watch
the waveform applied to the input of the regulators, and load/unload teh system
so that the ripple can be quantified.
(c) Check whether the regs are running hot to touch.
(d) Try a 12VAC wall wart, unless the DCin to the regs is already 23V or more.
Quite frankly, I am astounded and reviled that such a simple request has
generated such a vile reaction from this group. I can only assume that it is
because they, like me, were unable to Goggle a suitable design. So as to
maintain their allusion of expertise and so they hide their inability to
self-design a suitable solution, with insults!
Quite frankly, I am astounded that you seemed to think that these problems are
best solved by shooting for an (affordable) transformer offering 1% line
regulation. If that were the best solution, such transformers would abound, and
Google would have turned up a mutitude of solutions for you. Rather, the
solution to such requirements have been de rigeur for decades, earning the tag
classical.
Your request was further obfuscated - and the task of steering you towards a
sound and achievable/affordable solution - by giving more of your planned
solution than the description of the application and the problem.
I wonder what would happen if I took a simular request for such a simple
thing to another newsgroup, say 'us.electronics', and pointed out to them
that 'aus.electronics' does not have a single contributor who can satisfy
the request. I wonder what they would say. "Too difficult", do you think?
No, they'd wonder why you were searching for an electronic sledgehammer.
Whether you choose to follow my, or anyone else's, suggestions is up to you.
But among the flaming, you have been given more than enough leads to achieve a
sensible outcome.
And please don't top-post
--
