https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25

3.1 GHz unity-gain jfet buffer. Noise ain't bad.

--

If a man will begin with certainties, he shall end with doubts,
but if he will be content to begin with doubts he shall end in certainties.
Francis Bacon

On Tuesday, December 7, 2021 at 11:01:16 AM UTC+11, John Larkin wrote:
> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>
> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.

And if enough people show an interest n the data sheet, Texas Instrument may actually make a few.

I got burned by one of Texas Instruments "test the market" data sheets a back in 1972, and I've not trusted them ever since.

Don't design it in until you've got your hands on a couple of parts. It says "pre-proproduction - only available from TI" and that usually means a lot of stories about low yields and waiting for the next batch.

--
Bill Sloman, Sydney

John Larkin wrote:
> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>
> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.

Wow, almost as good as an LH0063, circa 1980. ;)

It's got higher bandwidth, for sure, and about the same slew rate, but
can't do 40V supplies.

The LH4009, which arrived with the NS 1989 Linear Databook and was gone
by the 1992 one, had a maximum slew of 10kV/us, and produced +-250 mA
from 40V supplies.

The extra bandwidth is useful, for sure, as long as the supply voltages
permit. ;)

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 Mon, 6 Dec 2021 22:09:07 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>John Larkin wrote:
>> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>>
>> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.
>
>Wow, almost as good as an LH0063, circa 1980. ;)
>
>It's got higher bandwidth, for sure, and about the same slew rate, but
>can't do 40V supplies.
>
>The LH4009, which arrived with the NS 1989 Linear Databook and was gone
>by the 1992 one, had a maximum slew of 10kV/us, and produced +-250 mA
>from 40V supplies.
>
>The extra bandwidth is useful, for sure, as long as the supply voltages
>permit. ;)
>
>Cheers
>
>Phil Hobbs

TI is pretty good about making fast amps that work at nose-bleed
voltages... like 10 volts.

--

Father Brown's figure remained quite dark and still;
but in that instant he had lost his head. His head was
always most valuable when he had lost it.

jlarkin@highlandsniptechnology.com wrote:
> On Mon, 6 Dec 2021 22:09:07 -0500, Phil Hobbs
> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>
>> John Larkin wrote:
>>> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>>>
>>> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.
>>
>> Wow, almost as good as an LH0063, circa 1980. ;)
>>
>> It's got higher bandwidth, for sure, and about the same slew rate, but
>> can't do 40V supplies.
>>
>> The LH4009, which arrived with the NS 1989 Linear Databook and was gone
>> by the 1992 one, had a maximum slew of 10kV/us, and produced +-250 mA
>>from 40V supplies.
>>
>> The extra bandwidth is useful, for sure, as long as the supply voltages
>> permit. ;)
>>

> TI is pretty good about making fast amps that work at nose-bleed
> voltages... like 10 volts.

The really dark days were about a dozen years ago, when there were no
new HV op amps at all. The OPA140 and OPA2188 are the bomb.

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 Tue, 7 Dec 2021 11:19:49 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>jlarkin@highlandsniptechnology.com wrote:
>> On Mon, 6 Dec 2021 22:09:07 -0500, Phil Hobbs
>> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>>
>>> John Larkin wrote:
>>>> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>>>>
>>>> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.
>>>
>>> Wow, almost as good as an LH0063, circa 1980. ;)
>>>
>>> It's got higher bandwidth, for sure, and about the same slew rate, but
>>> can't do 40V supplies.
>>>
>>> The LH4009, which arrived with the NS 1989 Linear Databook and was gone
>>> by the 1992 one, had a maximum slew of 10kV/us, and produced +-250 mA
>>>from 40V supplies.
>>>
>>> The extra bandwidth is useful, for sure, as long as the supply voltages
>>> permit. ;)
>>>
>
>> TI is pretty good about making fast amps that work at nose-bleed
>> voltages... like 10 volts.
>
>The really dark days were about a dozen years ago, when there were no
>new HV op amps at all. The OPA140 and OPA2188 are the bomb.
>
>Cheers
>
>Phil Hobbs

OPA552 is great.

OPA547 is 60 volts with programmable current limit.

ADA4522 is a 55 volt chopamp! Who needs a 55 volt chopamp?

I roll my own kilovolt-level opamps. That's fun.

--

Father Brown's figure remained quite dark and still;
but in that instant he had lost his head. His head was
always most valuable when he had lost it.

Phil Hobbs schrieb:
> John Larkin wrote:
>> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>>
>>
>> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.
>
> Wow, almost as good as an LH0063, circa 1980. ;)

Yes, that "damn fast" ;-) (later censored) buffer.

SCNR

Reinhard

On Monday, December 6, 2021 at 7:10:02 PM UTC-8, Phil Hobbs wrote:
> Wow, almost as good as an LH0063, circa 1980. ;)

ctrl-f noise
"Adobe Acrobat has finished searching the document. No matches were found."

Okey-dokey, then.

-- john, KE5FX

jlarkin@highlandsniptechnology.com wrote:
> On Tue, 7 Dec 2021 11:19:49 -0500, Phil Hobbs
> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>
>> jlarkin@highlandsniptechnology.com wrote:
>>> On Mon, 6 Dec 2021 22:09:07 -0500, Phil Hobbs
>>> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>>>
>>>> John Larkin wrote:
>>>>> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>>>>>
>>>>> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.
>>>>
>>>> Wow, almost as good as an LH0063, circa 1980. ;)
>>>>
>>>> It's got higher bandwidth, for sure, and about the same slew rate, but
>>>> can't do 40V supplies.
>>>>
>>>> The LH4009, which arrived with the NS 1989 Linear Databook and was gone
>>>> by the 1992 one, had a maximum slew of 10kV/us, and produced +-250 mA
>>> >from 40V supplies.
>>>>
>>>> The extra bandwidth is useful, for sure, as long as the supply voltages
>>>> permit. ;)
>>>>
>>
>>> TI is pretty good about making fast amps that work at nose-bleed
>>> voltages... like 10 volts.
>>
>> The really dark days were about a dozen years ago, when there were no
>> new HV op amps at all. The OPA140 and OPA2188 are the bomb.
>>
>> Cheers
>>
>> Phil Hobbs
>
>
> OPA552 is great.
>
> OPA547 is 60 volts with programmable current limit.
>
> ADA4522 is a 55 volt chopamp! Who needs a 55 volt chopamp?
>
> I roll my own kilovolt-level opamps. That's fun.

I'm doing a ridiculous APD front end for a customer--it uses our
tried-and-true pHEMT bootstrap, with probably a single-ended pHEMT/SiGe
cascode common-source gain stage with a noninverting CFA for the back
end, hooked up as a bootstrapped TIA with a bandwidth in the high VHF.

I can't get the slew rate I need using a unity-gain compensated VFA, and
the noise floor and supply headroom requirements have me blocked in
between 1k and 2k for the feedback resistor, so I can't use one of the
fast-slewing decompensated ones such as the OPA858 (A_V > 7).

They want it to recover from overload very very fast, so the APD voltage
has to be clamped during overload. Otherwise the bootstrap cap will see
almost the full DC current and start charging up to the wrong voltage.
It could easily take tens of milliseconds to recover from that.

The kicker is, all that has to happen floating on a several-hundred-volt
bias supply.

I suspect I may need to do one of your patented GaN FET tricks to drop
the bias until the overload is finished--at 450V, the poor APD will get
very hot very fast. That will decrease its gain sharply, and it could
easily take tens of seconds for the temperature to recover. So there's
some clarification and goal post moving required.

Fun.

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

John Miles, KE5FX wrote:
> On Monday, December 6, 2021 at 7:10:02 PM UTC-8, Phil Hobbs wrote:
>> Wow, almost as good as an LH0063, circa 1980. ;)
>
> ctrl-f noise
> "Adobe Acrobat has finished searching the document. No matches were found."
>
> Okey-dokey, then.
>
> -- john, KE5FX
>
Funny, it found it in your post. ;)

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 Tue, 7 Dec 2021 14:11:58 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>jlarkin@highlandsniptechnology.com wrote:
>> On Tue, 7 Dec 2021 11:19:49 -0500, Phil Hobbs
>> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>>
>>> jlarkin@highlandsniptechnology.com wrote:
>>>> On Mon, 6 Dec 2021 22:09:07 -0500, Phil Hobbs
>>>> <pcdhSpamMeSenseless@electrooptical.net> wrote:
>>>>
>>>>> John Larkin wrote:
>>>>>> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>>>>>>
>>>>>> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.
>>>>>
>>>>> Wow, almost as good as an LH0063, circa 1980. ;)
>>>>>
>>>>> It's got higher bandwidth, for sure, and about the same slew rate, but
>>>>> can't do 40V supplies.
>>>>>
>>>>> The LH4009, which arrived with the NS 1989 Linear Databook and was gone
>>>>> by the 1992 one, had a maximum slew of 10kV/us, and produced +-250 mA
>>>> >from 40V supplies.
>>>>>
>>>>> The extra bandwidth is useful, for sure, as long as the supply voltages
>>>>> permit. ;)
>>>>>
>>>
>>>> TI is pretty good about making fast amps that work at nose-bleed
>>>> voltages... like 10 volts.
>>>
>>> The really dark days were about a dozen years ago, when there were no
>>> new HV op amps at all. The OPA140 and OPA2188 are the bomb.
>>>
>>> Cheers
>>>
>>> Phil Hobbs
>>
>>
>> OPA552 is great.
>>
>> OPA547 is 60 volts with programmable current limit.
>>
>> ADA4522 is a 55 volt chopamp! Who needs a 55 volt chopamp?
>>
>> I roll my own kilovolt-level opamps. That's fun.
>
>
>I'm doing a ridiculous APD front end for a customer--it uses our
>tried-and-true pHEMT bootstrap, with probably a single-ended pHEMT/SiGe
>cascode common-source gain stage with a noninverting CFA for the back
>end, hooked up as a bootstrapped TIA with a bandwidth in the high VHF.
>
>I can't get the slew rate I need using a unity-gain compensated VFA, and
>the noise floor and supply headroom requirements have me blocked in
>between 1k and 2k for the feedback resistor, so I can't use one of the
>fast-slewing decompensated ones such as the OPA858 (A_V > 7).
>
>They want it to recover from overload very very fast, so the APD voltage
>has to be clamped during overload. Otherwise the bootstrap cap will see
>almost the full DC current and start charging up to the wrong voltage.
>It could easily take tens of milliseconds to recover from that.
>
>The kicker is, all that has to happen floating on a several-hundred-volt
>bias supply.
>
>I suspect I may need to do one of your patented GaN FET tricks to drop
>the bias until the overload is finished--at 450V, the poor APD will get
>very hot very fast. That will decrease its gain sharply, and it could
>easily take tens of seconds for the temperature to recover. So there's
>some clarification and goal post moving required.
>
>Fun.
>
>Cheers
>
>Phil Hobbs

We do have an unfair amount of fun doing this stuff, and get paid too.
Usually.

Email me the circuit if you'd like a review. I can do ridiculous.

--

If a man will begin with certainties, he shall end with doubts,
but if he will be content to begin with doubts he shall end in certainties.
Francis Bacon

On Tue, 7 Dec 2021 19:02:54 +0100, Reinhard Zwirner
<reinhard.zwirner@t-online.de> wrote:

>Phil Hobbs schrieb:
>> John Larkin wrote:
>>> https://www.ti.com/sitesearch/docs/universalsearch.tsp?langPref=en-US&searchTerm=buf802&nr=16#q=buf802&numberOfResults=25
>>>
>>>
>>> 3.1 GHz unity-gain jfet buffer. Noise ain't bad.
>>
>> Wow, almost as good as an LH0063, circa 1980. ;)
>
>Yes, that "damn fast" ;-) (later censored) buffer.
>
>SCNR
>
>Reinhard

http://www.repeater-builder.com/humor/pdfs/national-lh0033c-damn-fast.pdf

--
Best regards,
Spehro Pefhany