Searching \ for ' [PIC] [EE]: Improved LDO regulator circu' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: '[EE]: Improved LDO regulator circu'.

No exact or substring matches. trying for part
PICList Thread
'[PICLIST] [PIC] [EE]: Improved LDO regulator circu'
2002\04\06@175106 by Tom Messenger

flavicon
face
I hadn't paid attention to the TIP30 base resistor, 15 ohms in the version
you sent out. After you mentioned it in your last email about it
(especially the mention of zero ohms would be ok aside from output shorts),
I took another look. A very low value there has a side effect - maybe ok,
maybe not.

It is a 5 volt LDO so a likely use is with a 6 volt battery. At the point
where the input drops below what it takes to get the zener going the
quiescent current goes up sharply. My model showed it going up to almost
200ma with 10 ohms there. What you will actually get will be determined by
which particular 2N3904 and TIP30 you use. As soon as the battery gets a
little low, it gets much higher drain. If the battery is supplying a 250ma
load, at low line, the battery must supply 450ma.

If you have yours on your breadboard, maybe you can check to see if this is
simply spice gone bad.

Best regards,
Tom M.

--
http://www.piclist.com hint: To leave the PICList
spam_OUTpiclist-unsubscribe-requestTakeThisOuTspammitvma.mit.edu


2002\04\06@194935 by Dave Dilatush

picon face
The behavior you're describing (a large increase in quiescent
current as the pass transistor goes into saturation due to a
low-input condition) isn't "spice gone bad"--it's very typical of
low-dropout regulator ICs as well, although they usually include
some mechanism for moderating it.

I use National's LP2951 LDO IC in a lot of my designs and it does
this too; but the peak quiescent current is only about 25% above
that required to sustain the applied load in a non-saturated
condition.  They do this by sensing when the LP2951's pass
transistor is saturating and reducing the base drive as its Vce
drops toward zero.

The LP2951 data sheet has an internal schematic of the part that
shows how this is done.  The same technique could be applied to
Fr. Tom's circuit, though that would add yet another transistor
(or maybe several) and that's not good.

Dave

Tom Messenger wrote...

{Quote hidden}

--
http://www.piclist.com hint: To leave the PICList
.....piclist-unsubscribe-requestKILLspamspam@spam@mitvma.mit.edu


2002\04\07@124957 by Thomas McGahee

flavicon
face
In practice it makes good sense to use a base resistor that is just
large enough to provide the max base current to the TIP30 that will
give you an output of 5.00 volts at the desired load current.

What I SHOULD have said in my post describing the circuit operation
was that under NORMAL operating condition the value of the base
resistor makes little difference, though it must be small enough
to supply the max base current required. Anything smaller than that
doesn't gain you anything (and even zero ohms WORKS), but as you
pointed out, if it is too low it will cause a rise in base-emitter
current once the input drops below the desired output voltage.

Note that the best value for the TIP30 base resistor depends on
the max input voltage and the max output current. Once you know
what your operating parameters are you can then select a best-fit
resistance value.

By the way, I fired up Electronics Workbench and simulated
the circuit. The simulation and the actual circuit agree within a
percent or so. I don't really like the version of the software that
I have too much, though. It uses a copy protection scheme
that requires me to look up a word in the manual about every sixth time
I use the program. It only allows one instance of things like the
oscilloscope (though you can have as many voltmeters and ammeters as you
want). The device selection leaves much to be desired. But my
biggest gripe is that it doesn't handle things like inductive
buck/boost circuits very well. In fact, most of the time it won't
simulate them at all. What it does, it does quite well, but I think
it is about time for me to find a better tool.

Fr. Thomas McGahee

{Original Message removed}

2002\04\07@150805 by Larry Bradley

flavicon
face
Father Tom - I've just started playing with LT Spice - from the Linear
Technologies Web site. It is intended to be used to design switching
regulators (using their chips, of course :)  but seems to work quite well
with other things. The price is right.

http://www.linear-tech.com/software

They call it SwitchCAD


Larry
Larry Bradley
Orleans (Ottawa), Ontario, CANADA

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\04\07@151359 by Tom Messenger

flavicon
face
At 12:56 PM 4/7/02 -0400, you wrote:
>By the way, I fired up Electronics Workbench and simulated
>the circuit.

-snip-

>but I think
>it is about time for me to find a better tool.
>
>Fr. Thomas McGahee

I'd recommend LT SwitcherCad which despite the name, is a full spice
implementation. Plus, it really does switchers nicely too.  I've mentioned
this program several times of late and someone might well get the
impression that I'm recommending it in place of circuit testing with real
parts. I'm not.

It is a very good tool to try out many ideas and see which ones look
promising.  For beginners, if their circuit doesn't work like they thought
it would, sometimes they are at a loss to understand what's going on. But
with a spice run, you can not only see ac and dc operating points of any
and all voltages but it shows currents as well.  This can be *very*
illuminating when modeling switchers as it lets you see the peak to peak
ripple current your circuit will run at and tells you what you can get away
with regarding inductor and capacitor selection. Not to mention more
troubling parameters like esr, etc.

But by all means, it is only a starting point. There's no substitute for
real parts running from real electricity. Plus the fun of generating real
heat burning real fingers. :)

Best regards,
Tom M.

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\04\07@163656 by Peter L. Peres

picon face
>The behavior you're describing (a large increase in quiescent
>current as the pass transistor goes into saturation due to a
>low-input condition) isn't "spice gone bad"--it's very typical of
>low-dropout regulator ICs as well, although they usually include
>some mechanism for moderating it.

For which well-known reason LDORs meant to be used at near full capacity
use a low Rdson P-FET instead of a bipolar. This usually buys low
quiescent current too. The Seiko LDOs are built like this afaik.

Another feature of these CMOS regulators is the very low temperature
coefficient (wrt discretes and bipolar usual regs).

The simplest CMOS LDO is a depletion type P-FET in series with the load
and an enhancement type FET shunting its gate to GND. Add three
polarization R's (one pulls the pass transistor's gate up to S, and the
other two form a divider from the output voltage to feed the gate of the
error amplifier (the enhancement FET) - of course they are actually just
other FETs in pinch resistor mode when integrated).

A more evil version of this scheme uses a back-diode-less depletion mode
P-FET (I admit I've never seen one of those but they exist), and the
divider alone. The FET must be diode-less because it runs backwards (S to
load, D to +Vcc, G to GND or divider between S and gnd). The FET will
stabilize to its threshold voltage between S and G which is the output
voltage (unless you use a divider for higher V).

And the last and most evil scheme consists of a single NJFET (like BF244
or 245 or 2N equivalents) with G to gnd and D and S to load and Vcc
respectively, and NO other parts. Beware of decounpling input and output
properly as this is also a very popular UHF oscillator configuration. It
works like the single PFET described in the previous paragraph. By
selecting one with Vp ~= 5V (A suffix on BF245 is usually like this), you
have 5V out. This latter scheme has the strange advantage of also working
backwards ! Swap load and supply at will. I have used it to equalize
charge in two small NiCd accumulators. The last two schemes described have
poor regulation but they have a number of advantages for special
applications (besides being nice ideas).

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\04\07@170015 by Thomas McGahee

flavicon
face
Larry,

Thanks for the url. I downloaded it and hope to play with it
some tonight.

Fr. Thomas McGahee

{Original Message removed}

2002\04\07@170212 by Thomas McGahee

flavicon
face
Yeah! Give me real smoke any day of the week!

(But I admit, a good simulator is a wonderful tool!)

Fr. Tom McGahee

{Original Message removed}

2002\04\07@182042 by Sean H. Breheny

face picon face
Dear Larry (and others who advocate SwitchCAD),

I just downloaded and installed this and used it to simulate a diode RF
switch. It works wonderfully! It is much faster than most expensive SPICE
implementations I have used and it has a wide range of real world practical
parts built into it! For example, instead of having to look up the exact
specs for a particular capacitor, you can just select the exact make and
model from a comprehensive list they have!

Sean

At 03:06 PM 4/7/02 -0500, you wrote:
{Quote hidden}

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\04\07@190259 by Larry Bradley

flavicon
face
And you can download other spice models from other manufacturers and use
them. For example, I downloaded a LF411 op amp from National, then modified
a basic op-amp "symbol" that LTC provided to use the LF411 model and
presto! It works.

Larry


At 06:18 PM 4/7/2002 -0400, you wrote:
{Quote hidden}

Larry Bradley
Orleans (Ottawa), Ontario, CANADA

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads


2002\04\08@075924 by Edson Brusque

face
flavicon
face
Hello Larry,

> And you can download other spice models from other manufacturers and use
> them. For example, I downloaded a LF411 op amp from National, then
modified
> a basic op-amp "symbol" that LTC provided to use the LF411 model and
> presto! It works.

   I've tried to make this some weeks ago with no sucess. Can you provide
more details?

   Actually I'm using B2Spice (pointed by Lawrence Lile) and very happy
with it.

   Best regards,

   Brusque

-----------------------------------------------------------------
Edson Brusque                 C.I.Tronics Lighting Designers Ltda
Researcher and Developer               Blumenau  -  SC  -  Brazil
Say NO to HTML mail                          http://www.citronics.com.br
-----------------------------------------------------------------

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\04\08@085811 by Larry Bradley

flavicon
face
Yup. I downloaded the LF411 SPICE model from National.

Then I copied the "opamp2.asy" symbol and renamed it 'lf411.asy"
Made the following changes in the lf411.asy file(changed the two lines to
read as follows - note that in the SPICE model file the LF411 is actually
named LF411/NS):

SYMATTR SpiceModel LF411/NS
SYMATTR Value LF411/NS

Finally, you need to add the following SPICE command to your SwCad simulation

.lib LF411.sub

Larry

At 08:46 AM 4/8/2002 -0300, you wrote:
{Quote hidden}

Larry Bradley
Orleans (Ottawa), Ontario, CANADA

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\04\09@190122 by Barry Gershenfeld

picon face
>And the last and most evil scheme consists of a single NJFET (like BF244
>or 245 or 2N equivalents) with G to gnd and D and S to load and Vcc
>respectively, and NO other parts. Beware of decounpling input and output
>properly as this is also a very popular UHF oscillator configuration. It
>works like the single PFET described in the previous paragraph. By
>selecting one with Vp ~= 5V (A suffix on BF245 is usually like this), you
>have 5V out. This latter scheme has the strange advantage of also working
>backwards ! Swap load and supply at will. I have used it to equalize
>charge in two small NiCd accumulators.
>Peter

This sounds like something I ran across that lets you connect
a bus running at 3.3v with one running at 5v.  Bidirectional!
Six or eight of 'em on a chip (made by Quality Semiconductor,
who, the Web claims, got bought by IDT).  And it is nothing
more than a FET, connected in similar fashion.

Barry

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\04\10@174531 by Peter L. Peres

picon face
>This sounds like something I ran across that lets you connect
>a bus running at 3.3v with one running at 5v.  Bidirectional!
>Six or eight of 'em on a chip (made by Quality Semiconductor,
>who, the Web claims, got bought by IDT).  And it is nothing
>more than a FET, connected in similar fashion.

I suspect that there is more to it. I know three types of devices that
resemble what you describe:

a. Bus separators/surge suppressors, which indeed contain one MOSFET (or
two in antiparallel switch config) for each pass channel and a common gate
that can be used to switch the whole thing (it also includes surge
suppressors). It can switch analog and digital signals. Lots of uses for
hot-pluggable boards and systems where parts shut down to save power.

b. Bidirectional level converters that use several (4) MOSFETs. These are
two each complementary MOSFET drain followers, wired in antiparallel.
Since each has subunity gain they are stable. They have different input
thresholds and gains to adjust for the level translation.

Maybe this can be done with a single transistor as you say. I suspect that
there will be problems with this, or that it may be hard to do, at least
for bidirectional drive (sink & source).

Me I like the Seiko (CMOS) LDOs very much because the LSB of an 8bit A/D
converter will not be affected over the entire commercial temperature
range usually. This is a big headache with usual bipolar LDO regs, because
LDO does not automatically mean low tempco. 1LSB @8 bits and 5V is ~20mV
and most bipolar LDOs will change output far more than that over 0 to 70C.
Any non-ratiometric calibration you have on the device will be off.

Peter

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listservspamKILLspammitvma.mit.edu with SET PICList DIGEST in the body


2002\04\10@192500 by Barry Gershenfeld

picon face
I found the relevant document and it's not quite magic.  I think the
5v side would see a 3v drive signal.  But it's pretty close and
I did use it in a project some time ago.

http://www.idt.com/docs/AN_11.pdf

Barry

At 10:25 PM 4/10/02 +0300, Peter L. Peres wrote:
{Quote hidden}

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email .....listservKILLspamspam.....mitvma.mit.edu with SET PICList DIGEST in the body


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...