signal pickoff

[John Larkin]

On Mon, 6 May 2019 13:13:15 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 5/6/19 12:38 AM, John Larkin wrote:
On Sun, 05 May 2019 08:39:00 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:



I'm daydreaming a new product, and this is one problem:

https://www.dropbox.com/s/oo9gixme2mocsc3/kbox_pickoff.JPG?dl=0


The IC that I want to drive will be about a 2 pF load, and that wrecks
the 50 ohm microstrip transmission line that connects the two SMAs.
The idea is that a user can input a signal and either loop it into
another (several?) boxes, or screw on a terminator.

The 2 pF makes really bad stuff happen around 2 GHz or so, for the
loop-through path, and I'd like to go to 5. What I need is a
zero-capacitance buffer.

A phemt source follower might work, with a SAV551 or one of the
Skyworks parts. But fast source followers are also known as
oscillators.

Maybe I could use a series resistor into a 50 ohm MMIC amplifier. I'd
lose some signal but at least wouldn't have the horrible resonances
that a lumped capacitor makes.

A decent compromise is a pickoff that uses a 100 ohm to 50 ohm
divider, with a small inductor in series with the 50 to AC ground.
That gives me 1/3 of the customer voltage for my IC (a comparator) and
a bit of high frequency peaking where I probably want it. The damage
to the thru signal is small, even less at high frequencies.

It's an ADCMP572, and the 50 ohm resistor is actually inside the chip.

The phemt source follower is still interesting. I might try one for
fun.

The SKY65050 makes a semi-decent follower, about like your average JFET.
The Avago ones (ATF38143 etc.) have such low drain impedances that
their gains as a follower are down around 0.6 or something horrible like
that. Small pHEMTs are surprisingly stable due to their very low
interelectrode capacitances--you have to really work to make one oscillate.

I got burned recently with a BFT25 follower, oscillating at invisibly
high frequencies. Base resistance fixed it. It's interesting that
phemts make stable followers.

I haven't tried the Mini Circuits ones yet, though I have a hundred or
so in stock. The new Renesas 12-GHz class pHEMTs look pretty
interesting too.

Bootstrapping a pHEMT with a SiGe:C transistor (BFP640) makes a nearly
perfect gain stage, except that you have to de-Q the base of the BJT or
it'll oscillate at 14 GHz or so. I usually use a BLM18BB05 (5 ohms at
100 MHz) for that. Dunno how it would work at 5 GHz--my application was
at a bandwidth of 100 MHz.

Cheers

Phil Hobbs

I'll try the loopthru pickoff circuit with a MiniCircuits phemt
follower. I don't need voltage gain of 1... something like 0.6 would
be fine.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Phil Hobbs]

On 5/6/19 4:42 PM, John Larkin wrote:

On Mon, 6 May 2019 13:13:15 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 5/6/19 12:38 AM, John Larkin wrote:
On Sun, 05 May 2019 08:39:00 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:



I'm daydreaming a new product, and this is one problem:

https://www.dropbox.com/s/oo9gixme2mocsc3/kbox_pickoff.JPG?dl=0


The IC that I want to drive will be about a 2 pF load, and that wrecks
the 50 ohm microstrip transmission line that connects the two SMAs.
The idea is that a user can input a signal and either loop it into
another (several?) boxes, or screw on a terminator.

The 2 pF makes really bad stuff happen around 2 GHz or so, for the
loop-through path, and I'd like to go to 5. What I need is a
zero-capacitance buffer.

A phemt source follower might work, with a SAV551 or one of the
Skyworks parts. But fast source followers are also known as
oscillators.

Maybe I could use a series resistor into a 50 ohm MMIC amplifier. I'd
lose some signal but at least wouldn't have the horrible resonances
that a lumped capacitor makes.

A decent compromise is a pickoff that uses a 100 ohm to 50 ohm
divider, with a small inductor in series with the 50 to AC ground.
That gives me 1/3 of the customer voltage for my IC (a comparator) and
a bit of high frequency peaking where I probably want it. The damage
to the thru signal is small, even less at high frequencies.

It's an ADCMP572, and the 50 ohm resistor is actually inside the chip.

The phemt source follower is still interesting. I might try one for
fun.

The SKY65050 makes a semi-decent follower, about like your average JFET.
The Avago ones (ATF38143 etc.) have such low drain impedances that
their gains as a follower are down around 0.6 or something horrible like
that. Small pHEMTs are surprisingly stable due to their very low
interelectrode capacitances--you have to really work to make one oscillate.

I got burned recently with a BFT25 follower, oscillating at invisibly
high frequencies. Base resistance fixed it. It's interesting that
phemts make stable followers.


I haven't tried the Mini Circuits ones yet, though I have a hundred or
so in stock. The new Renesas 12-GHz class pHEMTs look pretty
interesting too.

Bootstrapping a pHEMT with a SiGe:C transistor (BFP640) makes a nearly
perfect gain stage, except that you have to de-Q the base of the BJT or
it'll oscillate at 14 GHz or so. I usually use a BLM18BB05 (5 ohms at
100 MHz) for that. Dunno how it would work at 5 GHz--my application was
at a bandwidth of 100 MHz.

Cheers

Phil Hobbs

I'll try the loopthru pickoff circuit with a MiniCircuits phemt
follower. I don't need voltage gain of 1... something like 0.6 would
be fine.

Yes, except that then it doesn't bootstrap Cgs very well.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Winfield Hill]

Phil Hobbs wrote...

The SKY65050 makes a semi-decent follower, about like
your average JFET.

What values do you use for the SKY65050 capacitances,
when you're doing back-of-the envelope calcs? I did
a bunch of SPICE modeling for this part about 8 years
ago, but it makes little sense now. Maybe Ciss = 1.1
to 1.5pF, and Crss must be much smaller than that.


--
Thanks,
- Win

 

[John Larkin]

On Mon, 6 May 2019 17:02:39 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 5/6/19 4:42 PM, John Larkin wrote:
On Mon, 6 May 2019 13:13:15 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 5/6/19 12:38 AM, John Larkin wrote:
On Sun, 05 May 2019 08:39:00 -0700, John Larkin
jjlarkin@highlandtechnology.com> wrote:



I'm daydreaming a new product, and this is one problem:

https://www.dropbox.com/s/oo9gixme2mocsc3/kbox_pickoff.JPG?dl=0


The IC that I want to drive will be about a 2 pF load, and that wrecks
the 50 ohm microstrip transmission line that connects the two SMAs.
The idea is that a user can input a signal and either loop it into
another (several?) boxes, or screw on a terminator.

The 2 pF makes really bad stuff happen around 2 GHz or so, for the
loop-through path, and I'd like to go to 5. What I need is a
zero-capacitance buffer.

A phemt source follower might work, with a SAV551 or one of the
Skyworks parts. But fast source followers are also known as
oscillators.

Maybe I could use a series resistor into a 50 ohm MMIC amplifier. I'd
lose some signal but at least wouldn't have the horrible resonances
that a lumped capacitor makes.

A decent compromise is a pickoff that uses a 100 ohm to 50 ohm
divider, with a small inductor in series with the 50 to AC ground.
That gives me 1/3 of the customer voltage for my IC (a comparator) and
a bit of high frequency peaking where I probably want it. The damage
to the thru signal is small, even less at high frequencies.

It's an ADCMP572, and the 50 ohm resistor is actually inside the chip.

The phemt source follower is still interesting. I might try one for
fun.

The SKY65050 makes a semi-decent follower, about like your average JFET.
The Avago ones (ATF38143 etc.) have such low drain impedances that
their gains as a follower are down around 0.6 or something horrible like
that. Small pHEMTs are surprisingly stable due to their very low
interelectrode capacitances--you have to really work to make one oscillate.

I got burned recently with a BFT25 follower, oscillating at invisibly
high frequencies. Base resistance fixed it. It's interesting that
phemts make stable followers.


I haven't tried the Mini Circuits ones yet, though I have a hundred or
so in stock. The new Renesas 12-GHz class pHEMTs look pretty
interesting too.

Bootstrapping a pHEMT with a SiGe:C transistor (BFP640) makes a nearly
perfect gain stage, except that you have to de-Q the base of the BJT or
it'll oscillate at 14 GHz or so. I usually use a BLM18BB05 (5 ohms at
100 MHz) for that. Dunno how it would work at 5 GHz--my application was
at a bandwidth of 100 MHz.

Cheers

Phil Hobbs

I'll try the loopthru pickoff circuit with a MiniCircuits phemt
follower. I don't need voltage gain of 1... something like 0.6 would
be fine.



Yes, except that then it doesn't bootstrap Cgs very well.

Right. I measured 0.92 pF at the gate with the source and drain
grounded, about 1 pF at the drain with others grounded.

0.6 bootstrapping of Cgs would be about 0.6 pF, a decent improvement
in my situation.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Phil Hobbs]

On 5/6/19 5:44 PM, Winfield Hill wrote:

Phil Hobbs wrote...

The SKY65050 makes a semi-decent follower, about like
your average JFET.

What values do you use for the SKY65050 capacitances,
when you're doing back-of-the envelope calcs? I did
a bunch of SPICE modeling for this part about 8 years
ago, but it makes little sense now. Maybe Ciss = 1.1
to 1.5pF, and Crss must be much smaller than that.

Ciss is less than that--my biochip preamp had a total input capacitance
of about 1 pF. Crss is probably in the 0.1 pF range.

I built a calibration source into my pHEMT/SiGe:C cascode TIA, using
discrete pHEMTs for followers and Schmitt triggers and stuff--it was a
very triangular triangle wave oscillator coupled into the summing
junction via about 0.01 pF of board capacitance. With the Avago pHEMTs
it completely failed to work on account of the low drain impedance, but
it worked OK with the Skyworks ones after I fiddled with the biasing a bit.

That was one of those moonshot projects--some biophysics types at
Samsung's main research lab were doing DNA sequencing the way a kid eats
spaghetti: sucking half strands through a micropore in a Si3N4 membrane
electrostatically, and measuring the change in transverse conductance as
the different amino acid bases went between the pads on the top side.

The spec was to measure 1 nA at the shot noise limit in a 100 MHz
bandwidth. The equivalent measurement time is 5 ns, and 1 nA times 5 ns
is 31 electrons. I pointed out to them that their maximum SNR was going
to be 15 dB, but they wanted to go ahead anyway.

The only lucky thing was that I insisted that they mount their biochip
on my board and wire bond it to the amp input.

I got within a factor of 2 (11 electrons RMS), but you really can't
measure anything with a SNR of 9 dB. I suppose that they could have
done pattern matching to combine data from many strands, sort of how you
do with tree ring dating.

It was great fun, and I think I extended the state of the amplifier art
in an unusual direction.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Phil Hobbs]

On 5/6/19 6:10 PM, Phil Hobbs wrote:

On 5/6/19 5:44 PM, Winfield Hill wrote:
Phil Hobbs wrote...

The SKY65050 makes a semi-decent follower, about like
your average JFET.

  What values do you use for the SKY65050 capacitances,
  when you're doing back-of-the envelope calcs?  I did
  a bunch of SPICE modeling for this part about 8 years
  ago, but it makes little sense now.  Maybe Ciss = 1.1
  to 1.5pF, and Crss must be much smaller than that.


Ciss is less than that--my biochip preamp had a total input capacitance
of about 1 pF.  Crss is probably in the 0.1 pF range.

I built a calibration source into my pHEMT/SiGe:C cascode TIA, using
discrete pHEMTs for followers and Schmitt triggers and stuff--it was a
very triangular triangle wave oscillator coupled into the summing
junction via about 0.01 pF of board capacitance.  With the Avago pHEMTs
it completely failed to work on account of the low drain impedance, but
it worked OK with the Skyworks ones after I fiddled with the biasing a bit.

That was one of those moonshot projects--some biophysics types at
Samsung's main research lab were doing DNA sequencing the way a kid eats
spaghetti: sucking half strands through a micropore in a Si3N4 membrane
electrostatically, and measuring the change in transverse conductance as
the different amino acid bases went between the pads on the top side.

The spec was to measure 1 nA at the shot noise limit in a 100 MHz
bandwidth.  The equivalent measurement time is 5 ns, and 1 nA times 5 ns
is 31 electrons.  I pointed out to them that their maximum SNR was going
to be 15 dB, but they wanted to go ahead anyway.

The only lucky thing was that I insisted that they mount their biochip
on my board and wire bond it to the amp input.

I got within a factor of 2 (11 electrons RMS), but you really can't
measure anything with a SNR of 9 dB.  I suppose that they could have
done pattern matching to combine data from many strands, sort of how you
do with tree ring dating.

It was great fun, and I think I extended the state of the amplifier art
in an unusual direction.

Cheers

Phil Hobbs

Oh, and the ATF38143 datasheet has a super detailed SPICE model for the
chip and package, so I used that.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

On Monday, May 6, 2019 at 12:30:35 PM UTC-4, John Larkin wrote:

On Mon, 6 May 2019 08:50:23 -0700 (PDT),
bloggs.fredbloggs.fred@gmail.com wrote:

On Sunday, May 5, 2019 at 8:10:15 PM UTC-4, John Larkin wrote:
On Sun, 5 May 2019 16:56:52 -0700 (PDT),
bloggs.fredbloggs.fred@gmail.com wrote:

On Sunday, May 5, 2019 at 11:39:09 AM UTC-4, John Larkin wrote:
I'm daydreaming a new product, and this is one problem:

https://www.dropbox.com/s/oo9gixme2mocsc3/kbox_pickoff.JPG?dl=0


The IC that I want to drive will be about a 2 pF load, and that wrecks
the 50 ohm microstrip transmission line that connects the two SMAs.
The idea is that a user can input a signal and either loop it into
another (several?) boxes, or screw on a terminator.

The 2 pF makes really bad stuff happen around 2 GHz or so, for the
loop-through path, and I'd like to go to 5. What I need is a
zero-capacitance buffer.

A phemt source follower might work, with a SAV551 or one of the
Skyworks parts. But fast source followers are also known as
oscillators.

Maybe I could use a series resistor into a 50 ohm MMIC amplifier. I'd
lose some signal but at least wouldn't have the horrible resonances
that a lumped capacitor makes.

Make one of these:
https://www.jfwindustries.com/product/50c-051-20-coupler/

Should just be something like so:
https://www.microwaves101.com/encyclopedias/resistive-taps


20 dB is a lot of signal to give up. 3 dB would be OK but wrecks the
thru path. A zero-pF follower would be ideal. Or else a resistive
pickoff and some wideband (as in 5 GHz) gain to get the signal level
back.

Your signal level is nearly 4dBm, which in the RF world is HUUUUGE, so HUUUGE in fact many small signal monolithic amps introduce serious harmonic + spur content hitting their intercept point with that kind of *input* level. You knock that thing down to -16dBm where you can get reasonably harmonic and spur free amplification. But if you want to pay $100 for a high IIP3 amp to do a $0.99 job, have at it, your call. You want that 20dB coupling to isolate your through channel from the pickoff channel, the more attenuation the better. Attenuators are deliberately introduced in all kinds of circuits to kill non-linearities.


Did I mention 4 dBm?

We have the LEE-19 in stock, an 8 GHz 12 dB MMIC. Costs $1.19. A
resistor from the microstrip trace, hundreds of ohms, would get me
about net unity voltage gain at the MMIC output.

Too much gain doesn't worry me. I can sell the users attenuators.

The signal is a clock, into a comparator, so RF type distortion specs
don't matter.

Does jitter worry you? That LEE-19 has 4dB greater NF than any amplifier in the product family.

--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Simon S Aysdie]

On Sunday, May 5, 2019 at 8:39:09 AM UTC-7, John Larkin wrote:

I'm daydreaming a new product, and this is one problem:

https://www.dropbox.com/s/oo9gixme2mocsc3/kbox_pickoff.JPG?dl=0


The IC that I want to drive will be about a 2 pF load, and that wrecks
the 50 ohm microstrip transmission line that connects the two SMAs.
The idea is that a user can input a signal and either loop it into
another (several?) boxes, or screw on a terminator.

The 2 pF makes really bad stuff happen around 2 GHz or so, for the
loop-through path, and I'd like to go to 5. What I need is a
zero-capacitance buffer.

A phemt source follower might work, with a SAV551 or one of the
Skyworks parts. But fast source followers are also known as
oscillators.

Maybe I could use a series resistor into a 50 ohm MMIC amplifier. I'd
lose some signal but at least wouldn't have the horrible resonances
that a lumped capacitor makes.

I often use a directional coupler for this, usually about 10 dB. Mini-Circuits is good for that. I has nice control of impedance at all ports and can be quite wide-band.

IN
>-. .-< Coupled
| |
| |
| |
| |

-' '-R-/
OUT

 

[John Larkin]

On Tue, 14 May 2019 11:10:51 -0700 (PDT), Simon S Aysdie
<gwhite@ti.com> wrote:

On Sunday, May 5, 2019 at 8:39:09 AM UTC-7, John Larkin wrote:
I'm daydreaming a new product, and this is one problem:

https://www.dropbox.com/s/oo9gixme2mocsc3/kbox_pickoff.JPG?dl=0


The IC that I want to drive will be about a 2 pF load, and that wrecks
the 50 ohm microstrip transmission line that connects the two SMAs.
The idea is that a user can input a signal and either loop it into
another (several?) boxes, or screw on a terminator.

The 2 pF makes really bad stuff happen around 2 GHz or so, for the
loop-through path, and I'd like to go to 5. What I need is a
zero-capacitance buffer.

A phemt source follower might work, with a SAV551 or one of the
Skyworks parts. But fast source followers are also known as
oscillators.

Maybe I could use a series resistor into a 50 ohm MMIC amplifier. I'd
lose some signal but at least wouldn't have the horrible resonances
that a lumped capacitor makes.

I often use a directional coupler for this, usually about 10 dB. Mini-Circuits is good for that. I has nice control of impedance at all ports and can be quite wide-band.

IN
-. .-< Coupled
| |
| |
| |
| |
-' '-R-/
OUT

We've just finished some tests, using a few different MMICS with
series input resistors, rather than the usual 50 ohm drive.

A LEE-59 with a 510 ohm series input resistor, driving a 50 ohm
comparator on its output side, has an overall voltage gain (pickoff
point to comparator input) of 1. The thru-path trace won't notice the
560 ohm load, and an 0603 resistor's capacitance will be tiny,
something like 0.05 pF.

A LEE-59 is rated at 5 GHz. Costs about $2.

This would be super wideband. I'd put a pretty big cap (physically
small!) as a DC block in series with the 510 ohm resistor.



--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[John Miles, KE5FX]

On Tuesday, May 14, 2019 at 4:14:58 PM UTC-7, John Larkin wrote:

We've just finished some tests, using a few different MMICS with
series input resistors, rather than the usual 50 ohm drive.

As another alternative, there's a name for the sort of coupler
used on these boards, but I can't remember it:

http://www.ke5fx.com/stellex.htm (specifically the photos near the
July 10, 2010 entry)


But yeah, a ~500 ohm pickoff resistor should be fine if you don't care
about any directional properties.

-- john, KE5FX

 

[John Larkin]

On Sat, 18 May 2019 19:11:54 -0700 (PDT), "John Miles, KE5FX"
<jmiles@gmail.com> wrote:

On Tuesday, May 14, 2019 at 4:14:58 PM UTC-7, John Larkin wrote:

We've just finished some tests, using a few different MMICS with
series input resistors, rather than the usual 50 ohm drive.


As another alternative, there's a name for the sort of coupler
used on these boards, but I can't remember it:

http://www.ke5fx.com/stellex.htm (specifically the photos near the
July 10, 2010 entry)


But yeah, a ~500 ohm pickoff resistor should be fine if you don't care
about any directional properties.

-- john, KE5FX

I want an invisible, minimally invasive voltage pickoff, from maybe 1
MHz to several GHz. A phemt source follower might work too - haven't
tried that yet - but it would probably be worse than the 500 ohms at
high frequencies. The s-g capacitance would be bootstrapped, but the
follower gain would be closer to 0.5 than to 1.

500 ohms costs about 5% of my signal amplitude and is flat over
frequency.

MMICS are cool when used off-label.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

On Monday, May 20, 2019 at 1:13:02 PM UTC-4, John Larkin wrote:

On Sat, 18 May 2019 19:11:54 -0700 (PDT), "John Miles, KE5FX"
jmiles@gmail.com> wrote:

On Tuesday, May 14, 2019 at 4:14:58 PM UTC-7, John Larkin wrote:

We've just finished some tests, using a few different MMICS with
series input resistors, rather than the usual 50 ohm drive.


As another alternative, there's a name for the sort of coupler
used on these boards, but I can't remember it:

http://www.ke5fx.com/stellex.htm (specifically the photos near the
July 10, 2010 entry)


But yeah, a ~500 ohm pickoff resistor should be fine if you don't care
about any directional properties.

-- john, KE5FX

I want an invisible, minimally invasive voltage pickoff, from maybe 1
MHz to several GHz. A phemt source follower might work too - haven't
tried that yet - but it would probably be worse than the 500 ohms at
high frequencies. The s-g capacitance would be bootstrapped, but the
follower gain would be closer to 0.5 than to 1.

500 ohms costs about 5% of my signal amplitude and is flat over
frequency.

MMICS are cool when used off-label.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

I did this recently:

Vdd
-+-
|
'--|
|
.--|<--.
| |
|--' |
| |
>------>|--. |
| |
+------'
|
(is)
|
===

It's not the world's best bootstrap, but it's dead simple, works
at d.c., and a pretty big improvement on several fronts.

Cheers,
James Arthur

 

[John Larkin]

On Mon, 20 May 2019 13:53:51 -0700 (PDT), dagmargoodboat@yahoo.com
wrote:

On Monday, May 20, 2019 at 1:13:02 PM UTC-4, John Larkin wrote:
On Sat, 18 May 2019 19:11:54 -0700 (PDT), "John Miles, KE5FX"
jmiles@gmail.com> wrote:

On Tuesday, May 14, 2019 at 4:14:58 PM UTC-7, John Larkin wrote:

We've just finished some tests, using a few different MMICS with
series input resistors, rather than the usual 50 ohm drive.


As another alternative, there's a name for the sort of coupler
used on these boards, but I can't remember it:

http://www.ke5fx.com/stellex.htm (specifically the photos near the
July 10, 2010 entry)


But yeah, a ~500 ohm pickoff resistor should be fine if you don't care
about any directional properties.

-- john, KE5FX

I want an invisible, minimally invasive voltage pickoff, from maybe 1
MHz to several GHz. A phemt source follower might work too - haven't
tried that yet - but it would probably be worse than the 500 ohms at
high frequencies. The s-g capacitance would be bootstrapped, but the
follower gain would be closer to 0.5 than to 1.

500 ohms costs about 5% of my signal amplitude and is flat over
frequency.

MMICS are cool when used off-label.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

I did this recently:

Vdd
-+-
|
'--|
|
.--|<--.
| |
|--' |
| |
------>|--. |
| |
+------'
|
(is)
|
===

It's not the world's best bootstrap, but it's dead simple, works
at d.c., and a pretty big improvement on several fronts.

Cheers,
James Arthur

Bootstrap the upper jfet with another fet!


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

On Monday, May 20, 2019 at 10:08:44 PM UTC-4, John Larkin wrote:

On Mon, 20 May 2019 13:53:51 -0700 (PDT), dagmargoodboat@yahoo.com
wrote:

On Monday, May 20, 2019 at 1:13:02 PM UTC-4, John Larkin wrote:
On Sat, 18 May 2019 19:11:54 -0700 (PDT), "John Miles, KE5FX"
jmiles@gmail.com> wrote:

On Tuesday, May 14, 2019 at 4:14:58 PM UTC-7, John Larkin wrote:

We've just finished some tests, using a few different MMICS with
series input resistors, rather than the usual 50 ohm drive.


As another alternative, there's a name for the sort of coupler
used on these boards, but I can't remember it:

http://www.ke5fx.com/stellex.htm (specifically the photos near the
July 10, 2010 entry)


But yeah, a ~500 ohm pickoff resistor should be fine if you don't care
about any directional properties.

-- john, KE5FX

I want an invisible, minimally invasive voltage pickoff, from maybe 1
MHz to several GHz. A phemt source follower might work too - haven't
tried that yet - but it would probably be worse than the 500 ohms at
high frequencies. The s-g capacitance would be bootstrapped, but the
follower gain would be closer to 0.5 than to 1.

500 ohms costs about 5% of my signal amplitude and is flat over
frequency.

MMICS are cool when used off-label.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

I did this recently:

Vdd
-+-
|
'--|
|
.--|<--.
| |
|--' |
| |
------>|--. |
| |
+------'
|
(is)
|
===

It's not the world's best bootstrap, but it's dead simple, works
at d.c., and a pretty big improvement on several fronts.

Cheers,
James Arthur

Bootstrap the upper jfet with another fet!


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

The actual circuit's source load (is) was this...

R2 | |
J1 '-' |
|- | |<'
>| +------| Q2
|--. | Q1 |\
| |/ |
+----| Vg2<--+
| |>. |
.-. | .-.
R1 | | | R4 | |
'-' | '-'
| | |
'------+--------'
|
.-.
| | R3
'-'
|
===

For a 100 or 200-ish fF integrator.

Vg2 allows tuning the cascode's gate drive to null Cdg.

Q2 does all the R3-driving, so Q1 sees little d(ie)/dV,
and J1 sees a very decent current-source load approximation.

The net result was a follower gain of approximately .98,
with an effective input capacitance in the femtofarads.

But you don't need all that. Just wiping out even part of the
Miller capacitance gives you a big boost over a bare fet.

Cheers,
James