fast breadboard, MC10EP08

On 7/17/19 8:21 PM, George Herold wrote:
On Friday, July 12, 2019 at 9:23:15 PM UTC-4, Phil Hobbs wrote:
On 7/12/19 7:55 PM, John Larkin wrote:


I was wondering how an MC10EP08 (4 GHz XOR gate) behaves with a slow
input, specifically sine waves from 1 MHz up until it quits. Old 10KH
type ECL gates made decent analog amplifiers, but the EP parts are
much faster and probably have more voltage gain, so I don't know what
they might do.

So I built this:

https://www.dropbox.com/s/dbuf1ixgfs8soc8/Z466_WB.JPG?raw=1

https://www.dropbox.com/s/9hmq0u1jv03oqgk/Z466_1.JPG?raw=1

https://www.dropbox.com/s/opnxfnk79o5lk1s/Z466_2.JPG?raw=1

Yes, it is barely possible to scrunch two SMA cables onto the output
connectors. Barely.

Now I have to pull rank on a junior staff member and swipe the big ole
Wayne Kerr RF signal generator and test it.

I could front this with an Analog Devices fast comparator, which has
hysteresis and differential ECL or CML outputs, but they are expensive
and need more power supplies.

I really need another RF signal generator.

A PTS1000 is super good medicine.

Your protos are so Hollywood. ;)
Yeah but imagine showing your proto made out of
low end copper clad to the same customer.

It's gold baby! :^)

John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference? I guess I'm thinking
TDR-wise. (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

George H.
Click to expand...

For CuClad protos that have to look nice, I spray both the board and box
with clear acrylic before building it. Acrylic is easy to solder
through, and stays nice looking for a long time.

I have a box fan in the window next to me while doing it, though!

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 18/7/19 10:54 am, John Larkin wrote:
On Wed, 17 Jul 2019 17:21:45 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:
John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference? I guess I'm thinking
TDR-wise. (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

I don't think I'd see any difference on a TDR, bare copper vs gold. I
think that most of the loss is on the bottom side of a microstrip
trace.
Click to expand...
Right. Surface flatness is the main issue (not conductivity), and a
micron of gold won't change that.
 
On Mon, 22 Jul 2019 09:39:21 +1000, Clifford Heath
<no.spam@please.net> wrote:

On 18/7/19 10:54 am, John Larkin wrote:
On Wed, 17 Jul 2019 17:21:45 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:
John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference? I guess I'm thinking
TDR-wise. (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

I don't think I'd see any difference on a TDR, bare copper vs gold. I
think that most of the loss is on the bottom side of a microstrip
trace.
Right. Surface flatness is the main issue (not conductivity), and a
micron of gold won't change that.
Click to expand...

The bottom side of a trace on a conventional PCB is a horror; peel one
up and see. Most of the e-field energy is on the bottom.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics
 
Am 22.07.19 um 01:39 schrieb Clifford Heath:
On 18/7/19 10:54 am, John Larkin wrote:
On Wed, 17 Jul 2019 17:21:45 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:
John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference?  I guess I'm thinking
TDR-wise.  (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

I don't think I'd see any difference on a TDR, bare copper vs gold. I
think that most of the loss is on the bottom side of a microstrip
trace.
Right. Surface flatness is the main issue (not conductivity), and a
micron of gold won't change that.
Click to expand...

but the Ni barrier between the Cu and Au can be a disaster
in microwave circuits.

cheers, Gerhard
 
On 22/7/19 9:59 am, John Larkin wrote:
On Mon, 22 Jul 2019 09:39:21 +1000, Clifford Heath
no.spam@please.net> wrote:

On 18/7/19 10:54 am, John Larkin wrote:
On Wed, 17 Jul 2019 17:21:45 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:
John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference? I guess I'm thinking
TDR-wise. (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

I don't think I'd see any difference on a TDR, bare copper vs gold. I
think that most of the loss is on the bottom side of a microstrip
trace.
Right. Surface flatness is the main issue (not conductivity), and a
micron of gold won't change that.

The bottom side of a trace on a conventional PCB is a horror; peel one
up and see.
Click to expand...

It's deliberate, too, for reasons of adhesion.
 
On Mon, 22 Jul 2019 02:24:46 +0200, Gerhard Hoffmann <dk4xp@arcor.de>
wrote:

Am 22.07.19 um 01:39 schrieb Clifford Heath:
On 18/7/19 10:54 am, John Larkin wrote:
On Wed, 17 Jul 2019 17:21:45 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:
John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference?  I guess I'm thinking
TDR-wise.  (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

I don't think I'd see any difference on a TDR, bare copper vs gold. I
think that most of the loss is on the bottom side of a microstrip
trace.
Right. Surface flatness is the main issue (not conductivity), and a
micron of gold won't change that.

but the Ni barrier between the Cu and Au can be a disaster
in microwave circuits.

cheers, Gerhard
Click to expand...

At high power? Are there low power issues?


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
 
Am 22.07.19 um 20:56 schrieb John Larkin:
On Mon, 22 Jul 2019 02:24:46 +0200, Gerhard Hoffmann <dk4xp@arcor.de
wrote:

Am 22.07.19 um 01:39 schrieb Clifford Heath:
On 18/7/19 10:54 am, John Larkin wrote:
On Wed, 17 Jul 2019 17:21:45 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:
John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference?  I guess I'm thinking
TDR-wise.  (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

I don't think I'd see any difference on a TDR, bare copper vs gold. I
think that most of the loss is on the bottom side of a microstrip
trace.
Right. Surface flatness is the main issue (not conductivity), and a
micron of gold won't change that.

but the Ni barrier between the Cu and Au can be a disaster
in microwave circuits.

cheers, Gerhard


At high power? Are there low power issues?
Click to expand...

Frequency dependent losses. Kills the Q, forget about stripline filters.

I have made small stamp size boards for 10EP & friends.
10EP has a consistent pinout. One side is inputs, the other side is
outputs. Power, Vbb etc are on the same locations.
The input side has Thevenin terminations, there is the standard
decoupling and the bottom side is GND only.
With 2 or three stamps, you have the entire family.

Twisted AWG30 wrapping wire has about 100 Ohms, ok for wiring
differential between the stamps. I just solder the stamps onto a
clean FR4 board. I admit, gold has more style.
I also have 1mm HF-50 coax cable.
That makes test circuits quite fast & easy.

BTW I wonder why you did choose a XOR gate. That is 2-level logic.
An or/nor/nand is probably faster.

Cheers, Gerhard
 
On Mon, 22 Jul 2019 23:23:31 +0200, Gerhard Hoffmann <dk4xp@arcor.de>
wrote:

Am 22.07.19 um 20:56 schrieb John Larkin:
On Mon, 22 Jul 2019 02:24:46 +0200, Gerhard Hoffmann <dk4xp@arcor.de
wrote:

Am 22.07.19 um 01:39 schrieb Clifford Heath:
On 18/7/19 10:54 am, John Larkin wrote:
On Wed, 17 Jul 2019 17:21:45 -0700 (PDT), George Herold
gherold@teachspin.com> wrote:
John L, If you make some sort of micro strip
out of copper clad and copper clad with gold on top,
can you tell the difference?  I guess I'm thinking
TDR-wise.  (Of which I know almost nothing... if you made
a video/ white broad thing about TDR I would watch it,
and recommend it to my all my buds on SED. :^)

I don't think I'd see any difference on a TDR, bare copper vs gold. I
think that most of the loss is on the bottom side of a microstrip
trace.
Right. Surface flatness is the main issue (not conductivity), and a
micron of gold won't change that.

but the Ni barrier between the Cu and Au can be a disaster
in microwave circuits.

cheers, Gerhard


At high power? Are there low power issues?


Frequency dependent losses. Kills the Q, forget about stripline filters.
Click to expand...

I think our ENIG boards all have bare copper traces under the solder
mask. Only the pads are plated. Solder dissolves the gold but (may?)
leave the nickel there.

I have made small stamp size boards for 10EP & friends.
10EP has a consistent pinout. One side is inputs, the other side is
outputs. Power, Vbb etc are on the same locations.
The input side has Thevenin terminations, there is the standard
decoupling and the bottom side is GND only.
With 2 or three stamps, you have the entire family.

Twisted AWG30 wrapping wire has about 100 Ohms, ok for wiring
differential between the stamps. I just solder the stamps onto a
clean FR4 board. I admit, gold has more style.
I also have 1mm HF-50 coax cable.
That makes test circuits quite fast & easy.

BTW I wonder why you did choose a XOR gate. That is 2-level logic.
An or/nor/nand is probably faster.
Click to expand...

The EP08 is rated for 3 GHz, which is a bit faster than many other EP
parts. And we have it in stock and in the PADS library. I can play
with input levels and use it as a mux, too. It's faster than the EP57
mux.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top