Searching \ for '[EE]: 2-tran SMPS 5v reg now 90% efficicient!' 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/index.htm?key=tran+smps+reg+efficicient
Search entire site for: '2-tran SMPS 5v reg now 90 efficicient!'.

Exact match. Not showing close matches.
PICList Thread
'[EE]: 2-tran SMPS 5v reg now 90% efficicient!'
2002\09\22@165111 by hard Prosser

flavicon
face
Roman
Nice results & well presented  - I had a feeling that this circuit had some
potential!
Your version is now firmly copied into my "useful circuits" file.

Richard P



(Posted on [PIC] and [EE] as this is a 5v regulator design.)

I think i've found the "holy grail" of 2-transistor 5v
SMPS regulators!

After seeing the recent discussion on using a 5v switching
regulator chip I went back and spent a few hours on my
2-tran SMPS regulator with the goals of improving efficiency
snip.....

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


2002\09\22@175127 by Roman Black

flavicon
face
(posted on [EE] too)

I have done some more testing on my new 2-transistor
SMPS 5v regulator, and it continues to look good. :o)

Regulation is now good enough that there is no need for
a secondary regulator like a 7805, and this simple
2-transistor regulator will work at very high efficiencies
and do the complete 5v SMPS buck conversion and regulation
in the one process. No 7805 needed.

I have made 3 "application" circuits, built and tested.
Anyone who needs a high efficiency 5v switching regulator
can simply copy my parts values and it will perform as
advertised. :o)

Already efficiency is better than a LM3485 (switcher chip
that was discussd here recently) over a decent current
range and LM3485 solution uses a FET where my "2-transistor
Black regulator" only uses a cheap BC327. I have had it to
over 91% with the BC327 and will try some high efficiency
tests later once I buy a few common P-FETs.

While i'm testing application circuits I am accepting
suggestions for applications and will build and test any
apps you might have for specific input and output voltages
and currents etc. :o)

It's here:
http://www.romanblack.com/smps.htm
-Roman

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


2002\09\22@180911 by David Harris

picon face
Wow Roman, this looks great.  Nice website!
For higher amperage, can one substitute a more robust transistor for the
BC327.  Eg, for 1A, what would you use?
Thanks for your continuous stable output -- umm kinda like the SMPS!
David

Roman Black wrote:

{Quote hidden}

--
David Harris
 OmniPort Home Page:  http://www3.telus.net/OmniPort1/
   Discussion egroup: http://groups.yahoo.com/group/OmniPort
   Swiki:  http://omniport.swiki.net/1

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


2002\09\22@182003 by Roman Black

flavicon
face
Hi David, thanks for the nice words. For 1A maximum
I think I would use 2x BC327 in parallel and 2x 1N5819
in parallel. This is still smaller and cheaper than
just about any alternative.

I'm going to hunt for some cheap small low-Rds P-FETs,
and see how it goes. :o)
-Roman



David Harris wrote:
{Quote hidden}

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


2002\09\22@185114 by Roman Black

flavicon
face
Richard Prosser wrote:
>
> Roman
> Nice results & well presented  - I had a feeling that this circuit had some
> potential!
> Your version is now firmly copied into my "useful circuits" file.


Thanks for the nice words Richard, and thanks for the
original idea seed too! (Likewise thanks to Russell and
Dave and any others who contributed with ideas for 2
and 3 transistor regulators. :o)

With this circuit I feel that I have spent enough hours
thinking about it over the last 10 months, stripping it to the
bare bones, evaluating the entire operation and then
implementing a clever "monostable" system on the Zener voltage,
along with the current dumps to force extremely rapid switching,
to justify calling this new circuit "my" regulator.
I hope this has not offended you, your original 2-transistor
circuit is mentioned on my web page and I would also have
Russell's 3-transistor high voltage circuit on there but I lost
the gif. :o)
-Roman

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


2002\09\22@190816 by hard Prosser

flavicon
face
No problem - it's " yours"  alright!

Richard P





Thanks for the nice words Richard, and thanks for the
original idea seed too! (Likewise thanks to Russell and
Dave and any others who contributed with ideas for 2
and 3 transistor regulators. :o)

With this circuit I feel that I have spent enough hours
thinking about it over the last 10 months, stripping it to the
bare bones, evaluating the entire operation and then
implementing a clever "monostable" system on the Zener voltage,
along with the current dumps to force extremely rapid switching,
to justify calling this new circuit "my" regulator.
I hope this has not offended you, your original 2-transistor
circuit is mentioned on my web page and I would also have
Russell's 3-transistor high voltage circuit on there but I lost
the gif. :o)
-Roman

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


2002\09\23@012536 by lexandre_Guimar=E3es?=

flavicon
face
Hi, Roman

   Nice going.... Proves that what is good can get even better :-)

> While i'm testing application circuits I am accepting
> suggestions for applications and will build and test any
> apps you might have for specific input and output voltages
> and currents etc. :o)

   Oh !! Now you asked for it ;-)

   I would love to get one for up to 500 ma, 40v max input, short current
protection and overvoltage protection at the output. Let's not forget a
poorly regulated 12 v output. :-)

Best regards,
Alexandre Guimaraes

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


2002\09\23@040151 by Morgan Olsson

picon face
Thank you for your work Roman, this Black regulator really look nice :)

>For 1A maximum
>I think I would use 2x BC327 in parallel and 2x 1N5819
>in parallel.

Warning: Parallelling bipolar transisotrs is usually avoided, since their temperature characterstic will make only one of them operate:  Th eone whichis slightly warmer wil have lower Ube, so, it do more labour than the other, and get still warmer etc.  Finaly you have one very hot tranny and the other just sitting there.  Solution: use separate base resistor(s).

Parallelling bipolar diodes i have never seen, may be same problem, but cheaper to use one bigger eanyway.  But for shottkys like 1N5819 i have done and no problem.

>This is still smaller and cheaper than
>just about any alternative.

Maybe Bc640-16.  (that -16 defines hFE=100..250, plain Bc640 is only 40..250 which would require designing for higher base drive margin)

Btw for even cheaper design use i.e Bc546B for Q2.

-----

Possible problem: verify peak negative Ube of Q2.  I fear that for higher input voltages it will be over spec (which usually is 5V for transistors like theese)  Problem is when it switches off, and C2 pulls it far down.
Securely limited by the zener to less than minus 1V (w regard to neg supply) peak.
But for 5V out we may only go down to neg supply.
If output voltage is 4V or lower then we can rely on the zener to protect Q2 from too high reverse Ube.  But for higher out voltage C1/C2 ratio is to be checked:

Simply, C3/C2 divides the amplitude over D1 to make the amplitude at Q2 base.
Example: at 30V in the amplitude at D1 is almost 31V.
Say we want to design for an amplitude at Q2 base of 5V (Ube then swings between about +1 and -4 V)

Then C2=(C1 * 5/31) ;  If C1=6n8 then C2=1n1

(actually the amplitude will be slightly less due to positive clamping by the zener)

------
Side note: all paragraphs on i.e http://www.romanblack.com/a02.htm gets line wrapped in browser Opera, making a mess of ASCII art and test data. (NO paragraphs are wrapped using Mozilla.)

/Morgan

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


2002\09\23@042935 by Morgan Olsson

picon face
While i think this idea using C2 is really neat, i think it can be substituted for a wire... wound as a feedback winding on L1 and placed in series with Q2 base instead.

+ Cheaper parts (one less component)
- More work (winding)
+ Higher efficiency (no current pumping through C2 and dumped in Zener)


Copying from Romans web page and adjusting:

+    -------*-----------------*--------,
           |                 |        |  Q1
           |                 Rb       |  PNP
           |                 |        E
           R1                *------B
           |                 |        C
           |                 |        |
           |                 Rc       |
           |                 |        |
           |                 |        *--[+L1a-]----*------ +v out
           |                 C  Q2    |             |
           *----*--[-L1b+]-B          |             |
           |    |            E--------|-------------*
           |    |                     |             |
           |    |                     |             |
           |    |                     |             |
    zener  |    |                     |             |
     ZD1   -    |                     -             C 16v
           ^    C1                    ^ D1          |
           |    |                     |             |
           |    |                     |             |
           |    |                     |             |
Gnd --------*----*---------------------*-------------*------ Gnd

Values:

L1b:
Turns: design for about 3v amplitude at highest input voltage.
Maybe 1V is enough (less to wind) but never more than 5V (getting near Q2 Ube reverse limit)

As we are winding this on a inductor with unknown number of turns it is trial, or if you have a signal generator to drive L1a and measure output from i.e ten external turns L1b, then it can be measured.

R1:
No current pumping through C2 and dumped in Zener, so R1 can be increased (doubled?) without decreasing Q2 base drive, = lower loss.

/Morgan

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


2002\09\23@053120 by Roman Black

flavicon
face
Alexandre Guimarães wrote:

> > While i'm testing application circuits I am accepting
> > suggestions for applications and will build and test any
> > apps you might have for specific input and output voltages
> > and currents etc. :o)
>
>     Oh !! Now you asked for it ;-)
>
>     I would love to get one for up to 500 ma, 40v max input, short current
> protection and overvoltage protection at the output. Let's not forget a
> poorly regulated 12 v output. :-)


Hmm, that's an easy one, 40v in, 12v out, 500mA, overvoltage
output. The short circuit protection I can do to, but it will
cost a little efficiency. Regulation should be excellent.

So is max efficiency from 30 to 300ma ok?
With current limiting kicking in at 500mA?
-Roman

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


2002\09\23@054817 by Roman Black

flavicon
face
Morgan Olsson wrote:
>
> Thank you for your work Roman, this Black regulator really look nice :)

Thank you! :o)

> >I think I would use 2x BC327 in parallel and 2x 1N5819
> >in parallel.
>
> Warning: Parallelling bipolar transisotrs is usually avoided, since their temperature characterstic will make only one of them operate:  Th eone whichis slightly warmer wil have lower Ube, so, it do more labour than the other, and get still warmer etc.  Finaly you have one very hot tranny and the other just sitting there.  Solution: use separate base resistor(s).

Yes, apart from you need 2 emitter resistors (not base)
as base resistors will have the wrong effect. :o)

> Parallelling bipolar diodes i have never seen, may be same problem, but cheaper to use one bigger eanyway.  But for shottkys like 1N5819 i have done and no problem.

That's good to hear, thanks!


> >This is still smaller and cheaper than
> >just about any alternative.
>
> Maybe Bc640-16.  (that -16 defines hFE=100..250, plain Bc640 is only 40..250 which would require designing for higher base drive margin)
>
> Btw for even cheaper design use i.e Bc546B for Q2.

Thanks again for the suggestions.

> Possible problem: verify peak negative Ube of Q2.  I fear that for higher input voltages it will be over spec (which usually is 5V for transistors like theese)  Problem is when it switches off, and C2 pulls it far down.

Yes I anticipated this problem, the ratio of C1:C2
capacitances determines how far C1 will discharge
during the turn-off phase, and this is set at 3.0v
so the base is exposed to (3.0v - 5.0v(regulated))
= -2v
This should be within spec for the circuits I showed.

With larger power circuits I suggest trying a resistor
in series with C2 to limit dump current.


> Side note: all paragraphs on i.e http://www.romanblack.com/a02.htm gets line wrapped in browser Opera, making a mess of ASCII art and test data. (NO paragraphs

Thanks and sorry about that. I chose to use the text
circuit as it is easy to edit. I don't think the line
width can be reduced. Can you change Opera to allow
longer line widths?
-Roman

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


2002\09\23@061242 by Roman Black

flavicon
face
Morgan Olsson wrote:
{Quote hidden}

The main inductor has THOUSANDS of turns, and is swinging
from 0v to 14v, to get 3v out of feedback winding will
need another few hundred turns on top. Not me. :o)

> No current pumping through C2 and dumped in Zener, so R1 can be increased (doubled?) without decreasing Q2 base drive, = lower loss.

I take your point of using feedback to Q2 base with
the new coil but this is only about 1/3 of the picture.
The other 2/3 efficiency I gained was from the DISCHARGE
of C1 giving the monostable delay effect and slowing freq
by 2 or 3 times. Your circuit won't do that as the discharge
current must go to the inductor, not into the base of Q2. :o)

-Roman

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


2002\09\23@090350 by Morgan Olsson

picon face
Roman Black wrote:
>Morgan Olsson wrote:

>Yes, apart from you need 2 emitter resistors (not base)
>as base resistors will have the wrong effect. :o)

No, as long as you calculate the base drive to worst case hFE there is no problem, as this is switching -not linear- operation.

Emitter resistors will loose more power / total efficiency without gaining anything in this case.

Btw, paralleling mosfets is generally no problam as their ON-resistance increase for higher temp, balancing it more equal.  But as ON-resistance is not identical between two of the same kind, you can´t double the load for double mosfets.

And, you can in most cases eliminate Rc if you use Mosfet, as they usually can take up to 20 or 30 V gate drive.

Regarding display in Opera:
It seems Opera thinks you told it to wrap, or don´t understand your type of manual newlines.  (it autowraps all paragraphs on that page) I know too little to evaluate if it is your page, or Opera that don´t comply to W3C.
Might be syntactic; Opera is a little picky on syntax while other browsers guess what the user meant.  Try validating at i.e http://validator.w3.org/

/Morgan

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


2002\09\23@101028 by Morgan Olsson

picon face
Roman Black wrote:
>The main inductor has THOUSANDS of turns,

NOPE... i measured now, a cheap tiny 470uH inductor, Farnell artno 432-120.
Feeding it 10V p-p 300kHz (used high freq because i wanted to eliminate any Rloss errors) And on one turn it give me 100mV p-p.

This make a turn ratio of 1:10, so for a converter from 12V, ten turn will give 1.2 V which i think is more than enough.  Five turn might suffice.

And this was without taking off the original shrink tubing.  Removing it will probably make slightly better magnetic coupling, plus it is easier to make the wire stick around with the other wire on the ferrite bobbin.

>I take your point of using feedback to Q2 base with
etc..

L1b will make use of C1 as "tank".  Using the amplitude L1b delivers to vary C1 voltage by the same amplitude.  But.. oops, clumsy me... current design will pump peak currents through Q2 base... remedy is to add a series base resistor... or series emitter resistor..... maybe put series emitter resistor on Q2, and replace Rc with wire?  Note resistors must be low enough ohm to make Q2 fully turn on/off, but not too low as that will make excessive peak base current.

good luck experimenting

/Morgan

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


2002\09\23@115749 by lexandre_Guimar=E3es?=

flavicon
face
Hi, Roman

> Hmm, that's an easy one, 40v in, 12v out, 500mA, overvoltage
> output. The short circuit protection I can do to, but it will
> cost a little efficiency. Regulation should be excellent.
>
> So is max efficiency from 30 to 300ma ok?
> With current limiting kicking in at 500mA?

   Would be almost perfect ! I was not clear about the 12V out, I would
like to have both, a 5V well regulated output and a 12V poor regulated one.
One would power the board and the other one would power the sensors. I think
I could just drop another coil and use a zener for the 12V out. Not very
efficient but maybe enough for many applications.

   Another great thing about your circuit is that it is makes much less
noise than other SMPS i have seen. Not problems with radios nearby.

   I have one of the old versions running on my desktop for over 2 months
and it is still running perfectly.

   Did I ever sent you the one I made with the MPSA42 that can run from 150
volts ?? It also performs beautifully !

   The circuit is simple, clean, efficient and cheap !! What else could
anyone ask for ??? Congratulations.

   My only concerns about using it in a product is the lack of overvoltage
and overcurrent protection. It would not be nice to have 100 bucks going
away because the switching transistor got short circuited :-(

Best regards,
Alexandre Guimaraes

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


2002\09\23@153440 by Philip Pemberton

face picon face
Alexandre Guimarães wrote:
>     My only concerns about using it in a product is the lack of
> overvoltage and overcurrent protection. It would not be nice to have
> 100 bucks going away because the switching transistor got short
> circuited :-(
What about an SCR + Zener crowbar? Rig one SCR to short out L1, rig the
other to short Q1 base to Vin. Remind me to try that out later :-)

Later,
--
Phil.
RemoveMEphilpemTakeThisOuTspamspamdsl.pipex.com
http://www.philpem.dsl.pipex.com/

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


2002\09\23@162459 by lexandre_Guimar=E3es?=

flavicon
face
Hi,

> What about an SCR + Zener crowbar? Rig one SCR to short out L1, rig the
> other to short Q1 base to Vin. Remind me to try that out later :-)

   I really do not like crowbar protection ! I have seem it fail many, many
times. It should work but in practice I have seem it fail much more than the
power supply itself. If you make it with a fast response it will false
trigger on any small transients and if it is too slow things will burn
before it activates ! I would prefer just a big zener and a polyfuse.
Knowing Roman for a while I always expect that he can come up with another
pearl of simplicity and make much better than the zener and polyfuse :-)

Best regards,
Alexandre Guimaraes

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


2002\09\23@220545 by Roman Black

flavicon
face
Alexandre Guimarães wrote:

> > So is max efficiency from 30 to 300ma ok?
> > With current limiting kicking in at 500mA?
>
>     Would be almost perfect ! I was not clear about the 12V out, I would
> like to have both, a 5V well regulated output and a 12V poor regulated one.
> One would power the board and the other one would power the sensors. I think
> I could just drop another coil and use a zener for the 12V out. Not very
> efficient but maybe enough for many applications.

I've got the 12v version tested. More below.


>     Another great thing about your circuit is that it is makes much less
> noise than other SMPS i have seen. Not problems with radios nearby.

Thanks, but this new version will be slightly
noisier now it's switching quicker. :o)

>     Did I ever sent you the one I made with the MPSA42 that can run from 150
> volts ?? It also performs beautifully !

I would like to see the circuit. :o)


>     My only concerns about using it in a product is the lack of overvoltage
> and overcurrent protection. It would not be nice to have 100 bucks going
> away because the switching transistor got short circuited :-(


I have finished the current limiting and short circuit
protection now, it only cost one transistor and 2 resistors.

Here is the new 26v to 12v converter, 700mA, 92% efficient:
http://www.romanblack.com/a03.htm

Here is the new 26v to 12v converter, 450mA, 91% efficient
AND WITH CURRENT LIMITING!
http://www.romanblack.com/a04.htm

The 3-tran circuit will handle continuous short
circuits and all overload conditions.

Re your over voltage question and possibility of a
shorted switching transistor, I would just use a small
fusible resistor 0.33 ohms etc in series with the
input voltage. This will save your $100 equipment! :o)
-Roman

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


2002\09\23@222933 by hard Prosser

flavicon
face
Magic !

Richard P




I have finished the current limiting and short circuit
protection now, it only cost one transistor and 2 resistors.

Here is the new 26v to 12v converter, 700mA, 92% efficient:
http://www.romanblack.com/a03.htm

Here is the new 26v to 12v converter, 450mA, 91% efficient
AND WITH CURRENT LIMITING!
http://www.romanblack.com/a04.htm

The 3-tran circuit will handle continuous short
circuits and all overload conditions.

Re your over voltage question and possibility of a
shorted switching transistor, I would just use a small
fusible resistor 0.33 ohms etc in series with the
input voltage. This will save your $100 equipment! :o)
-Roman

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

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


2002\09\23@231843 by lexandre_Guimar=E3es?=

flavicon
face
part 1 1826 bytes content-type:text/plain; (decoded 7bit)

Hi, Roman

> >     Did I ever sent you the one I made with the MPSA42 that can run from
150
> > volts ?? It also performs beautifully !
>
> I would like to see the circuit. :o)

   It is a pleasure :-) It works nicely and goes all the way from 160 volts
downto 12 volts. I tested it for some weeks connected directly to the mains
trough a simple rectifier and filter and another one connected to a car
battery for the same period. Efficiency is very bad but reliability should
be quite nice. Even with all the series resistors I still get 150ma with no
heat at all in all test conditions.

   The ugly and inefficient circuit is attached. :-) It runs nicely, I can
assure that. I tested it in all conditions I could simulate and regulation
and efficiency is poor but it makes very little noise and should be quite
reliable. There should still be a prototype running inside a train for some
months with it and nobody called to say it stoped :-)

> The 3-tran circuit will handle continuous short
> circuits and all overload conditions.

   Thanks a lot.. I will surely try it out as I did with the others. I will
surely find a nice product to put it in someday ;-) When that happens you
can be sure I will get you a nice bottle of Champagne :-) What brand would
you prefer ? Does that satisfies the "hippyware" aggrement ? ;-)

> Re your over voltage question and possibility of a
> shorted switching transistor, I would just use a small
> fusible resistor 0.33 ohms etc in series with the
> input voltage. This will save your $100 equipment! :o)
> -Roman

   You mean with a zener or something else at the output to clamp the high
voltage and make the fusistor open, right ? Am I missing something ?

Best regards,
Alexandre Guimaraes




part 2 5871 bytes content-type:image/gif; (decode)


part 3 105 bytes
--
http://www.piclist.com hint: To leave the PICList
EraseMEpiclist-unsubscribe-requestspamEraseMEmitvma.mit.edu


2002\09\24@084853 by Russell McMahon

face
flavicon
face
>  My only concerns about using it in a product is the lack of overvoltage
> and overcurrent protection. It would not be nice to have 100 bucks going
> away because the switching transistor got short circuited :-(

I have my 3 transistor version in service with unknown thousands in use so
far. I have had a very few failures reported to me although how many I don't
know because the system they use has no mechanism for reporting this back -
this is not under my control at all and not how I would like it to be.
(Designed in NZ, made in Taiwan, sold in Europe :-) ).

Because I didn't want a failed pass transistor etc to take out the low
voltage electronics I specd a fused output with a 12v zener behind the fuse.
(1w glass bead body DOxx Philips can't remember the number and too lazy to
go and look it up). Zeners almost always (essentially always) fail short and
the leads on these are like mini crowbars compared to PCB tracks etc. All of
the (few) failures I am aware of have shorted the zener and blown the fuse
without hurting the low voltage electronics. Arguably a reasonably cheap way
of protecting against failure but may cost almost as much as Roman's whole
converter :-)


       RM

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


2002\09\24@090550 by lexandre_Guimar=E3es?=

flavicon
face
Hi, Russel


> >  My only concerns about using it in a product is the lack of overvoltage
> > and overcurrent protection. It would not be nice to have 100 bucks going
> > away because the switching transistor got short circuited :-(
>
> Because I didn't want a failed pass transistor etc to take out the low
> voltage electronics I specd a fused output with a 12v zener behind the
fuse.
> (1w glass bead body DOxx Philips can't remember the number and too lazy to
> go and look it up). Zeners almost always (essentially always) fail short
and
> the leads on these are like mini crowbars compared to PCB tracks etc. All
of
> the (few) failures I am aware of have shorted the zener and blown the fuse
> without hurting the low voltage electronics. Arguably a reasonably cheap
way
> of protecting against failure but may cost almost as much as Roman's whole
> converter :-)

   I was thinking about the zener and a polyfuse before it. Should give
good protection but would really cost almost the same as the full power
supply :-( I just asked about it because, as we all know, Roman can bring
some very innovative solutions to some classic problems. We never know.. I
will test it in the next few weeks with the zener and fuse and make sure
that I hit the switching transistor hard enough to make it fail and see how
the zener and fuse acts. Should be quite reliable. I will let the list know
how the testing goes.

Best regards,
Alexandre Guimaraes

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


2002\09\24@170459 by D. Schouten

flavicon
face
I have implemented overvoltage protection with an output fuse followed by a
parallel zener too. Works perfect! I have bypassed the pass transistor (not
from Roman's circuit but from an IC solution) and applied 180VDC at once on
the input. The zeners (two 10V/5W types in series) stay healty but the
200mAT fuse blows at once.

My DC-DC circuit output's 15VDC but needs a minimum load current of 20mA to
stay in continuous mode, or else the output voltage will rise almost
linearly with Vin. So I thought, instead of using a 'dump' resistor load and
trowing away efficiency, why not use the 20V zener clamp on the output to
'regulate' at 20V when there is no load applied to the converter? This
worked ok, so now my overvoltage circuit solves the minimum load requirement
too :-)

Daniel...

{Original Message removed}

2002\09\25@153723 by Roman Black

flavicon
face
D. Schouten wrote:
>
> I have implemented overvoltage protection with an output fuse followed by a
> parallel zener too. Works perfect! I have bypassed the pass transistor (not
> from Roman's circuit but from an IC solution) and applied 180VDC at once on
> the input. The zeners (two 10V/5W types in series) stay healty but the
> 200mAT fuse blows at once.

I'm curious, why fuse the output and not the input?
Also, they are two big and expensive Zeners! Couldn't
you use a smaller zener clamped with a larger capacitor?
It's not as though you need 10W of zeners to handle 4W
of surge for a few mS.

You can even reduce the Zener surge current even more by
adding a resistor between the cap and zener, like this,
(if your load can tolerate a slight overvoltage)

psu out ----*-------*--
           |       |
           |       R
           |       |
     big   C       |
     cap   |       |
           |       Zener
           |       |
           |       |
gnd --------*-------*--

So the zener is normally held a volt or so below
conduction, and on surge fault the cap has to charge an
entire volt before the zener dissipates any power,
even then the cap must charge another volt before the
zener dissipates maximum (4W).

On surge fault the fuse will blow well before the zener
dissipates maximum due to the large inrush currents
into the cap during charging those couple of volts. :o)

> My DC-DC circuit output's 15VDC but needs a minimum load current of 20mA to
> stay in continuous mode, or else the output voltage will rise almost
> linearly with Vin. So I thought, instead of using a 'dump' resistor load and
> trowing away efficiency, why not use the 20V zener clamp on the output to
> 'regulate' at 20V when there is no load applied to the converter? This
> worked ok, so now my overvoltage circuit solves the minimum load requirement
> too :-)

That's a smart way of doing it. :o) And like I said
above, as you have 5 volts or so between normal running
and the fault turning the zener on, i'm sure you can
reduce the zener size a LOT.
-Roman

--
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\09\25@170857 by Brent Brown

picon face
On 26 Sep 2002 at 5:32, Roman Black wrote:
> D. Schouten wrote:
> >
> > I have implemented overvoltage protection with an output fuse
> > followed by a parallel zener too. Works perfect! I have bypassed the
> > pass transistor (not from Roman's circuit but from an IC solution)
> > and applied 180VDC at once on the input. The zeners (two 10V/5W
> > types in series) stay healty but the 200mAT fuse blows at once.
>
> I'm curious, why fuse the output and not the input?
> Also, they are two big and expensive Zeners! Couldn't
> you use a smaller zener clamped with a larger capacitor?
> It's not as though you need 10W of zeners to handle 4W
> of surge for a few mS.
>

Just to add my 2 cents worth... I think a TVS would be way better
than a zener. A TVS or a VDR have advantages over zeners for this
sort of thing. Correct me if I'm wrong, but in general:

A TVS (transient voltage suppressor, goes by other names too) is
faster than a zener, can absorb way more energy, not so sharp voltage
clamping. You can get them in AC & DC  types.

A VDR (voltage dependent resistor, goes by other names too) is I
think slower than a zener, can absorb more energy than a TVS and is
coarser voltage clamping than a TVS. These are AC.

PS. These circuits are really interesting, keep up the good work!

Regards, Brent.

--
Brent Brown, Electronic Design Solutions
16 English Street, Hamilton, New Zealand
Ph/fax: +64 7 849 0069
Mobile/txt: 025 334 069
eMail:  @spam@brent.brown@spam@spamspam_OUTclear.net.nz

--
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\09\25@175550 by D. Schouten

flavicon
face
Hi Roman,

> I'm curious, why fuse the output and not the input?
> Also, they are two big and expensive Zeners! Couldn't
> you use a smaller zener clamped with a larger capacitor?
> It's not as though you need 10W of zeners to handle 4W
> of surge for a few mS.

I have fused the output to protect the circuit from applying an external
power source to the output. Since this DC-DC converter is a stand-alone
unit, it might be wired visa-versa. Applying an external DC source with
correct polarity doesn't harm the unit because of a blocking diode. But when
the polarity is reversed, the output fuse would blow.
But indeed, when protecting for external power sources on the output is no
issue, I could move the fuse to the input.

The reason why I use these two 5W zeners in series can indeed not be fully
justified. I had these 10V parts lying around and needed 20V. I am sure just
one 5W zener could do the job too. However, since I am using a slow blow
200mA fuse (fast types cannot be used due to relatively high capacitive
loads), even these two 5W series zeners get pretty hot when shorting out
180VDC.

I know TVS's are able to absorb more energy than a simple zener. But these
parts are twice as expensive than my zener too.


> You can even reduce the Zener surge current even more by
> adding a resistor between the cap and zener, like this,
> (if your load can tolerate a slight overvoltage)

Yes, I have thought of that too. But that will cost me an extra resistor. Of
course the zener could be smaller too then, however the smaller the zener
the larger the series resistor value has to be. And that will only increase
the overvoltage level. But perhaps, some tweaking might be interesting since
I can handle some overvoltage (larger than 20V). :-)

Daniel...

--
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\09\25@180750 by Roman Black

flavicon
face
Morgan Olsson wrote:
>
> Regarding efficiency and other spec i use to be very careful.

> while also have margins for component variations such as min/max hFE
> of transistors, ESR of caps, etc. Maybe some combinations cause stability
>problems.
> Also remember zeners have typically 5% tolerance, add voltage change due to
> current change due to input voltage change, regulation of input volt, and load,
> and i guess simple regulators like theese have problems qualifying in a lot of areas.

Good points. Do you think I was hasty making that
"replaces a 7805" claim? Basically the web page is
written from my perspective, with the enthusiasm
thrown in. :o) It is what it is, I spent some hours
fiddling with the circuits and then put up the results
obtained. The efficiency tests were done just to see
that it IS possible, and improvements or de-provements
are up to the user.

Any serious manufacturer using the circuit would be
sure to spice it and/or lab test it with their chosen
parts. And for hobby/PIC use etc the circuits I gave
should work pretty good off the shelf.

> Using some combinations of worst case values is sometimes very tough
> for a cirquit,

Any suggestions to improving this? :o)

> It takes much calculations and verifying labs to make a design that
> performes good in all conditions High/low temp, High/low input volt etc,
> while also have margins for component variations such as min/max hFE of
> transistors, ESR of caps, etc.

How much work did you expect me to do? <grin>

>  Also note that the relatively bigger inductors in the Black is
> making slower load pulse reponse and also cost more in component cost
> and board estate if talking production.

Not totally true, in the smaller size "RF choke" type
parts they are usually the same price so comparing a
50uH or 100uH choke to the larger 470uH show the same
price. The cost is in increased losses in the choke,
not $$. As long as overall efficiency is still ok for
the task at hand. There should be substantial $$ saved
from not having to buy a dedicated switcher chip.


> Btw, new cadidates for the switch: BD438 (high gain high current
> pretty low cost TO126) and also check out Zetex "Matrix" range ZTXnnn
> IIRC for low voltage drop and very high gain.

Thanks, I will check them out. :o)
-Roman

--
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\09\30@131056 by 4HAZ

flavicon
face
Hey Roman hows the high voltage version coming?
I am in need of a smps that can take 200 - 300 DC volts input and deliver
~13.6Vdc out at 10 amps

Lonnie - KF4HAZ -

--
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\09\30@183902 by Peter L. Peres

picon face
On Mon, 30 Sep 2002, FalconWireless Tech Support - KF4HAZ wrote:

*>Hey Roman hows the high voltage version coming?
*>I am in need of a smps that can take 200 - 300 DC volts input and deliver
*>~13.6Vdc out at 10 amps

Can it use more than 3 BC639 transistors ?

Peter

--
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\09\30@195032 by Russell McMahon

face
flavicon
face
> Hey Roman hows the high voltage version coming?
> I am in need of a smps that can take 200 - 300 DC volts input and deliver
> ~13.6Vdc out at 10 amps

I'd think really hard about using something else for a design of that
voltage and power level. I suspect Roman may agree. While Roman's design
appears superb it is still relatively untried and may yet have subtle
"gotchas"  (although it's hard to see what they would be)(because, if they
exist,  they are subtle :-) ).  Also at those power levels well proven
current limit with fold back and bullet proof operation under all
conceivable transient conditions is required. My 3 transistor design
operates at up to 200v in and 600 mA out in practice but I wouldn't
recommend it for the above application without considerable investigation.

Amongst other things, at high voltages the energy levels become non trivial
in the components linking the upper and lower sections of the circuit. As
the current requirement rises the drive requirement of the pass transistor
also rises again increasing dissipation. - especially the power required in
the resistor from lower transistor collector to upper transistor base. This
resistor's average (as opposed to peak) dissipation tends to stay constant
with increasing voltage because of the changing mark space ratio as voltage
is increased BUT beware what happens under eg short circuit conditions.
Depending on its design a current limit may protect the output but expose
this resistor to full dissipation with "interesting" results.

Even the zener resistor can become non trivial with increasing voltage as it
has to be "stiff enough" to provide reference voltage at lowest input
voltage and not die at highest input voltage. Addition of a current source
to drive the zener may be indicated. A useful alternative is to drive the
zener from both the input (for startup) and from the output (most of its
current once started). In this design this would require "tapping down" the
output voltage resistively to feed the emitter of the control (lower)
transistor so that the zener has some headroom. Nasty complexity for such a
beautifully simple circuit.

Use of a P FET* as the pass element can be very worthwhile here as it
greatly reduce drive power requirements - BUT may then need extra bipolar
drivers depending on switching speed and desired efficiency. I ended up
adding several bipolars to the basic design to improve performance over the
widest possible range of input voltages. This may not be needed if you can
optimise for a limited input range. A few extra small transistors cost very
little but you may wish to avoid them if the "purity" (and fun value!) of
the 2 transistor design is valued.

* 300 volt PFETS are about as rare as hen's teeth. The advantage of a buck
design is that you can operate the pass element almost at its rated voltage.
Especial care with switching transients is then indicated although the buck
regulator is very forgiving in this respect. .


       Russell McMahon

--
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



'[EE]: 2-tran SMPS 5v reg now 90% efficicient!'
2002\10\01@025009 by Mike Singer
picon face
part 1 1269 bytes content-type:text/plain; (decoded 7bit)

Russell McMahon wrote:
> > Hey Roman hows the high voltage version coming?
> > I am in need of a smps that can take 200 - 300 DC
> > volts input and deliver ~13.6Vdc out at 10 amps
>
> I'd think really hard about using something else for
> a design of that voltage and power level. I suspect
> Roman may agree. While Roman's design appears superb...

Russell McMahon wrote on 2002-07-17:
Subject: [EE]: ST Electronics selector guide
Newsgroups: comp.arch.embedded.piclist

> Allegedly "dedicated devices"
> Wide range of interesting components.
>       http://us.st.com/stonline/products/selector/

ST VIPer100A:

VIPer100]/100A, made using VIPower M0 Technology,
combines on the same silicon chip a state-of-the-art
PWM circuit together with an optimized high voltage
avalanche rugged Vertical Power MOSFET (620V or 700V / 3A).
Typical applications cover off line power supplies
with a secondary power capability of 50 W in wide
range condition and 100W in single range or with
doubler configuration. It is compatible from both
primary or secondary regulation loop despite using
around 50% less components when compared with a
discrete solution.



+++ .gif is attached +++

Mike.



part 2 10628 bytes content-type:image/gif; (decode)


part 3 131 bytes
--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.


2002\10\01@040431 by Alan B. Pearce

face picon face
>> Hey Roman hows the high voltage version coming?
>> I am in need of a smps that can take 200 - 300 DC volts input and deliver
>> ~13.6Vdc out at 10 amps
>
>I'd think really hard about using something else for a design of that
>voltage and power level. I suspect Roman may agree.

Well I certainly agree. With that voltage ratio I suspect you are really
into a transformer than a single inductor, especially with that load current
requirement.

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


2002\10\01@084516 by Roman Black

flavicon
face
Russell McMahon wrote:
>
> > Hey Roman hows the high voltage version coming?
> > I am in need of a smps that can take 200 - 300 DC volts input and deliver
> > ~13.6Vdc out at 10 amps
>
> I'd think really hard about using something else for a design of that
> voltage and power level. I suspect Roman may agree.

You are right! :o)


> While Roman's design
> appears superb it is still relatively untried and may yet have subtle
> "gotchas"

Some things are quite safe, like it can't feed any
current to the output UNLESS output is below the
set voltage. BUT, if the output is fed from the load
(like a decelerating motor) the output voltage will
rise accordingly. However this will also happen with
many other series regulators etc.

One "gotcha" is that to be efficient it is tuned for
an input voltage "range". If the input voltage gets
too far below that range it may go linear, and increase
dissipation of the switching transistor. This needs
to be taken into account. :o)
-Roman

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


2002\10\01@084840 by Roman Black

flavicon
face
Mike Singer wrote:

> Russell McMahon wrote on 2002-07-17:
> Subject: [EE]: ST Electronics selector guide

> >       http://us.st.com/stonline/products/selector/
>
> ST VIPer100A:
>
> VIPer100]/100A, made using VIPower M0 Technology,
> combines on the same silicon chip a state-of-the-art
> PWM circuit together with an optimized high voltage
> avalanche rugged Vertical Power MOSFET (620V or 700V / 3A).
> Typical applications cover off line power supplies
> with a secondary power capability of 50 W in wide
> range condition and 100W in single range or with
> doubler configuration.


Price?? :o)
-Roman

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


2002\10\01@085435 by Roman Black

flavicon
face
FalconWireless Tech Support - KF4HAZ wrote:
>
> Hey Roman hows the high voltage version coming?
> I am in need of a smps that can take 200 - 300 DC volts input and deliver
> ~13.6Vdc out at 10 amps
>
> Lonnie - KF4HAZ -


Ha ha! :o) Hi Lonnie, so you want me to convert my
tiny cheap 2-transistor regulator to 300v input and
total 140 watts! Wow!

There are *many* issues, probably the worst being that
13.6v @ 10A sounds like the needs of ham radio equipment,
and you will require ISOLATED output that is NOT at
mains potential!! My circuit is totally UNSAFE for
mains "off line" use, as it has no isolation of input
and output. This is illegal for any PSU that may be
touched by the end user and run from the AC mains.

As far as technical issues go, both Russell and Alexandre
have built high voltage 2-tran circuits and tackled many
of the tech issues... One being that the buck diode must
be rated for high voltage, and hence will be inefficient.
Then the buck inefficiencies from the very short pulse
duration 340vdc -> 13.6vdc (about a 4% width pulse), then
the problems with transistors communicating over a 340v
drop, etc etc.

You need an "off line" SMPS converter, with a transformer
so it can use low voltage schottky output diodes, and
MUCH better feedback with optocoupler etc.
These are common and cheap now, i suggest trying;

http://www.jaycar.com.au

who have item MP-3110 which is a 240vac -> 12vdc 150W
SMPS supply for $109 AUD (about $60 USD). They are on
special this month.

You can change the 12vdc output to 13.6vdc output by
changing one zener diode next to the optocoupler.
:o)
-Roman

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


2002\10\01@103555 by Mike Singer

picon face
Roman Black wrote:
>
> ST VIPer100A:

Price?? :o)
-Roman

Here in Ukraine (large quantities):

VIPER-100 STMicroelectronics    2,25
VIPER-20 STMicroelectronics    1,12

(Have no idea about transformer price.)

> > Hey Roman hows the high voltage version coming?
> > I am in need of a smps that can take 200 - 300 DC volts
> > input and deliver ~13.6Vdc out at 10 amps

The kind of question, asked twice a month for years.
Answer: _ PC PSU _  ~ $10.(Case+fan+HV capacitors+
+power transistors+transformer+...)

Mike.

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


2002\10\01@130056 by 4HAZ

flavicon
face
> Hey Roman hows the high voltage version coming?
> I am in need of a smps that can take 200 - 300 DC volts input and deliver
> ~13.6Vdc out at 10 amps

Just got a quote from the only place I could find that offered anything that
was designed to take "dirty" unfiltered unregulated high voltage DC and
deliver clean regulated 13.x Vdc. @ 10 Amps.

$3300.00
Guess I can afford to let the smoke out of a few more components trying to
design something myself.

Lonnie - KF4HAZ -

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


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