Conical inductors--still $10!...

[Tom Gardner]

On 22/07/20 04:01, Phil Hobbs wrote:

On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
You don\'t need to be able to design circuits to get an EE job at most large
companies as a newly-minted EE, seems like it\'s expected that\'s one of the
things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too.  With my new astronomy and physics bachelor\'s degree, and a hobby
background in electronics. I got hired to do 2/3 of the timing and frequency
control electronics for the first civilian direct-broadcast satellite system.
I\'d heard of PLLs but had never actually come across one to know what it was.
Talk about drinking from a fire hose.

My first job involved that newfangled digital logic; easy.

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

 

[Tom Gardner]

On 22/07/20 04:01, Phil Hobbs wrote:

On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
You don\'t need to be able to design circuits to get an EE job at most large
companies as a newly-minted EE, seems like it\'s expected that\'s one of the
things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too.  With my new astronomy and physics bachelor\'s degree, and a hobby
background in electronics. I got hired to do 2/3 of the timing and frequency
control electronics for the first civilian direct-broadcast satellite system.
I\'d heard of PLLs but had never actually come across one to know what it was.
Talk about drinking from a fire hose.

My first job involved that newfangled digital logic; easy.

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

 

[Phil Hobbs]

On 2020-07-22 02:57, Tom Gardner wrote:

On 22/07/20 04:01, Phil Hobbs wrote:
On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
You don\'t need to be able to design circuits to get an EE job at
most large companies as a newly-minted EE, seems like it\'s expected
that\'s one of the things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too.  With my new astronomy and physics bachelor\'s degree, and a
hobby background in electronics. I got hired to do 2/3 of the timing
and frequency control electronics for the first civilian
direct-broadcast satellite system. I\'d heard of PLLs but had never
actually come across one to know what it was. Talk about drinking from
a fire hose.

My first job involved that newfangled digital logic; easy.

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I often use TIAs with two outputs, one low and one high gain. One
approach is to hang an op amp on each end of the PD, and clamp the
high-gain one so it doesn\'t lose control of its summing junction at
large signals. You can apply bias between the + inputs of the op amps
and let feedback force that across the PD. Needs a diff amp on one
side, of course.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

The Tayloe thing was actually published by Ed Oxner of Siliconix some
years earlier, so don\'t feel so bad.

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

 

[Tom Gardner]

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:

On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird stuff,
TTL) were horrible.

TTL was blissfully simple compared with the others!

No, I haven\'t dealt with magnetic logic (although
the first computer I used in anger did).

Yes, I have looked at hydraulic logic running at
2000psi. It was a design study for replacing it
with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it
was on an offshore oil rig with zero electricity
due to the inflammable atmosphere

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent them and
get rich.

I can\'t complain. I didn\'t follow up the concept for
20+ years!

 

[Tom Gardner]

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:

On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird stuff,
TTL) were horrible.

TTL was blissfully simple compared with the others!

No, I haven\'t dealt with magnetic logic (although
the first computer I used in anger did).

Yes, I have looked at hydraulic logic running at
2000psi. It was a design study for replacing it
with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it
was on an offshore oil rig with zero electricity
due to the inflammable atmosphere

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent them and
get rich.

I can\'t complain. I didn\'t follow up the concept for
20+ years!

 

[Tom Gardner]

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:

On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird stuff,
TTL) were horrible.

TTL was blissfully simple compared with the others!

No, I haven\'t dealt with magnetic logic (although
the first computer I used in anger did).

Yes, I have looked at hydraulic logic running at
2000psi. It was a design study for replacing it
with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it
was on an offshore oil rig with zero electricity
due to the inflammable atmosphere

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent them and
get rich.

I can\'t complain. I didn\'t follow up the concept for
20+ years!

 

[John Larkin]

On Wed, 22 Jul 2020 18:52:46 +0100, Tom Gardner
<spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:
On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird stuff,
TTL) were horrible.

TTL was blissfully simple compared with the others!

Agree, but the early plastic DIPs failed a lot.

No, I haven\'t dealt with magnetic logic (although
the first computer I used in anger did).

Yes, I have looked at hydraulic logic running at
2000psi. It was a design study for replacing it
with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it
was on an offshore oil rig with zero electricity
due to the inflammable atmosphere


It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent them and
get rich.

I can\'t complain. I didn\'t follow up the concept for
20+ years!

I invented the dual-slope ADC when I was a kid, but I figured that
relay bounce would make it inaccurate. Relays were the only analog
switches I could imagine then.

I also invented the successive log RF detector, but my boss thought it
was useless.

 

[Sjouke Burry]

On 22.07.20 20:19, John Larkin wrote:

On Wed, 22 Jul 2020 18:52:46 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:
On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird stuff,
TTL) were horrible.

TTL was blissfully simple compared with the others!

Agree, but the early plastic DIPs failed a lot.


No, I haven\'t dealt with magnetic logic (although
the first computer I used in anger did).

Yes, I have looked at hydraulic logic running at
2000psi. It was a design study for replacing it
with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it
was on an offshore oil rig with zero electricity
due to the inflammable atmosphere


It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent them and
get rich.

I can\'t complain. I didn\'t follow up the concept for
20+ years!

I invented the dual-slope ADC when I was a kid, but I figured that
relay bounce would make it inaccurate. Relays were the only analog
switches I could imagine then.

I also invented the successive log RF detector, but my boss thought it
was useless.

In about 25 years of ttl, I have not had any chip failure.
( Lab equipment, computer interfaces, experiment controls,
etc.)
Computers with ttl abandoned, but never failed.

 

[Phil Hobbs]

On 2020-07-22 14:19, John Larkin wrote:

On Wed, 22 Jul 2020 18:52:46 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:
On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird
stuff, TTL) were horrible.

TTL was blissfully simple compared with the others!

Agree, but the early plastic DIPs failed a lot.


No, I haven\'t dealt with magnetic logic (although the first
computer I used in anger did).

Yes, I have looked at hydraulic logic running at 2000psi. It was a
design study for replacing it with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it was on an offshore
oil rig with zero electricity due to the inflammable atmosphere


It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed between
exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB optical)
dynamic range, using a large photodiode (BPW34) and a LF351
based transimpedance amp. To recover the signal I decided I
couldn\'t predict how a PLL would work in a range switched
design, so I made a filter with a Q of 4000 using 10%
components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked, I
found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent
them and get rich.

I can\'t complain. I didn\'t follow up the concept for 20+ years!

I invented the dual-slope ADC when I was a kid, but I figured that
relay bounce would make it inaccurate. Relays were the only analog
switches I could imagine then.

Well, spark gaps might have worked.

I also invented the successive log RF detector, but my boss thought
it was useless.

Successive-detection DLVAs are cool. It\'s fun explaining the idea to a
bright youngster--you can see the light go on.

I invented the modulation-generated carrier approach to phase-sensitive
interferometry a couple of years after somebody else published it. You
FM a diode laser with a modulation index of 2.405 where the carrier
disappears, and detect Q at f_mod and I at 2f_mod. The J1 and J2 terms
are the same strength (or very nearly) and together account for about
85% of the signal power. Very pretty for things with unstable phases
such as fibre interferometers, for instance.

My fanciest PLL to date was from 1982--it controlled a VCXO by using a
fractional-N synth made out of a regular dual-modulus prescaler, two
synchronously cascaded 74S161 decade counters, and two decades of 4527
rate multipliers twiddling the carry input of the bottom counter.

It made a fair amount of jitter, of course, but it was all well outside
the loop bandwidth and so it all worked fine. Increasing the comparison
frequency from 9.09 kHz to 909 kHz improved the phase noise on the 14
GHz output a fair amount.

(When I started that job I had no idea how demanding some communications
applications can be.)

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 2020-07-22 14:19, John Larkin wrote:

On Wed, 22 Jul 2020 18:52:46 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:
On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird
stuff, TTL) were horrible.

TTL was blissfully simple compared with the others!

Agree, but the early plastic DIPs failed a lot.


No, I haven\'t dealt with magnetic logic (although the first
computer I used in anger did).

Yes, I have looked at hydraulic logic running at 2000psi. It was a
design study for replacing it with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it was on an offshore
oil rig with zero electricity due to the inflammable atmosphere


It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed between
exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB optical)
dynamic range, using a large photodiode (BPW34) and a LF351
based transimpedance amp. To recover the signal I decided I
couldn\'t predict how a PLL would work in a range switched
design, so I made a filter with a Q of 4000 using 10%
components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked, I
found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent
them and get rich.

I can\'t complain. I didn\'t follow up the concept for 20+ years!

I invented the dual-slope ADC when I was a kid, but I figured that
relay bounce would make it inaccurate. Relays were the only analog
switches I could imagine then.

Well, spark gaps might have worked.

I also invented the successive log RF detector, but my boss thought
it was useless.

Successive-detection DLVAs are cool. It\'s fun explaining the idea to a
bright youngster--you can see the light go on.

I invented the modulation-generated carrier approach to phase-sensitive
interferometry a couple of years after somebody else published it. You
FM a diode laser with a modulation index of 2.405 where the carrier
disappears, and detect Q at f_mod and I at 2f_mod. The J1 and J2 terms
are the same strength (or very nearly) and together account for about
85% of the signal power. Very pretty for things with unstable phases
such as fibre interferometers, for instance.

My fanciest PLL to date was from 1982--it controlled a VCXO by using a
fractional-N synth made out of a regular dual-modulus prescaler, two
synchronously cascaded 74S161 decade counters, and two decades of 4527
rate multipliers twiddling the carry input of the bottom counter.

It made a fair amount of jitter, of course, but it was all well outside
the loop bandwidth and so it all worked fine. Increasing the comparison
frequency from 9.09 kHz to 909 kHz improved the phase noise on the 14
GHz output a fair amount.

(When I started that job I had no idea how demanding some communications
applications can be.)

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 2020-07-22 14:19, John Larkin wrote:

On Wed, 22 Jul 2020 18:52:46 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:
On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird
stuff, TTL) were horrible.

TTL was blissfully simple compared with the others!

Agree, but the early plastic DIPs failed a lot.


No, I haven\'t dealt with magnetic logic (although the first
computer I used in anger did).

Yes, I have looked at hydraulic logic running at 2000psi. It was a
design study for replacing it with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it was on an offshore
oil rig with zero electricity due to the inflammable atmosphere


It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed between
exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB optical)
dynamic range, using a large photodiode (BPW34) and a LF351
based transimpedance amp. To recover the signal I decided I
couldn\'t predict how a PLL would work in a range switched
design, so I made a filter with a Q of 4000 using 10%
components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked, I
found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent
them and get rich.

I can\'t complain. I didn\'t follow up the concept for 20+ years!

I invented the dual-slope ADC when I was a kid, but I figured that
relay bounce would make it inaccurate. Relays were the only analog
switches I could imagine then.

Well, spark gaps might have worked.

I also invented the successive log RF detector, but my boss thought
it was useless.

Successive-detection DLVAs are cool. It\'s fun explaining the idea to a
bright youngster--you can see the light go on.

I invented the modulation-generated carrier approach to phase-sensitive
interferometry a couple of years after somebody else published it. You
FM a diode laser with a modulation index of 2.405 where the carrier
disappears, and detect Q at f_mod and I at 2f_mod. The J1 and J2 terms
are the same strength (or very nearly) and together account for about
85% of the signal power. Very pretty for things with unstable phases
such as fibre interferometers, for instance.

My fanciest PLL to date was from 1982--it controlled a VCXO by using a
fractional-N synth made out of a regular dual-modulus prescaler, two
synchronously cascaded 74S161 decade counters, and two decades of 4527
rate multipliers twiddling the carry input of the bottom counter.

It made a fair amount of jitter, of course, but it was all well outside
the loop bandwidth and so it all worked fine. Increasing the comparison
frequency from 9.09 kHz to 909 kHz improved the phase noise on the 14
GHz output a fair amount.

(When I started that job I had no idea how demanding some communications
applications can be.)

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

 

[Commander Kinsey]

On Tue, 14 Jul 2020 02:01:11 +0100, Brian Gregory <void-invalid-dead-dontuse@email.invalid> wrote:

On 12/07/2020 13:10, Commander Kinsey wrote:
Why are CPUs only about 80W TDP? Can\'t they make ones with three times
as many cores that have 250W TDP like graphics cards?

A lower clock speed it fine for a GPU since you can compensate by having
the equivalent of lots of cores. Thus it\'s easy to make a nice big hot
chip for a GPU.

In a CPU things need to be more precision, we need to be able to boost
any core up to a high clock speed because single core performance is
important in a CPU. Therefore a big chip with lots of cores would have a
low yield and be expensive to produce. That\'s way AMD invented a way to
do it with multiple smaller chip units for their most powerful processors.

So a 64 core AMD is actually 8 8 core CPUs?

 

[Commander Kinsey]

On Tue, 14 Jul 2020 02:01:11 +0100, Brian Gregory <void-invalid-dead-dontuse@email.invalid> wrote:

On 12/07/2020 13:10, Commander Kinsey wrote:
Why are CPUs only about 80W TDP? Can\'t they make ones with three times
as many cores that have 250W TDP like graphics cards?

A lower clock speed it fine for a GPU since you can compensate by having
the equivalent of lots of cores. Thus it\'s easy to make a nice big hot
chip for a GPU.

In a CPU things need to be more precision, we need to be able to boost
any core up to a high clock speed because single core performance is
important in a CPU. Therefore a big chip with lots of cores would have a
low yield and be expensive to produce. That\'s way AMD invented a way to
do it with multiple smaller chip units for their most powerful processors.

So a 64 core AMD is actually 8 8 core CPUs?

 

[John Larkin]

On Wed, 22 Jul 2020 15:56:03 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-22 14:19, John Larkin wrote:
On Wed, 22 Jul 2020 18:52:46 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:
On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird
stuff, TTL) were horrible.

TTL was blissfully simple compared with the others!

Agree, but the early plastic DIPs failed a lot.


No, I haven\'t dealt with magnetic logic (although the first
computer I used in anger did).

Yes, I have looked at hydraulic logic running at 2000psi. It was a
design study for replacing it with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it was on an offshore
oil rig with zero electricity due to the inflammable atmosphere


It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed between
exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB optical)
dynamic range, using a large photodiode (BPW34) and a LF351
based transimpedance amp. To recover the signal I decided I
couldn\'t predict how a PLL would work in a range switched
design, so I made a filter with a Q of 4000 using 10%
components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked, I
found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent
them and get rich.

I can\'t complain. I didn\'t follow up the concept for 20+ years!

I invented the dual-slope ADC when I was a kid, but I figured that
relay bounce would make it inaccurate. Relays were the only analog
switches I could imagine then.

Well, spark gaps might have worked.


I also invented the successive log RF detector, but my boss thought
it was useless.

Successive-detection DLVAs are cool. It\'s fun explaining the idea to a
bright youngster--you can see the light go on.

I invented the modulation-generated carrier approach to phase-sensitive
interferometry a couple of years after somebody else published it. You
FM a diode laser with a modulation index of 2.405 where the carrier
disappears, and detect Q at f_mod and I at 2f_mod. The J1 and J2 terms
are the same strength (or very nearly) and together account for about
85% of the signal power. Very pretty for things with unstable phases
such as fibre interferometers, for instance.

Those Bessel things are cool. You can set FM deviation really close by
disappearing the carrier or some sidebands.

My fanciest PLL to date was from 1982--it controlled a VCXO by using a
fractional-N synth made out of a regular dual-modulus prescaler, two
synchronously cascaded 74S161 decade counters, and two decades of 4527
rate multipliers twiddling the carry input of the bottom counter.

It made a fair amount of jitter, of course, but it was all well outside
the loop bandwidth and so it all worked fine. Increasing the comparison
frequency from 9.09 kHz to 909 kHz improved the phase noise on the 14
GHz output a fair amount.

(When I started that job I had no idea how demanding some communications
applications can be.)

Cheers

Phil Hobbs

PLLs are fun. My latest one triggers an LC oscillator from an external
input, and phase-locks it, at its natural frequency, to an OCXO within
a microsecond or so. It still hurts my head to think about that. I did
the analog stuff and delegated the FPGA and the hard math.

A good OCXO has a few picoseconds of RMS jitter, timing out a 1-second
delay. A cheap XO has many nanoseconds.

Bang-bang PLLs are interesting too, with a d-flop as the phase
detector. I invented that in my youth too, for an analog time-slot
telephone system. They also hurt my head.

 

[John Larkin]

On Wed, 22 Jul 2020 15:56:03 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-22 14:19, John Larkin wrote:
On Wed, 22 Jul 2020 18:52:46 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 15:20, jlarkin@highlandsniptechnology.com wrote:
On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:
My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird
stuff, TTL) were horrible.

TTL was blissfully simple compared with the others!

Agree, but the early plastic DIPs failed a lot.


No, I haven\'t dealt with magnetic logic (although the first
computer I used in anger did).

Yes, I have looked at hydraulic logic running at 2000psi. It was a
design study for replacing it with the newfangled micros.

I concluded it wasn\'t worth the aggro, since it was on an offshore
oil rig with zero electricity due to the inflammable atmosphere


It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed between
exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB optical)
dynamic range, using a large photodiode (BPW34) and a LF351
based transimpedance amp. To recover the signal I decided I
couldn\'t predict how a PLL would work in a range switched
design, so I made a filter with a Q of 4000 using 10%
components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked, I
found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent
them and get rich.

I can\'t complain. I didn\'t follow up the concept for 20+ years!

I invented the dual-slope ADC when I was a kid, but I figured that
relay bounce would make it inaccurate. Relays were the only analog
switches I could imagine then.

Well, spark gaps might have worked.


I also invented the successive log RF detector, but my boss thought
it was useless.

Successive-detection DLVAs are cool. It\'s fun explaining the idea to a
bright youngster--you can see the light go on.

I invented the modulation-generated carrier approach to phase-sensitive
interferometry a couple of years after somebody else published it. You
FM a diode laser with a modulation index of 2.405 where the carrier
disappears, and detect Q at f_mod and I at 2f_mod. The J1 and J2 terms
are the same strength (or very nearly) and together account for about
85% of the signal power. Very pretty for things with unstable phases
such as fibre interferometers, for instance.

Those Bessel things are cool. You can set FM deviation really close by
disappearing the carrier or some sidebands.

My fanciest PLL to date was from 1982--it controlled a VCXO by using a
fractional-N synth made out of a regular dual-modulus prescaler, two
synchronously cascaded 74S161 decade counters, and two decades of 4527
rate multipliers twiddling the carry input of the bottom counter.

It made a fair amount of jitter, of course, but it was all well outside
the loop bandwidth and so it all worked fine. Increasing the comparison
frequency from 9.09 kHz to 909 kHz improved the phase noise on the 14
GHz output a fair amount.

(When I started that job I had no idea how demanding some communications
applications can be.)

Cheers

Phil Hobbs

PLLs are fun. My latest one triggers an LC oscillator from an external
input, and phase-locks it, at its natural frequency, to an OCXO within
a microsecond or so. It still hurts my head to think about that. I did
the analog stuff and delegated the FPGA and the hard math.

A good OCXO has a few picoseconds of RMS jitter, timing out a 1-second
delay. A cheap XO has many nanoseconds.

Bang-bang PLLs are interesting too, with a d-flop as the phase
detector. I invented that in my youth too, for an analog time-slot
telephone system. They also hurt my head.

 

[server]

On Tue, 21 Jul 2020 23:01:02 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
On 7/20/2020 12:52 PM, Tom Gardner wrote:
On 20/07/20 16:22, jlarkin@highlandsniptechnology.com wrote:
On Mon, 20 Jul 2020 15:29:29 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 20/07/20 14:52, jlarkin@highlandsniptechnology.com wrote:
On Mon, 20 Jul 2020 08:42:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 20/07/20 01:52, Joe Gwinn wrote:
On Sat, 18 Jul 2020 08:29:35 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 18/07/20 00:23, Joe Gwinn wrote:
On Fri, 17 Jul 2020 01:32:33 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 21:42, Joe Gwinn wrote:
On Thu, 16 Jul 2020 15:54:54 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 15:32, jlarkin@highlandsniptechnology.com wrote:
On Thu, 16 Jul 2020 14:42:56 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 14:26, jlarkin@highlandsniptechnology.com wrote:
On Thu, 16 Jul 2020 07:28:17 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 06:32, dagmargoodboat@yahoo.com wrote:
E.g., why would someone on the dole ever work for
something they
already get free?  It doesn\'t make any sense.

True for some people, false for many more.

Many people feel defined by their work, and feel
pointless without it. Such people have a tendency
to \"give up and die\"  relatively shortly after
retiring.

You seem to understand Theory X companies, but
have no clue about Theory Y companies, as described
by McGregor in the 1950s.

Long before McGregor, Hewlett and Packard knew the
difference instinctively, and created a rather
successful Theory Y company. You may have heard
of it.

\"Theory Y managers assume employees are internally
motivated, enjoy their job, and work to better
themselves without a direct reward in return. These
managers view their employees as one of the most
valuable assets to the company, driving the internal
workings of the corporation. Employees additionally
tend to take full responsibility for their work and
do not need close supervision to create a quality
product.\"
https://en.wikipedia.org/wiki/Theory_X_and_Theory_Y

Or, as famously noted at the time of Princess Fiorina,
http://www.satirewire.com/news/0105/loyal.shtml

Sure, but a company doesn\'t become a Y just with a
policy statement.
It requires finding and hiring the right workers,
treating them right,
and firing the ones that don\'t work out.

It didn\'t cross my mind anybody could think mere
policy statements could be sufficient.

In HP, the HP Way was continually reinforced and
re-explained by use of Bill and Dave anecdotes,
wheeled out to show how they thought and wanted
things to be done. Apparently when they were setting
up new sites the first hires became a little sick
and tired of them!

OTOH, Princess Fiorina made very animated policy
pronouncements, which nobody could understand.
That\'s one of the things that made me decide
to leave.

I have Packard\'s book, The HP Way. And I have Fiorina\'s
book, The
Journey. The contrast is hilarious.

Not if you were in HP!


HP did that early on. By about 1980, not so well.

HP was /very/ careful about its hiring process, at least
until shortly before Fiorina ascended in 1999.

I interviewed at HP in about 1980. The guy was obnoxious.
He would
have been my boss.

He looked at my resume and said \"The first thing you need
to do is
decide if you are an engineer or a programmer.\"

What I decided to do was walk out.

Snap!

I had an interview at a GEC site in ~1981. After explaining
the hardware and software and systems I had designed, the
HRdroid asked me whether I was \"really a hardware of software
engineer\".

Somewhat surprisingly, I managed not to give him an earful.
I suspect the expression on my face and my answers becoming
terser might have alerted him to his faux pas. The idiot still
offered me a job.

I have a similar story from the 1970s, but it turned out rather
better.

I was applying to a middle size defense contractor in the
Baltimore
suburbs, and the hiring manager looked over my resume, and
asked which
I preferred, hardware or software.  I replied that it was
very useful
to be bilingual, to be able to speak hardware to software
and vice
versa.

A very sensible response of course.

\"My\" GEC HRdroid couldn\'t comprehend anything beyond
square holes, and all round candidates has to be
force fitted into one of the square holes.

If the interviewer asks questions but listens to the
answers and avoids such destructive idiocies, that\'s
just fine.

One technique I developed was to ask ever wilder
questions, with the objective of getting them to
(sensibly) say \"no\". That gave me good insight into
the validity of their \"yes\" responses.

This would be for interviewing, versus being interviewed?

I was thinking of being interviewed. After starting my
second job (at a contract design and consultancy company)
they told me that I asked far more questions than most
candidates. I haven\'t stopped since

I haven\'t found it necessary when I\'m the technical
interviewer; there are other more fun and fruitful way
of smoking out blatherers.

What would that be?

Nothing magic.

I get them to describe what they have done in the past,
listen (unlike a few interviewers!), ask them why they
made their choices, and what they would do differently
next time.

I take them to a whiteboard and design something with them.

So do I, either directly or indirectly. That will be
something that is adjacent to their experience, and/or
something that illustrates how the company thinks/works.

The key word is \"with\"; interviews aren\'t (or at least
shouldn\'t be) willy waving competitions.


One thing you can\'t tell from a resume, or the usual interview, is if
they understand electricity. Lots of presumed engineers don\'t. Most
recent EE grads don\'t.

Wanna make a kid panic? Show him a 2-resistor voltage divider. Even
worse, a batery and a resistor charging a cap.

That\'s one reason why I like to see evidence on their
CV that they have done things on their own, without
it being any part of any course.

Doesn\'t matter what it is, so long as they can describe
what, why, and what they would do differently next time.



You don\'t need to be able to design circuits to get an EE job at most
large companies as a newly-minted EE, seems like it\'s expected that\'s
one of the things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too. With my new astronomy and physics bachelor\'s degree, and a
hobby background in electronics. I got hired to do 2/3 of the timing and
frequency control electronics for the first civilian direct-broadcast
satellite system. I\'d heard of PLLs but had never actually come across
one to know what it was. Talk about drinking from a fire hose.

The times I\'ve made the most progress in my life are the times I jumped
off a cliff--that satcom job, getting married, going off to grad school
in another country with a family to support, yada yada.

It\'s been a good ride so far.

Cheers

Phil Hobbs

I was lucky to get a college-freshman job with a company that had no
coherent business plan or marketing. They would take an order for
anything and assume the engineers could do it. Learned a lot.

They paid me $400 a month, which was a fortune in New Orleans back
then.

Good ride indeed. Electronics is universal and gets us into all sorts
on interesting situations.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[server]

On Tue, 21 Jul 2020 23:01:02 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
On 7/20/2020 12:52 PM, Tom Gardner wrote:
On 20/07/20 16:22, jlarkin@highlandsniptechnology.com wrote:
On Mon, 20 Jul 2020 15:29:29 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 20/07/20 14:52, jlarkin@highlandsniptechnology.com wrote:
On Mon, 20 Jul 2020 08:42:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 20/07/20 01:52, Joe Gwinn wrote:
On Sat, 18 Jul 2020 08:29:35 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 18/07/20 00:23, Joe Gwinn wrote:
On Fri, 17 Jul 2020 01:32:33 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 21:42, Joe Gwinn wrote:
On Thu, 16 Jul 2020 15:54:54 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 15:32, jlarkin@highlandsniptechnology.com wrote:
On Thu, 16 Jul 2020 14:42:56 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 14:26, jlarkin@highlandsniptechnology.com wrote:
On Thu, 16 Jul 2020 07:28:17 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 06:32, dagmargoodboat@yahoo.com wrote:
E.g., why would someone on the dole ever work for
something they
already get free?  It doesn\'t make any sense.

True for some people, false for many more.

Many people feel defined by their work, and feel
pointless without it. Such people have a tendency
to \"give up and die\"  relatively shortly after
retiring.

You seem to understand Theory X companies, but
have no clue about Theory Y companies, as described
by McGregor in the 1950s.

Long before McGregor, Hewlett and Packard knew the
difference instinctively, and created a rather
successful Theory Y company. You may have heard
of it.

\"Theory Y managers assume employees are internally
motivated, enjoy their job, and work to better
themselves without a direct reward in return. These
managers view their employees as one of the most
valuable assets to the company, driving the internal
workings of the corporation. Employees additionally
tend to take full responsibility for their work and
do not need close supervision to create a quality
product.\"
https://en.wikipedia.org/wiki/Theory_X_and_Theory_Y

Or, as famously noted at the time of Princess Fiorina,
http://www.satirewire.com/news/0105/loyal.shtml

Sure, but a company doesn\'t become a Y just with a
policy statement.
It requires finding and hiring the right workers,
treating them right,
and firing the ones that don\'t work out.

It didn\'t cross my mind anybody could think mere
policy statements could be sufficient.

In HP, the HP Way was continually reinforced and
re-explained by use of Bill and Dave anecdotes,
wheeled out to show how they thought and wanted
things to be done. Apparently when they were setting
up new sites the first hires became a little sick
and tired of them!

OTOH, Princess Fiorina made very animated policy
pronouncements, which nobody could understand.
That\'s one of the things that made me decide
to leave.

I have Packard\'s book, The HP Way. And I have Fiorina\'s
book, The
Journey. The contrast is hilarious.

Not if you were in HP!


HP did that early on. By about 1980, not so well.

HP was /very/ careful about its hiring process, at least
until shortly before Fiorina ascended in 1999.

I interviewed at HP in about 1980. The guy was obnoxious.
He would
have been my boss.

He looked at my resume and said \"The first thing you need
to do is
decide if you are an engineer or a programmer.\"

What I decided to do was walk out.

Snap!

I had an interview at a GEC site in ~1981. After explaining
the hardware and software and systems I had designed, the
HRdroid asked me whether I was \"really a hardware of software
engineer\".

Somewhat surprisingly, I managed not to give him an earful.
I suspect the expression on my face and my answers becoming
terser might have alerted him to his faux pas. The idiot still
offered me a job.

I have a similar story from the 1970s, but it turned out rather
better.

I was applying to a middle size defense contractor in the
Baltimore
suburbs, and the hiring manager looked over my resume, and
asked which
I preferred, hardware or software.  I replied that it was
very useful
to be bilingual, to be able to speak hardware to software
and vice
versa.

A very sensible response of course.

\"My\" GEC HRdroid couldn\'t comprehend anything beyond
square holes, and all round candidates has to be
force fitted into one of the square holes.

If the interviewer asks questions but listens to the
answers and avoids such destructive idiocies, that\'s
just fine.

One technique I developed was to ask ever wilder
questions, with the objective of getting them to
(sensibly) say \"no\". That gave me good insight into
the validity of their \"yes\" responses.

This would be for interviewing, versus being interviewed?

I was thinking of being interviewed. After starting my
second job (at a contract design and consultancy company)
they told me that I asked far more questions than most
candidates. I haven\'t stopped since

I haven\'t found it necessary when I\'m the technical
interviewer; there are other more fun and fruitful way
of smoking out blatherers.

What would that be?

Nothing magic.

I get them to describe what they have done in the past,
listen (unlike a few interviewers!), ask them why they
made their choices, and what they would do differently
next time.

I take them to a whiteboard and design something with them.

So do I, either directly or indirectly. That will be
something that is adjacent to their experience, and/or
something that illustrates how the company thinks/works.

The key word is \"with\"; interviews aren\'t (or at least
shouldn\'t be) willy waving competitions.


One thing you can\'t tell from a resume, or the usual interview, is if
they understand electricity. Lots of presumed engineers don\'t. Most
recent EE grads don\'t.

Wanna make a kid panic? Show him a 2-resistor voltage divider. Even
worse, a batery and a resistor charging a cap.

That\'s one reason why I like to see evidence on their
CV that they have done things on their own, without
it being any part of any course.

Doesn\'t matter what it is, so long as they can describe
what, why, and what they would do differently next time.



You don\'t need to be able to design circuits to get an EE job at most
large companies as a newly-minted EE, seems like it\'s expected that\'s
one of the things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too. With my new astronomy and physics bachelor\'s degree, and a
hobby background in electronics. I got hired to do 2/3 of the timing and
frequency control electronics for the first civilian direct-broadcast
satellite system. I\'d heard of PLLs but had never actually come across
one to know what it was. Talk about drinking from a fire hose.

The times I\'ve made the most progress in my life are the times I jumped
off a cliff--that satcom job, getting married, going off to grad school
in another country with a family to support, yada yada.

It\'s been a good ride so far.

Cheers

Phil Hobbs

I was lucky to get a college-freshman job with a company that had no
coherent business plan or marketing. They would take an order for
anything and assume the engineers could do it. Learned a lot.

They paid me $400 a month, which was a fortune in New Orleans back
then.

Good ride indeed. Electronics is universal and gets us into all sorts
on interesting situations.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[server]

On Tue, 21 Jul 2020 23:01:02 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
On 7/20/2020 12:52 PM, Tom Gardner wrote:
On 20/07/20 16:22, jlarkin@highlandsniptechnology.com wrote:
On Mon, 20 Jul 2020 15:29:29 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 20/07/20 14:52, jlarkin@highlandsniptechnology.com wrote:
On Mon, 20 Jul 2020 08:42:36 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 20/07/20 01:52, Joe Gwinn wrote:
On Sat, 18 Jul 2020 08:29:35 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 18/07/20 00:23, Joe Gwinn wrote:
On Fri, 17 Jul 2020 01:32:33 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 21:42, Joe Gwinn wrote:
On Thu, 16 Jul 2020 15:54:54 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 15:32, jlarkin@highlandsniptechnology.com wrote:
On Thu, 16 Jul 2020 14:42:56 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 14:26, jlarkin@highlandsniptechnology.com wrote:
On Thu, 16 Jul 2020 07:28:17 +0100, Tom Gardner
spamjunk@blueyonder.co.uk> wrote:

On 16/07/20 06:32, dagmargoodboat@yahoo.com wrote:
E.g., why would someone on the dole ever work for
something they
already get free?  It doesn\'t make any sense.

True for some people, false for many more.

Many people feel defined by their work, and feel
pointless without it. Such people have a tendency
to \"give up and die\"  relatively shortly after
retiring.

You seem to understand Theory X companies, but
have no clue about Theory Y companies, as described
by McGregor in the 1950s.

Long before McGregor, Hewlett and Packard knew the
difference instinctively, and created a rather
successful Theory Y company. You may have heard
of it.

\"Theory Y managers assume employees are internally
motivated, enjoy their job, and work to better
themselves without a direct reward in return. These
managers view their employees as one of the most
valuable assets to the company, driving the internal
workings of the corporation. Employees additionally
tend to take full responsibility for their work and
do not need close supervision to create a quality
product.\"
https://en.wikipedia.org/wiki/Theory_X_and_Theory_Y

Or, as famously noted at the time of Princess Fiorina,
http://www.satirewire.com/news/0105/loyal.shtml

Sure, but a company doesn\'t become a Y just with a
policy statement.
It requires finding and hiring the right workers,
treating them right,
and firing the ones that don\'t work out.

It didn\'t cross my mind anybody could think mere
policy statements could be sufficient.

In HP, the HP Way was continually reinforced and
re-explained by use of Bill and Dave anecdotes,
wheeled out to show how they thought and wanted
things to be done. Apparently when they were setting
up new sites the first hires became a little sick
and tired of them!

OTOH, Princess Fiorina made very animated policy
pronouncements, which nobody could understand.
That\'s one of the things that made me decide
to leave.

I have Packard\'s book, The HP Way. And I have Fiorina\'s
book, The
Journey. The contrast is hilarious.

Not if you were in HP!


HP did that early on. By about 1980, not so well.

HP was /very/ careful about its hiring process, at least
until shortly before Fiorina ascended in 1999.

I interviewed at HP in about 1980. The guy was obnoxious.
He would
have been my boss.

He looked at my resume and said \"The first thing you need
to do is
decide if you are an engineer or a programmer.\"

What I decided to do was walk out.

Snap!

I had an interview at a GEC site in ~1981. After explaining
the hardware and software and systems I had designed, the
HRdroid asked me whether I was \"really a hardware of software
engineer\".

Somewhat surprisingly, I managed not to give him an earful.
I suspect the expression on my face and my answers becoming
terser might have alerted him to his faux pas. The idiot still
offered me a job.

I have a similar story from the 1970s, but it turned out rather
better.

I was applying to a middle size defense contractor in the
Baltimore
suburbs, and the hiring manager looked over my resume, and
asked which
I preferred, hardware or software.  I replied that it was
very useful
to be bilingual, to be able to speak hardware to software
and vice
versa.

A very sensible response of course.

\"My\" GEC HRdroid couldn\'t comprehend anything beyond
square holes, and all round candidates has to be
force fitted into one of the square holes.

If the interviewer asks questions but listens to the
answers and avoids such destructive idiocies, that\'s
just fine.

One technique I developed was to ask ever wilder
questions, with the objective of getting them to
(sensibly) say \"no\". That gave me good insight into
the validity of their \"yes\" responses.

This would be for interviewing, versus being interviewed?

I was thinking of being interviewed. After starting my
second job (at a contract design and consultancy company)
they told me that I asked far more questions than most
candidates. I haven\'t stopped since

I haven\'t found it necessary when I\'m the technical
interviewer; there are other more fun and fruitful way
of smoking out blatherers.

What would that be?

Nothing magic.

I get them to describe what they have done in the past,
listen (unlike a few interviewers!), ask them why they
made their choices, and what they would do differently
next time.

I take them to a whiteboard and design something with them.

So do I, either directly or indirectly. That will be
something that is adjacent to their experience, and/or
something that illustrates how the company thinks/works.

The key word is \"with\"; interviews aren\'t (or at least
shouldn\'t be) willy waving competitions.


One thing you can\'t tell from a resume, or the usual interview, is if
they understand electricity. Lots of presumed engineers don\'t. Most
recent EE grads don\'t.

Wanna make a kid panic? Show him a 2-resistor voltage divider. Even
worse, a batery and a resistor charging a cap.

That\'s one reason why I like to see evidence on their
CV that they have done things on their own, without
it being any part of any course.

Doesn\'t matter what it is, so long as they can describe
what, why, and what they would do differently next time.



You don\'t need to be able to design circuits to get an EE job at most
large companies as a newly-minted EE, seems like it\'s expected that\'s
one of the things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too. With my new astronomy and physics bachelor\'s degree, and a
hobby background in electronics. I got hired to do 2/3 of the timing and
frequency control electronics for the first civilian direct-broadcast
satellite system. I\'d heard of PLLs but had never actually come across
one to know what it was. Talk about drinking from a fire hose.

The times I\'ve made the most progress in my life are the times I jumped
off a cliff--that satcom job, getting married, going off to grad school
in another country with a family to support, yada yada.

It\'s been a good ride so far.

Cheers

Phil Hobbs

I was lucky to get a college-freshman job with a company that had no
coherent business plan or marketing. They would take an order for
anything and assume the engineers could do it. Learned a lot.

They paid me $400 a month, which was a fortune in New Orleans back
then.

Good ride indeed. Electronics is universal and gets us into all sorts
on interesting situations.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[server]

On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
<spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 04:01, Phil Hobbs wrote:
On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
You don\'t need to be able to design circuits to get an EE job at most large
companies as a newly-minted EE, seems like it\'s expected that\'s one of the
things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too.  With my new astronomy and physics bachelor\'s degree, and a hobby
background in electronics. I got hired to do 2/3 of the timing and frequency
control electronics for the first civilian direct-broadcast satellite system.
I\'d heard of PLLs but had never actually come across one to know what it was.
Talk about drinking from a fire hose.

My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird stuff,
TTL) were horrible.

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent them and
get rich.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard

 

[server]

On Wed, 22 Jul 2020 07:57:58 +0100, Tom Gardner
<spamjunk@blueyonder.co.uk> wrote:

On 22/07/20 04:01, Phil Hobbs wrote:
On 2020-07-20 14:34, Tom Gardner wrote:
On 20/07/20 19:19, bitrex wrote:
You don\'t need to be able to design circuits to get an EE job at most large
companies as a newly-minted EE, seems like it\'s expected that\'s one of the
things you learn on the job.

Not back in my day. I was thrown in at the deep
end on my first day.

I\'ve always had jobs like that, and wouldn\'t have
had it any other way.

Me too.  With my new astronomy and physics bachelor\'s degree, and a hobby
background in electronics. I got hired to do 2/3 of the timing and frequency
control electronics for the first civilian direct-broadcast satellite system.
I\'d heard of PLLs but had never actually come across one to know what it was.
Talk about drinking from a fire hose.

My first job involved that newfangled digital logic; easy.

Beg to differ. Early integrated logic chips (RTL, DTL, weird stuff,
TTL) were horrible.

It also involved creating a test set for the newfangled
multimode optical fibres that were just being installed
between exchanges. My knowledge: zero.

I ended up with a receiver with 180dB electrical (90dB
optical) dynamic range, using a large photodiode (BPW34)
and a LF351 based transimpedance amp. To recover the
signal I decided I couldn\'t predict how a PLL would work in
a range switched design, so I made a filter with a Q of 4000
using 10% components. The noise equivalent power was 1pW.

I\'ve always wanted to revisit that N-path filter for RF work,
since it has interesting properties. When I finally looked,
I found the Tayloe mixer had been patented, dammit.

It is annoying to invent things, and have someone else patent them and
get rich.



--

John Larkin Highland Technology, Inc

Science teaches us to doubt.

Claude Bernard