Covert emitter follower to PNP

[Jerry Norris]

I have decided on this design and tweaked the rails to 9V and 3V,
which still provides a nice 5V swing between 3V and 8V. Here is the
Spice sim.

http://www.mediafire.com/i/?fonqyjahu6wp13g

To begin my project, I also need a PNP version that will run form -9V
and -3V rails. IOW swing between -8V and -3V.

I have played around with the components in LTSpice, but without much
luck. There is nothing I can find on the net to work from.

Can anyone please set me on the right track regarding what changes
need to be made?

Jerry Norris

 

[Jerry Norris]

You are only referencing ground at your input supply, meaning that
you've made your self a great, big, fancy, expensive, nonlinear
resistor, probably with DC offset.

Where was ground _supposed_ to be in that circuit?

Thank you for your reply. The output ground is in common with the
input ground.

The power supplies may or may not be referenced to ground. It seems to
work either way. Yes, the 3V DC offset is a desired result.

As stated in my OP I would like to produce a second matching circuit
uisng PNP's where the ouptut swings between neg 3 and neg 8V, as
opposed to pos 8V and pos 3V.

I have tried a few things to no avail, and can't find anything similar
on the net. Any expert advice on this would be much appreciated.

Here is the link again.

http://www.mediafire.com/i/?fonqyjahu6wp13g

Jerry Norris

 

[petrus bitbyter]

Thank you for your reply. The output ground is in common with the
input ground.

The power supplies may or may not be referenced to ground. It seems to
work either way. Yes, the 3V DC offset is a desired result.

As stated in my OP I would like to produce a second matching circuit
uisng PNP's where the ouptut swings between neg 3 and neg 8V, as
opposed to pos 8V and pos 3V.

I have tried a few things to no avail, and can't find anything similar
on the net. Any expert advice on this would be much appreciated.

Here is the link again.

http://www.mediafire.com/i/?fonqyjahu6wp13g

Jerry Norris

If you replace all transistors by PNP types en revers all voltage sources it
should have to work. If not, there's something wrong, probably with the
simulator. Be aware that you cannot connect both inputs to the same source.

petrus bitbyter

 

[YD]

Late at night, by candle light, [email protected] (Jerry
Norris) penned this immortal opus:

I have decided on this design and tweaked the rails to 9V and 3V,
which still provides a nice 5V swing between 3V and 8V. Here is the
Spice sim.

http://www.mediafire.com/i/?fonqyjahu6wp13g

To begin my project, I also need a PNP version that will run form -9V
and -3V rails. IOW swing between -8V and -3V.

I have played around with the components in LTSpice, but without much
luck. There is nothing I can find on the net to work from.

Can anyone please set me on the right track regarding what changes
need to be made?

Jerry Norris

Whatsit supposed to do? Those DC sources are in floating anti-series,
effectively making a single floating 6V DC source. Since they have no
reference to the signal the whole circuit seems rather pointless.

ITWHBT.

- YD.

 

[Jerry Norris]

indeed... tell us what problem youre having

Well, it's trying to work, but the output amplitude has dropped to 1
volt and there is not the desired neg 3V offset I should have from the
neg 3 and neg 9 rails.

http://www.mediafire.com/?51jv33439s847fv

Any suggestions please?

Jerry Norris

 

[Jerry Norris]

Whatsit supposed to do? Those DC sources are in floating anti-series,
effectively making a single floating 6V DC source. Since they have no
reference to the signal the whole circuit seems rather pointless.

Yes, the output stage is floating at +6V. The negative of the two
supplies could be tied to input signal ground, but I am trying to
avoid that if possible.

IOW I want to replace the input signal ground with +3V at the output,
so the swing is between +3 to +9 without a common ground reference.

Is everyone saying this is "pointless" because it can't be done, or
because it hasn't been done?

Jerry Norris

 

[Jerry Norris]

It's easy. Just do it right.

John

If I knew how to do it right, would I be here asking for help?

Jerry

 

[petrus bitbyter]

Well, it's trying to work, but the output amplitude has dropped to 1
volt and there is not the desired neg 3V offset I should have from the
neg 3 and neg 9 rails.

http://www.mediafire.com/?51jv33439s847fv

Any suggestions please?

Jerry Norris

Your simulater does not seem to function the same way in both cases. I
suppose it has to do with the "ground" of your signal source that seems to
have no relation to the rest of the circuit having the voltage sources
floating. Guess the simulater chooses a common ground itself.

petrus bitbyter

 

[Jerry Norris]

It's pointless because a power supply that's not connected to anything
can't supply power!

It seems to me there is a -6V potential between the -3 and -9V
batteries. That is the power I want to use, not something referenced
to an arbitrary ground, as in the inout signal.

Jerry Norris

 

[Jerry Norris]

Newsgroups: alt.binaries.schematics.electronic
Subject: Re: Convert emitter follower to PNP - Follower_Norris.pdf
Date: Fri, 17 Jun 2011 17:59:49 -0700
Message-ID: <[email protected]>

Sorry, but downloded all the messages on the above group and could not
find a reference to this.

Please advise how to view it.

Many thanks,

Jerry Norris

 

[Jerry Norris]

V1 and V2 have no ground reference.

Yes, Jim that is the intent of the project. Please allow me to
explain.

I have a 20Hz sinewave from a function generator. It obviously has its
own ground reference, ie. 0-2.5Vpp.

I would like to convert this to two identical signals. One swinging 5V
(or so) between the +3 and +9 battery potential. The other likewise
swinging 5V, but between -3 and -9 batteries.

IOW the two output signals are to appear soley between the two
dissimilar plus battery terminals, and two dissimilar minus battery
terminals respectively ... and not with any reference to an "offset"
or common ground.

I assume, to see it, the CRO probe would be placed across the 1K load
resistor shown in my diagram.

http://www.mediafire.com/i/?fonqyjahu6wp13g

I realize this is not conventional practice, but the nature of my
project is a physics experiment.

Any way the above can be achieved, discrete or op amp, would be fine.

Jerry Norris

 

[Jerry Norris]

Double click on the message ID. If you can't get it, I'll put it up
on my website. Let me know.

I use Free Agent newsreader. It doesn't appear to support this
function. I will check the binary group tomorrow. Maybe my server is
slow.

Sorry for the trouble.

Jerry

 

[Jerry Norris]

Well, then just connect a single power supply between the input and
the output. Done.

I would then need to build two dedicated signal generators, one that
runs off +3 and +9, and the other from -3 and -9.

For example two XR2206's, which will run on split supply.

However, the output signals of the two emitter followers (npn and pnp
versions), shown in my posted circuits, need to be identical, except
for the differing potential.

Maybe you have something in mind, but I don't see how this can be
easily achieved without using one input signal for both.

Jerry Norris

 

[Jerry Norris]

The better way, use OpAmps...

http://analog-innovations.com/SED/Follower_Norris_OpAmps.pdf

Thanks for the schematics. I will breadboard them today.

Can they be separated into two circuits, one running off +9V and the
other -9V, so I can have the option of applying an inverted signal to
one?

I can see the positively and negatively shifted waveforms on your
sims, but where is the return for each?

As previously indicated, I did not want ot reference back to the input
signal ground. IOW the output returns, thorugh the load resistances,
should be eually offset from the signal ground.

For example, if the swing is 2.5 to 7.5V, the load return needs to be
at 2.5V, not signal ground. Same for the other signal, but with minus
signs added. Perhaps I am not reading the diagram correctly.

BTW did you ever know Harry Lang on E Briown St? He introduced me to
electronics when I lived in Phoenix years ago.

Jerry Norris

 

[YD]

Late at night, by candle light, [email protected] (Jerry
Norris) penned this immortal opus:

I would then need to build two dedicated signal generators, one that
runs off +3 and +9, and the other from -3 and -9.

For example two XR2206's, which will run on split supply.

However, the output signals of the two emitter followers (npn and pnp
versions), shown in my posted circuits, need to be identical, except
for the differing potential.

Maybe you have something in mind, but I don't see how this can be
easily achieved without using one input signal for both.

Jerry Norris

You need just one generator, outputting 6vpp (+/- 3Vp). Split the
signal, in one branch connect a 6V DC source between the signal and
the output with + going to the output. For the other output just
invert the source.


+| 6V
+--| |--------o -3 to -9 V
| |
+-----+
| | 6V|+
_ +--| |--------o +3 to +9 V
/ \ |
| ~ |-3 to + 3 V
\_/
|

 

[Jerry Norris]

SOMEWHERE there HAS to be a potential between the two GROUNDS/RETURNS.

How about something like either of these circuits then?

http://www.mediafire.com/i/?7z4xtae01cnqnbw

Jerry Norris

 

[Jamie]

I can see that you and Jim are going to be best of friends.

John

I gotta say, this has been one of the longest threads based on the
simple practice of a single transistor phase splitter circuit for which
you feed the emitter side a (-) source instead of the common to generate
your +&- outputs. One only needs to tailor the DC bias on the base to
get the desired DC offset!

Hell, you can even use a single supply source using a V-ground...
20 v source into a Vground network ...

Jamie

 

[Jamie]

Looks like you are over your head when trying to understand
electronics.

What kind of isolation do you need between input and the outputs?

Is the signal always 20Hz?

Source impedance?

...Jim Thompson

Not knowing everything that is needed for a proper design is a big issue
how ever, if you were to look at this problem from a basic electrical
point of view...!!!

One common configurable control transformer found in basic electrical
panels that allow various voltage configs. These xformers have 2
windings on the output, (X1,X2 and X3,X4), which you use as separate
outputs if you wish.. You config the input so that yields a 1:1 ratio in
the transfer..

Each secondary can be fed with a DC offset. One with + voltage and
the other, - voltage.

With that, the applied 20hz, which should pass fine on these
transformers, will be seen with the desired offsets on the other side.

If the signal is 180 unmatched, you simple flip one of the secondary
outputs. Also this gives isolation with the input signal..

Problem solved! Next?



Jamie

 

[Jerry Norris]

I was hoping no one would notice. You must have given up on the human
race learning from mistakes ;-)

With that, the applied 20hz, which should pass fine on these
transformers, will be seen with the desired offsets on the other side.

Hmm ... a fractionating transformer. Why didn't I think of that
before.

Thanks to all for their generous assistance.

Jerry