I have an opamp with a 10 volt positive supply and need the output to
swing to about 16 volts. I was thinking I could add a paralleled 6.8
volt zener, a cap, and a floating power supply at the opamp output.

The power supply could be a PV optocoupler like this:

https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804

My load is obviously light, a varicap in fact.

On Friday, 3 November 2023 at 15:02:35 UTC, John Larkin wrote:
> I have an opamp with a 10 volt positive supply and need the output to
> swing to about 16 volts. I was thinking I could add a paralleled 6.8
> volt zener, a cap, and a floating power supply at the opamp output.
>
> The power supply could be a PV optocoupler like this:
>
> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>
> My load is obviously light, a varicap in fact.

Why use a zener at all? Its not likely to add much voltage stability at the very
low current available from the optocoupler.
Alternatively, with a low power op-amp, put the optocoupler in series with
the positive power supply and its voltage instability will vanish from the
final output.
John

On Fri, 3 Nov 2023 08:20:56 -0700 (PDT), John Walliker
<jrwalliker@gmail.com> wrote:

>On Friday, 3 November 2023 at 15:02:35 UTC, John Larkin wrote:
>> I have an opamp with a 10 volt positive supply and need the output to
>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>> volt zener, a cap, and a floating power supply at the opamp output.
>>
>> The power supply could be a PV optocoupler like this:
>>
>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>
>> My load is obviously light, a varicap in fact.
>
>Why use a zener at all? Its not likely to add much voltage stability at the very
>low current available from the optocoupler.
>Alternatively, with a low power op-amp, put the optocoupler in series with
>the positive power supply and its voltage instability will vanish from the
>final output.
>John

Dumping the zener is a possibility, if the PV voltage were predictable
and stable, which it may not be. PVs have a big negative voltage
tempco.

A PV opto can only supply 10s of uA, and the opamp needs many mA of
supply voltage.

I'll test the pv+zener combo and see how stable it is.

Zeners can oscillate at low currents, but a hefty bypass cap should
fix that.

John Larkin <jl@997PotHill.com> wrote:
> I have an opamp with a 10 volt positive supply and need the output to
> swing to about 16 volts. I was thinking I could add a paralleled 6.8
> volt zener, a cap, and a floating power supply at the opamp output.
>
> The power supply could be a PV optocoupler like this:
>
> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>
> My load is obviously light, a varicap in fact.
>
>

Looks nice. You’ll want to filter the LED drive carefully, because with a
bang-bang phase detector your loop bandwidth will be limited.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC /
Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics

Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:
> John Larkin <jl@997PotHill.com> wrote:
>> I have an opamp with a 10 volt positive supply and need the output to
>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>> volt zener, a cap, and a floating power supply at the opamp output.
>>
>> The power supply could be a PV optocoupler like this:
>>
>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>
>> My load is obviously light, a varicap in fact.
>>
>>
>
> Looks nice. You’ll want to filter the LED drive carefully, because with a
> bang-bang phase detector your loop bandwidth will be limited.
>
> Cheers
>
> Phil Hobbs
>

Oh, and at 10 uA, the zener noise will probably be a worry.

--
Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC /
Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics

On Fri, 3 Nov 2023 15:47:12 -0000 (UTC), Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:
>> John Larkin <jl@997PotHill.com> wrote:
>>> I have an opamp with a 10 volt positive supply and need the output to
>>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>>> volt zener, a cap, and a floating power supply at the opamp output.
>>>
>>> The power supply could be a PV optocoupler like this:
>>>
>>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>>
>>> My load is obviously light, a varicap in fact.
>>>
>>>
>>
>> Looks nice. You’ll want to filter the LED drive carefully, because with a
>> bang-bang phase detector your loop bandwidth will be limited.
>>
>> Cheers
>>
>> Phil Hobbs
>>
>
>Oh, and at 10 uA, the zener noise will probably be a worry.

Hence the beefy bypass cap, 10 uF maybe. Gotta try it. I'm dealing
with a bunch of unspecified behavior. The current design has a 24 volt
power supply whose only function is to bias that zener. A PV coupler
is a small, cheap, quiet, wimpy isolated DC/DC converter.

The PV LED will just be driven from 5 volts through a resistor. The PV
current keeps the zener alive, so a little noise on the +5 shouldn't
get into the loop.

The loop isn't bang-bang, it's based on a fast 10-bit ADC observing
the waveform from a 50 MHz LC oscillator, and a lot of DSP.

On Friday, November 3, 2023 at 11:02:35 AM UTC-4, John Larkin wrote:
> I have an opamp with a 10 volt positive supply and need the output to
> swing to about 16 volts. I was thinking I could add a paralleled 6.8
> volt zener, a cap, and a floating power supply at the opamp output.
>
> The power supply could be a PV optocoupler like this:
>
> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>
> My load is obviously light, a varicap in fact.

Not seeing the need for a zener...

I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.

Is there some reason you can't drive current through the detector? Maybe too non-linearizing...

They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place..

On 03/11/2023 16:47, Fred Bloggs wrote:
> On Friday, November 3, 2023 at 11:02:35 AM UTC-4, John Larkin wrote:
>> I have an opamp with a 10 volt positive supply and need the output to
>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>> volt zener, a cap, and a floating power supply at the opamp output.
>>
>> The power supply could be a PV optocoupler like this:
>>
>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>
>> My load is obviously light, a varicap in fact.
>
> Not seeing the need for a zener...
>
> I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.
>
> Is there some reason you can't drive current through the detector? Maybe too non-linearizing...
>
> They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place.

Not clear to me from JL's original post if the op amp has a negative
supply rail so that even after the 7V output offset the varicap can be
biased below 7V?

Modulating the PV LED drive could have horrible delays and upset the
dynamics?

piglet

On Fri, 3 Nov 2023 09:47:09 -0700 (PDT), Fred Bloggs
<bloggs.fredbloggs.fred@gmail.com> wrote:

>On Friday, November 3, 2023 at 11:02:35?AM UTC-4, John Larkin wrote:
>> I have an opamp with a 10 volt positive supply and need the output to
>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>> volt zener, a cap, and a floating power supply at the opamp output.
>>
>> The power supply could be a PV optocoupler like this:
>>
>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>
>> My load is obviously light, a varicap in fact.
>
>Not seeing the need for a zener...

The PV voltage has a bad negative tempco, and that could cause us
problems. A zener helps that. An LM4040 bandgap would be great, but
they only come in 5 and 10 volts, too low and too high!
>
>I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.

If I have a pretty constant offset added to the opamp output, I don't
need to feed back from the offset, just from the opamp output. If I
use the PV for the offset, I don't have enough current for the opampm
feedback resistor.
>
>Is there some reason you can't drive current through the detector? Maybe too non-linearizing...
>
>They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place.

To slow.

Nice puzzle.

On 03/11/2023 19:19, john larkin wrote:
> On Fri, 3 Nov 2023 09:47:09 -0700 (PDT), Fred Bloggs
> <bloggs.fredbloggs.fred@gmail.com> wrote:
>
>> On Friday, November 3, 2023 at 11:02:35?AM UTC-4, John Larkin wrote:
>>> I have an opamp with a 10 volt positive supply and need the output to
>>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>>> volt zener, a cap, and a floating power supply at the opamp output.
>>>
>>> The power supply could be a PV optocoupler like this:
>>>
>>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>>
>>> My load is obviously light, a varicap in fact.
>>
>> Not seeing the need for a zener...
>
> The PV voltage has a bad negative tempco, and that could cause us
> problems. A zener helps that. An LM4040 bandgap would be great, but
> they only come in 5 and 10 volts, too low and too high!
>>
>> I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.
>
> If I have a pretty constant offset added to the opamp output, I don't
> need to feed back from the offset, just from the opamp output. If I
> use the PV for the offset, I don't have enough current for the opampm
> feedback resistor.
>>
>> Is there some reason you can't drive current through the detector? Maybe too non-linearizing...
>>
>> They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place.
>
> To slow.
>
> Nice puzzle.
>

There is the adjustable LM4041 but the minimum operating current will be
too much for the PV.

Most action in a varicap is in the 1-10V range, can't you simply pad the
varicap capacitance down so you avoid the need to go to 16V?

piglet

On Friday, November 3, 2023 at 3:17:48 PM UTC-4, piglet wrote:
> On 03/11/2023 16:47, Fred Bloggs wrote:
> > On Friday, November 3, 2023 at 11:02:35 AM UTC-4, John Larkin wrote:
> >> I have an opamp with a 10 volt positive supply and need the output to
> >> swing to about 16 volts. I was thinking I could add a paralleled 6.8
> >> volt zener, a cap, and a floating power supply at the opamp output.
> >>
> >> The power supply could be a PV optocoupler like this:
> >>
> >> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
> >>
> >> My load is obviously light, a varicap in fact.
> >
> > Not seeing the need for a zener...
> >
> > I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.
> >
> > Is there some reason you can't drive current through the detector? Maybe too non-linearizing...
> >
> > They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place.
> Not clear to me from JL's original post if the op amp has a negative
> supply rail so that even after the 7V output offset the varicap can be
> biased below 7V?

Good point. Shunt the PV with an anti-parallel diode.

>
> Modulating the PV LED drive could have horrible delays and upset the
> dynamics?
>
> piglet

On Friday, November 3, 2023 at 3:19:36 PM UTC-4, john larkin wrote:
> On Fri, 3 Nov 2023 09:47:09 -0700 (PDT), Fred Bloggs
> <bloggs.fred...@gmail.com> wrote:
> >On Friday, November 3, 2023 at 11:02:35?AM UTC-4, John Larkin wrote:
> >> I have an opamp with a 10 volt positive supply and need the output to
> >> swing to about 16 volts. I was thinking I could add a paralleled 6.8
> >> volt zener, a cap, and a floating power supply at the opamp output.
> >>
> >> The power supply could be a PV optocoupler like this:
> >>
> >> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
> >>
> >> My load is obviously light, a varicap in fact.
> >
> >Not seeing the need for a zener...
> The PV voltage has a bad negative tempco, and that could cause us
> problems. A zener helps that. An LM4040 bandgap would be great, but
> they only come in 5 and 10 volts, too low and too high!
> >
> >I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.
> If I have a pretty constant offset added to the opamp output, I don't
> need to feed back from the offset, just from the opamp output. If I
> use the PV for the offset, I don't have enough current for the opampm
> feedback resistor.

You do if you bootstrap a buffer into the feedback divider.

This example looks simplified:

https://www.digikey.com/en/articles/how-to-solve-analog-high-voltage-delivery-challenges

This approach is complicated and parts intensive, but the author is an IC designer so number of parts don't matter:

https://www.analog.com/en/analog-dialogue/raqs/raq-issue-168.html

The bootstrapped buffer drives the resistor divider feedback to the first OA.

> >
> >Is there some reason you can't drive current through the detector? Maybe too non-linearizing...
> >
> >They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place.
> To slow.
>
> Nice puzzle.

On Friday, November 3, 2023 at 3:19:36 PM UTC-4, john larkin wrote:
> On Fri, 3 Nov 2023 09:47:09 -0700 (PDT), Fred Bloggs
> <bloggs.fred...@gmail.com> wrote:
> >On Friday, November 3, 2023 at 11:02:35?AM UTC-4, John Larkin wrote:
> >> I have an opamp with a 10 volt positive supply and need the output to
> >> swing to about 16 volts. I was thinking I could add a paralleled 6.8
> >> volt zener, a cap, and a floating power supply at the opamp output.
> >>
> >> The power supply could be a PV optocoupler like this:
> >>
> >> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
> >>
> >> My load is obviously light, a varicap in fact.
> >
> >Not seeing the need for a zener...
> The PV voltage has a bad negative tempco, and that could cause us
> problems. A zener helps that. An LM4040 bandgap would be great, but
> they only come in 5 and 10 volts, too low and too high!
> >
> >I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.
> If I have a pretty constant offset added to the opamp output, I don't
> need to feed back from the offset, just from the opamp output. If I
> use the PV for the offset, I don't have enough current for the opampm
> feedback resistor.

Did you say you have unregulated 24VDC? Then just find an OA that can handle it and use that to buffer the feedback resistor divider.

> >
> >Is there some reason you can't drive current through the detector? Maybe too non-linearizing...
> >
> >They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place.
> To slow.
>
> Nice puzzle.

On Fri, 3 Nov 2023 19:29:46 +0000, piglet <erichpwagner@hotmail.com>
wrote:

>On 03/11/2023 19:19, john larkin wrote:
>> On Fri, 3 Nov 2023 09:47:09 -0700 (PDT), Fred Bloggs
>> <bloggs.fredbloggs.fred@gmail.com> wrote:
>>
>>> On Friday, November 3, 2023 at 11:02:35?AM UTC-4, John Larkin wrote:
>>>> I have an opamp with a 10 volt positive supply and need the output to
>>>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>>>> volt zener, a cap, and a floating power supply at the opamp output.
>>>>
>>>> The power supply could be a PV optocoupler like this:
>>>>
>>>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>>>
>>>> My load is obviously light, a varicap in fact.
>>>
>>> Not seeing the need for a zener...
>>
>> The PV voltage has a bad negative tempco, and that could cause us
>> problems. A zener helps that. An LM4040 bandgap would be great, but
>> they only come in 5 and 10 volts, too low and too high!
>>>
>>> I take it mean the standard non-inverting OA with the PV in series with OA output, and a high-Z resistor feedback divider from the composite output (PV cathode) to OA (-). Loop gain, which should be very large attenuates noise and the unknowns at the composite output.
>>
>> If I have a pretty constant offset added to the opamp output, I don't
>> need to feed back from the offset, just from the opamp output. If I
>> use the PV for the offset, I don't have enough current for the opampm
>> feedback resistor.
>>>
>>> Is there some reason you can't drive current through the detector? Maybe too non-linearizing...
>>>
>>> They datasheet is so fixed on detector Voc Isc with input LED drive, it's almost tempting to use the LED drive to modulate the final output into place.
>>
>> To slow.
>>
>> Nice puzzle.
>>
>
>There is the adjustable LM4041 but the minimum operating current will be
>too much for the PV.
>
>Most action in a varicap is in the 1-10V range, can't you simply pad the
>varicap capacitance down so you avoid the need to go to 16V?
>
>piglet

The varicap is violently nonlinear at low voltages. Something like 6
to 18 volts has about a 2:1 slope, which we can barely tolerate. Maybe
I can talk the boys into doing a 2nd order polynomial in the FPGA to
correct the varicap curve and keep the control loop happier.

But today's revelation is to bias the cold side of the varicap to -8V,
and not grounding it, instead of shifting the opamp output up!

On Friday, November 3, 2023 at 8:02:35 AM UTC-7, John Larkin wrote:
> I have an opamp with a 10 volt positive supply and need the output to
> swing to about 16 volts. I was thinking I could add a paralleled 6.8
> volt zener, a cap, and a floating power supply at the opamp output.
>
> The power supply could be a PV optocoupler like this:
>
> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>
> My load is obviously light, a varicap in fact.
So, put two lithium coin cells in series with the op amp output (it'll work for years if the load is really light).

Zeners sound good until you actually see one with very light current through it; they can be unstable and noisy.

On a sunny day (Fri, 03 Nov 2023 08:01:58 -0700) it happened John Larkin
<jl@997PotHill.com> wrote in <db2aki9v2gst0o57hamemq56brv7vscdol@4ax.com>:

>I have an opamp with a 10 volt positive supply and need the output to
>swing to about 16 volts. I was thinking I could add a paralleled 6.8
>volt zener, a cap, and a floating power supply at the opamp output.
>
>The power supply could be a PV optocoupler like this:
>
>https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>
>My load is obviously light, a varicap in fact.

Interesting, you get about 8 V at 10 mA and 1 M load.

On a sunny day (Fri, 03 Nov 2023 12:19:20 -0700) it happened john larkin
<jl@650pot.com> wrote in <7ugakit1jtm431b4jaml3cs43gla1amdi8@4ax.com>:

>On Fri, 3 Nov 2023 09:47:09 -0700 (PDT), Fred Bloggs
><bloggs.fredbloggs.fred@gmail.com> wrote:
>
>>On Friday, November 3, 2023 at 11:02:35?AM UTC-4, John Larkin wrote:
>>> I have an opamp with a 10 volt positive supply and need the output to
>>> swing to about 16 volts. I was thinking I could add a paralleled 6.8
>>> volt zener, a cap, and a floating power supply at the opamp output.
>>>
>>> The power supply could be a PV optocoupler like this:
>>>
>>> https://www.digikey.com/en/products/detail/toshiba-semiconductor-and-storage/TLP3905-TPR-E/7809804
>>>
>>> My load is obviously light, a varicap in fact.

If you want to tune a 50 MHz? oscillator with a varicap,
I tune those with a transistor Cce,
here tuning a crystal oscillator, 25 MHz f0, pot on the left.
https://panteltje.nl/pub/lnb_ref_circuit_update_IMG_6652.JPG
No need for high voltages, pick a low capacitance NPN or PNP if frequencies are very high.