Building switchers with LT1070
Eric Williams
have a couple in my parts box) but was concerned about how hard it might
be to make a reliable regulator. Anybody here have experience putting
these chips to use? Is layout critical, or can mere mortals armed with
a soldering iron expect success? TIA.
Rene Tschaggelar
It depends what you call success. Getting something out at all - yes.
tuning the lot for very little noise and spikes it a bit harder.
On the other hand there are dozends of application notes available.
Rene
--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
Greg Neff
wrote:
We have used this regulator and its derivatives in a few different
designs over the years. Our experience has been very good. The
boards have been stable and reliable in the field.
Linear Tech has an excellent application note (AN19 - LT1070 Design
Manual), plus free software that does the math for you to select
first-pass component values.
Personally I would use the LT1170 series, which operates at 100KHz
instead of 40KHz.
See the application note regarding layout requirements. I have never
attempted a breadboard implementation of a 5A switching power supply,
and I'm doubtful that this would work reliably.
===================================
Greg Neff
VP Engineering
*Microsym* Computers Inc.
gr...@guesswhichwordgoeshere.com
Eric Williams
> We have used this regulator and its derivatives in a few different
> designs over the years. Our experience has been very good. The
> boards have been stable and reliable in the field.
>
> Linear Tech has an excellent application note (AN19 - LT1070 Design
> Manual), plus free software that does the math for you to select
> first-pass component values.
Got it, and I'm working my way through it, but I was interested in
real-life anecdotes.
> Personally I would use the LT1170 series, which operates at 100KHz
> instead of 40KHz.
>
> See the application note regarding layout requirements. I have never
> attempted a breadboard implementation of a 5A switching power supply,
> and I'm doubtful that this would work reliably.
Fortunately, I only need an amp or less. I don't absolutely need a
switcher, but thought it would be a good opportunity to get my feet wet,
as long as it wasn't going to turn into a hairy mess. Thanks!
John Popelish
I don't have specific experience with this chip, but lots of
experience with buck regulators. Layout is pretty important for all
of them. For example the node that connects the input bypass
capacitor, the common end of the catch diode, the regulator chip
ground reference and the output capacitor should be very compact.
Likewise for the node that connects the output of the switch, the coil
and the catch diode. These two nodes dominate most layout problems
with these regulators. This applies to circuit board layout and air
circuits.
The other important thing to do is to parallel the input and output
caps with a good low inductance film capacitor. I like the Panasonic
V series stacked film units. Lots of people swear by ceramic
capacitors, but I have seen lots of trouble with high ESR ceramics,
unless you use the X7R (10%) types. Z5U and Y5V
are usually crap.
--
John Popelish
Mike
Mike
"John Popelish" <jpop...@rica.net> wrote in message
news:3ADFB2F8...@rica.net...
John Popelish
>
> And keep the feedback signal FAR away from the coil!
>
is used, it is pretty well self shielded. If a drum core is used, the
best way to get past the coil may be to go directly under the center
line. Running around the edge is the way to couple lots of the
switching waveform into the signal lines. Drum cores are cheap but
really bad as far as external fields are concerned. I have had to
mount them on top of thick spacers to reduce coupling into board
traces. Two piece cores that enclose the coils are much less trouble.
--
John Popelish