Negative voltage from Transformerless Capacitive power supply

ge

..

=A0| =A0 =A0 =A0|

=A0k =A0 =A0 =A0|

=A0 =A0 =A0c

=A0| =A0 =A0 =A0|

=A0| =A0 =A0 =A0|

=A0| =A0 =A0 =A0|

=A0k =A0 =A0 =A0|

=A0 =A0 =A0c

=A0| =A0 =A0 =A0|

=A0| =A0 =A0 =A0|

s e

Hey, now there's a good idea! Two C's but only one R. That should work.

George H.

...

That looks just like John's circuit. But it has lotsa ripple. At least the one I built. I could have made a mistake.

George H.

m...

=A0| =A0 =A0 =A0|

=A0k =A0 =A0 =A0|

a =A0 =A0 =A0c

=A0| =A0 =A0 =A0|

=A0| =A0 =A0 =A0|

=A0| =A0 =A0 =A0|

=A0k =A0 =A0 =A0|

a =A0 =A0 =A0c

=A0| =A0 =A0 =A0|

=A0| =A0 =A0 =A0|

OK, Maybe I made a mistake in my wiring. (It wouldn't be the first time.) I'll try again.

George H.

I posted the circuit above.

John

What? The one under your comb-over?

Listen DarkSplatter, I'm not going to put up with your angst-ridden verbal abuse the way your bewildered wife-swapping mom and pop do. If you're suffering from terminal gamete backup, I would prescribe cranking up the catheter's suction until it begins to yield shrapnel from your skateboard-accident-cracked parietal lobe, samples of which might prove useful later on in studies of the Hollow Skull Theory.

HTH HAND.

com...

=A0 =A0| =A0 =A0 =A0|

=A0 =A0k =A0 =A0 =A0|

=A0a =A0 =A0 =A0c

=A0 =A0| =A0 =A0 =A0|

=A0 =A0| =A0 =A0 =A0|

=A0 =A0| =A0 =A0 =A0|

=A0 =A0k =A0 =A0 =A0|

=A0a =A0 =A0 =A0c

=A0 =A0| =A0 =A0 =A0|

=A0 =A0| =A0 =A0 =A0|

ve

OK played around a bit this morning. A few observations; I put no load on the circuit the first time. So with a load all circuit configurations have more ripple. (Simple RC time constant I assume). Second the first time I built this I used a 1N5229 (4.3V) zener. (Just what I happened to grab from the box) This showed a lot of ripple in the circuit that had the rectifying diode before the zener. (Even with no load) But this went down dramatically when I used a 1N4733 (5.1V) zener. John L. said something about this earlier in the thread,

=93There would be more ripple only if the zener curve isn't flat, which is not an issue for zeners above 5 volts or so. Use bandgaps for flatter curves.=94

That=92s nice I can learn something about zeners.

Any good Zener references?

George H.

We were just playing with some zeners yesterday.

We need a low-power, minimum-parts 5.6 volt reference that's floating on the output of a power opamp. I suggested this:

+40----------------+ | e pnp b-------+ c | | |

but got an unenthusiastic reaction. Well, it does have two small gotchas.

John

What do you have where the asterisks are? In an earlier diagram, they were "c"s (for "capacitor").

Hope This Helps! Rich

One that jumps out at me is the lack of a b-e resistor on the PNP, which is asking for thermal runaway. (or so I was taught in techie school, or maybe it was in a hobbyist kit manual.)

Do I win a prize? ;-)

Thanks, Rich

Zener current depends on beta then, as r sets the base current pretty well (say 39.2/R) but that's all it does. Different parts will yield varying zener currents that vary also over temp (and Vce), as well. I think a minimum beta is often spec'd but it can vary by manufacturer and part. Depending too highly upon it is a problem, isn't it? And I'm not sure what temp ranges you wanted, but a wide enough range could be a problem? Wouldn't this be an application for a current mirror?

Jon

Beta isn't a problem. We stock BCX71J's, which have a guaranteed beta spread of about 2:1, and beta doesn't vary a huge amount over temperature. A bigger problem is that R would have to be something like 200M, which is expensive and hard to get. The other problem is that zeners sometimes oscillate at low currents, so we should bypass it or use a bandgap, if we can hit our voltage target with one.

So the current source will be a Supertex depletion fet with a resistor in its source (and maybe more current spread than the beta-limited transistor) and a bypass across the zener.

Darn, I told the guy doing this that, just once in his career, he should break the sacred rules and do a beta-biased stage.

John

Hmm, the circuit looked OK to me. But them I'm a Zener novice. If you can run enough current through the zener the transistor beta shouldn't matter too much. A current mirror is another transitor. You can make a current source with a three terminal voltage regulator and a resistor. George H.

John, why not just use a cheap current mirror part? They come in SOT23 (I've got a bunch) and they cost very little. Cripes, a 2nd BJT costs very little. Okay. I remember. Every part is 25 cents min, at least, counting all the ancillary costs. But still...

Just curious.

Jon

I'd need P and N polarities and 80 volt compliance. Are there any parts like that?

It's a wash on parts count between the depletion fet (which we have in stock for about 35 cents) and the two new mirrors we'd have to buy and stock. At least the resistor value would be sensible, a few hundred K maybe.

Whenever we add a part to the PADS library and to stock, I have to fill out a form. I hate to fill out forms.

John

John

[snip]

Crap design at Larkin's finest ;-)

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
Gourmet Puzzles:

        What part of the fish are the "sticks"?

        Likewise where are the chicken "fingers" located?

John Larkin a écrit :

Yup. Just ask...

You have :

PNP/150V

NPN/160V

All available at digikey, mouser, arrow.

Now, those are all for Joerg :-) They also have some other nice devices (BJT+MOS, all combinations) in interesting packages.

That's why I don't build stuff. I just design it and I never fill out forms.

--
Thanks,
Fred.

Those don't appear to be monolithic duals, just two transistors from the same wafer. The datasheet doesn't specify whether they are electrically and thermally isolated, which they probably are. So it will be a really bad current mirror without added external resistors... some possibility of thermal runaway, especially at my voltages.

Digikey shows no stock, "call" for price.

Maybe I should stick with the Supertex parts.

John

Can small binaries be posted here?

John

Not normally.

Some ISP's block ALL binaries.

Use A.B.S.E (alt.binaries.schematics.electronic) for the binary, then post a link to it here.

Or simply post on your website, then a link here.

...Jim Thompson

-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

| 1962 | Gourmet Puzzles:

What part of the fish are the "sticks"?

Likewise where are the chicken "fingers" located?