https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer capacitance 0.3 pF, cheap.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS. Assuming it’s quiet, of course—apart from extrinsic effects it should be about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

What I want is something

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

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer capacitance 0.3 pF, cheap.

Just looking at the photos it isn't obvious how to breadboard such a
package. Do they require a special socket?

Thanks for writing,

bob prohaska

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Fri, 20 Jun 2025 13:58:17 -0000 (UTC), [email protected] wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer
capacitance 0.3 pF, cheap.

Just looking at the photos it isn't obvious how to breadboard such a
package. Do they require a special socket?

Thanks for writing,

bob prohaska

I think I can Dremel a hunk of FR4 enough to start testing one, but it
will take a real proto PCB, with vias, to do it right. I have a summer
intern starting on Monday and it would be a good project for him.

One could solder small wires to it too, for basic DC tests.

I'm thinking about making a high-voltage pulse generator with this
part, with some sort of transmission-line transformer for polarity
control and isolation and maybe step-up.

As always, gate drive is the interesting part, and the low gate charge
sure helps.

If we do a proto PC board, we'll include some shear-off adapters.
Everybody who makes parts like this should offer adapters. Maybe
somebody sells something.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs <[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer
capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM >C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS. >Assuming it�s quiet, of course�apart from extrinsic effects it should be >about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide
rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make
VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 6/20/25 18:01, john larkin wrote:

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs <[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer
capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM
C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS.
Assuming it’s quiet, of course—apart from extrinsic effects it should be >> about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make
VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

0.0750 ns/div. So clumsy. Haven't they ever heard of picoseconds?

Jeroen Belleman

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Fri, 20 Jun 2025 21:59:58 +0200, Jeroen Belleman
<[email protected]> wrote:

On 6/20/25 18:01, john larkin wrote:

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs
<[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer
capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM >>> C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS.
Assuming it�s quiet, of course�apart from extrinsic effects it should be >>> about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide
rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make
VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

0.0750 ns/div. So clumsy. Haven't they ever heard of picoseconds?

Jeroen Belleman

ps means pounds sterling I think.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 21/06/2025 8:51 am, john larkin wrote:

On Fri, 20 Jun 2025 21:59:58 +0200, Jeroen Belleman
<[email protected]> wrote:

On 6/20/25 18:01, john larkin wrote:

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs
<[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer >>>>> capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM >>>> C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS. >>>> Assuming it’s quiet, of course—apart from extrinsic effects it should be
about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide
rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make
VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

0.0750 ns/div. So clumsy. Haven't they ever heard of picoseconds?

Jeroen Belleman

ps means pounds sterling I think.

Not anywhere I've lived. There is a special character for the job - £ - comparable with the $ symbol. Anyway, sensible people type nsec and psec.

--
Bill Sloman, Sydney


--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 20/06/2025 23:51, john larkin wrote:

On Fri, 20 Jun 2025 21:59:58 +0200, Jeroen Belleman
<[email protected]> wrote:

On 6/20/25 18:01, john larkin wrote:

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs
<[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer >>>>> capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM >>>> C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS. >>>> Assuming it’s quiet, of course—apart from extrinsic effects it should be
about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide
rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make
VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

0.0750 ns/div. So clumsy. Haven't they ever heard of picoseconds?

Jeroen Belleman

ps means pounds sterling I think.

gbp is the currency designation aka £ which is shift 3 on a UK keyboard.
$ is shift 4.


--
Martin Brown

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

"john larkin" <[email protected]> wrote in message news:[email protected]...

On Sat, 21 Jun 2025 23:45:45 +0100, Martin Brown
<'''newspam'''@nonad.co.uk> wrote:

On 20/06/2025 23:51, john larkin wrote:

On Fri, 20 Jun 2025 21:59:58 +0200, Jeroen Belleman
<[email protected]> wrote:

On 6/20/25 18:01, john larkin wrote:

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs
<[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer >>>>>>> capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM
C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS. >>>>>> Assuming it's quiet, of course-apart from extrinsic effects it should be >>>>>> about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide
rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make >>>>> VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

0.0750 ns/div. So clumsy. Haven't they ever heard of picoseconds?

Jeroen Belleman

ps means pounds sterling I think.

gbp is the currency designation aka � which is shift 3 on a UK keyboard.
$ is shift 4.

Shift 3 is # here, pronounced "pound sign."

I thought it was "number" in the States. As in Apartment #251
I've heard it called pound too, but no-one in the States has seen a �

A pound is of course a unit of weight. Or force. Or maybe mass.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Sat, 21 Jun 2025 23:45:45 +0100, Martin Brown
<'''newspam'''@nonad.co.uk> wrote:

On 20/06/2025 23:51, john larkin wrote:

On Fri, 20 Jun 2025 21:59:58 +0200, Jeroen Belleman
<[email protected]> wrote:

On 6/20/25 18:01, john larkin wrote:

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs
<[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer >>>>>> capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get. Analog switch FOM >>>>> C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200 mS. >>>>> Assuming it�s quiet, of course�apart from extrinsic effects it should be >>>>> about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide
rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make
VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

0.0750 ns/div. So clumsy. Haven't they ever heard of picoseconds?

Jeroen Belleman

ps means pounds sterling I think.

gbp is the currency designation aka � which is shift 3 on a UK keyboard.
$ is shift 4.

Shift 3 is # here, pronounced "pound sign."

A pound is of course a unit of weight. Or force. Or maybe mass.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 6/20/25 17:45, john larkin wrote:

On Fri, 20 Jun 2025 13:58:17 -0000 (UTC), [email protected] wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer
capacitance 0.3 pF, cheap.

Just looking at the photos it isn't obvious how to breadboard such a
package. Do they require a special socket?

Thanks for writing,

bob prohaska

I think I can Dremel a hunk of FR4 enough to start testing one, but it
will take a real proto PCB, with vias, to do it right. I have a summer
intern starting on Monday and it would be a good project for him.

One could solder small wires to it too, for basic DC tests.

I'm thinking about making a high-voltage pulse generator with this
part, with some sort of transmission-line transformer for polarity
control and isolation and maybe step-up.

As always, gate drive is the interesting part, and the low gate charge
sure helps.

If we do a proto PC board, we'll include some shear-off adapters.
Everybody who makes parts like this should offer adapters. Maybe
somebody sells something.

it would probably take two cups of coffee to design and order from
JLCPCB, something like $50 for 100 with a stencil delivered in a week

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 6/21/2025 6:45 PM, Martin Brown wrote:

On 20/06/2025 23:51, john larkin wrote:

On Fri, 20 Jun 2025 21:59:58 +0200, Jeroen Belleman
<[email protected]> wrote:

On 6/20/25 18:01, john larkin wrote:

On Fri, 20 Jun 2025 14:36:06 -0000 (UTC), Phil Hobbs
<[email protected]> wrote:

john larkin <[email protected]> wrote:

https://www.infineon.com/dgdl/Infineon-GS-065-004-6-L-TR-
DataSheet-v01_00-EN.pdf?fileId=8ac78c8c8e7ead30018ebd3e1aea2dc9

700 volts, turns on to 10 amps with +5 on the gate, reverse transfer >>>>>> capacitance 0.3 pF, cheap.

Yeah, $1.64 in hundreds, pretty good for what you get.  Analog
switch FOM
C_dg*R_on is about 1 ps, very impressive.

I want its little brother, with that sort of speed and a g_m of 200
mS.
Assuming it’s quiet, of course—apart from extrinsic effects it
should be
about sqrt(8/3 * kT/g_m) = 0.22 nV in 1 Hz.

The smaller EPC parts are good, but lower volts and amps.

Cheers

Phil Hobbs

Cheers

Phil Hobbs

If I park the drain at +700 and drive the gate from 0 to +5 with a
BUF602, I estimate a fall time of 1 ns.

Pretty good slew rate.

People are making such fun toys these days.

We spent yesterday at the MTTS microwave show downtown. All those
people are still stuck in the graph paper/slide
rule/s-parameter/VNA/load-pull world. I talked to a bunch of them
about using their parts in time domain, with Spice models, and got
vacant stares. Their world is all sine waves and dBm's.

I met one cool guy from a brit company, LA Techniques Ltd. They make
VNAs and he grabbed my proto RF switch board and VNA'd it. And,
amazingly, did a bunch of digital data eye diagrams. From the VNA
data.

https://www.dropbox.com/scl/fi/vsfe0bjm8a5ur7pp6fwms/
s21_eye_2gbps.png?rlkey=8p17rq5sttnd31mbyqttw2umj&raw=1

Hurts my head.

0.0750 ns/div. So clumsy. Haven't they ever heard of picoseconds?

Jeroen Belleman

ps means pounds sterling I think.

gbp is the currency designation aka £ which is shift 3 on a UK keyboard.
$ is shift 4.

I think some 1980s video game used "£" to represent its in-game currency
which was called "Ludder" so I've thought of UK currency as being called "Ludder" ever since

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)