little circuit problem

On Saturday, 11 May 2019 09:26:54 UTC+1, piglet wrote:

In the spirit of gray on gray sketches here is mine:

https://www.dropbox.com/s/v18svcy5kre4ymi/ActivityBlink.jpg?dl=0

The Mhz oscillator is gated off at a few Hz if there is a stream of data
pulses. Everything is AC coupled so a stuck high/ stuck low fault is
apparent.

The 50ms blink-on blink-off half shots could probably be tidied up into
a 74HC123

piglet

The collector cap needs a resistor though.


NT

 

[John Larkin]

On Sat, 11 May 2019 08:16:25 -0700 (PDT), tabbypurr@gmail.com wrote:

On Saturday, 11 May 2019 09:26:54 UTC+1, piglet wrote:

In the spirit of gray on gray sketches here is mine:

https://www.dropbox.com/s/v18svcy5kre4ymi/ActivityBlink.jpg?dl=0

The Mhz oscillator is gated off at a few Hz if there is a stream of data
pulses. Everything is AC coupled so a stuck high/ stuck low fault is
apparent.

The 50ms blink-on blink-off half shots could probably be tidied up into
a 74HC123

piglet

The collector cap needs a resistor though.


NT

Why?


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[John Larkin]

On Sat, 11 May 2019 15:39:22 +0100, Piglet <erichpwagner@hotmail.com>
wrote:

On 11/05/2019 15:27, John Larkin wrote:
On Sat, 11 May 2019 09:26:51 +0100, Piglet <erichpwagner@hotmail.com
wrote:

On 10/05/2019 17:31, Piglet wrote:
On 09/05/2019 16:40, John Larkin wrote:


Assume I have a transmitter box T. It drives four remote receiver
boxes R1 R2 R3 R4. There are four bidirectional AC-coupled (telecom
style) duplex fiberoptic links.

We only need to transmit from four ports on T to Rn, so the return
links are wasted. So we figured we could return a status indication
from each R box that could drive a couple of LEDs at each tx port of
the T box.

We'd send back three states from each R: I'm dead, I'm here, and I'm
here and receiving your signal. The tx ports would each have two LEDs,
called LINK and DATA or something. Cute little light pipe things.

So, how about some maximally simple circuits at each end to convey
those three states, available as two TTL levels at R, back to the LEDs
on the T box? The tosa/rosa things are all 3.3V differential CML logic
levels.



You could simplify two leds down to one. Off = no connection; solid-on =
connected; Flickering-on = connected and data traffic.

piglet


In the spirit of gray on gray sketches here is mine:

https://www.dropbox.com/s/v18svcy5kre4ymi/ActivityBlink.jpg?dl=0

The Mhz oscillator is gated off at a few Hz if there is a stream of data
pulses. Everything is AC coupled so a stuck high/ stuck low fault is
apparent.

The 50ms blink-on blink-off half shots could probably be tidied up into
a 74HC123

piglet


Interesting. But the light pipes come in twos, and we'd have to
explain the single-LED blink patterns to the users. The two LEDs can
be LINK and DATA which is arguably clear in meaning. Maybe.

The data seen at the recever is GHz stuff, so that needs a separate
detector that supplies a TTL level to this link. That will already
drive a local DATA led on the R box.





No problem. I saw the network activity light on my modem/router which
indicates the same kind of thing: disconnect, connect, active.

You asked for simple and I enjoyed the challenge.

piglet

Good. So few people who post to SED are interested in or skilled at
electronic design. Or seem to ever do it.

Little circuit puzzles are fun.





--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Piglet]

On 11/05/2019 18:13, tabbypurr@gmail.com wrote:

On Saturday, 11 May 2019 16:59:56 UTC+1, John Larkin wrote:
On Sat, 11 May 2019 08:16:25 -0700 (PDT), tabbypurr wrote:
On Saturday, 11 May 2019 09:26:54 UTC+1, piglet wrote:

In the spirit of gray on gray sketches here is mine:

https://www.dropbox.com/s/v18svcy5kre4ymi/ActivityBlink.jpg?dl=0

The Mhz oscillator is gated off at a few Hz if there is a stream of data
pulses. Everything is AC coupled so a stuck high/ stuck low fault is
apparent.

The 50ms blink-on blink-off half shots could probably be tidied up into
a 74HC123

piglet

The collector cap needs a resistor though.


NT

Why?

Otherwise tr is passing uncontrolled current peaks c to e from the 0.1uF.


NT

High current yes but "uncontrolled" is debatable. There is intrinsic
silicon resistance; capacitor esr; track resistance; limited base drive;
hfe depression at high Ic; extreme short duration; etc etc which often
means further resistance unnecessary. But by all means add a few ohms
when you build this circuit

piglet

 

On Saturday, 11 May 2019 16:59:56 UTC+1, John Larkin wrote:

On Sat, 11 May 2019 08:16:25 -0700 (PDT), tabbypurr wrote:
On Saturday, 11 May 2019 09:26:54 UTC+1, piglet wrote:

In the spirit of gray on gray sketches here is mine:

https://www.dropbox.com/s/v18svcy5kre4ymi/ActivityBlink.jpg?dl=0

The Mhz oscillator is gated off at a few Hz if there is a stream of data
pulses. Everything is AC coupled so a stuck high/ stuck low fault is
apparent.

The 50ms blink-on blink-off half shots could probably be tidied up into
a 74HC123

piglet

The collector cap needs a resistor though.


NT

Why?

Otherwise tr is passing uncontrolled current peaks c to e from the 0.1uF.


NT

 

[John Larkin]

On Sat, 11 May 2019 10:13:09 -0700 (PDT), tabbypurr@gmail.com wrote:

On Saturday, 11 May 2019 16:59:56 UTC+1, John Larkin wrote:
On Sat, 11 May 2019 08:16:25 -0700 (PDT), tabbypurr wrote:
On Saturday, 11 May 2019 09:26:54 UTC+1, piglet wrote:

In the spirit of gray on gray sketches here is mine:

https://www.dropbox.com/s/v18svcy5kre4ymi/ActivityBlink.jpg?dl=0

The Mhz oscillator is gated off at a few Hz if there is a stream of data
pulses. Everything is AC coupled so a stuck high/ stuck low fault is
apparent.

The 50ms blink-on blink-off half shots could probably be tidied up into
a 74HC123

piglet

The collector cap needs a resistor though.


NT

Why?

Otherwise tr is passing uncontrolled current peaks c to e from the 0.1uF.


NT

0.1 uF charged to a few volts is a microjoule at best. That's not
going to fry any transistor.

The pumped 1n cap limits base current anyhow.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

On Saturday, May 11, 2019 at 5:19:17 PM UTC-4, John Larkin wrote:

On Sat, 11 May 2019 10:13:09 -0700 (PDT), tabbypurr@gmail.com wrote:

On Saturday, 11 May 2019 16:59:56 UTC+1, John Larkin wrote:
On Sat, 11 May 2019 08:16:25 -0700 (PDT), tabbypurr wrote:
On Saturday, 11 May 2019 09:26:54 UTC+1, piglet wrote:

In the spirit of gray on gray sketches here is mine:

https://www.dropbox.com/s/v18svcy5kre4ymi/ActivityBlink.jpg?dl=0

The Mhz oscillator is gated off at a few Hz if there is a stream of data
pulses. Everything is AC coupled so a stuck high/ stuck low fault is
apparent.

The 50ms blink-on blink-off half shots could probably be tidied up into
a 74HC123

piglet

The collector cap needs a resistor though.


NT

Why?

Otherwise tr is passing uncontrolled current peaks c to e from the 0.1uF.


NT

0.1 uF charged to a few volts is a microjoule at best. That's not
going to fry any transistor.

The pumped 1n cap limits base current anyhow.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

one LED per Rx.

no tone = I'm not there= light off

continuous tone = I'm here = light on

tone pulseing 1 / second = I'm here and Rx data, = light flashing.

m

 

[John Larkin]

On Thu, 09 May 2019 08:40:13 -0700, John Larkin
<jjlarkin@highlandtechnology.com> wrote:

Assume I have a transmitter box T. It drives four remote receiver
boxes R1 R2 R3 R4. There are four bidirectional AC-coupled (telecom
style) duplex fiberoptic links.

We only need to transmit from four ports on T to Rn, so the return
links are wasted. So we figured we could return a status indication
from each R box that could drive a couple of LEDs at each tx port of
the T box.

We'd send back three states from each R: I'm dead, I'm here, and I'm
here and receiving your signal. The tx ports would each have two LEDs,
called LINK and DATA or something. Cute little light pipe things.

So, how about some maximally simple circuits at each end to convey
those three states, available as two TTL levels at R, back to the LEDs
on the T box? The tosa/rosa things are all 3.3V differential CML logic
levels.

This is slick, a minimal way to make two frequencies:

https://www.analog.com/en/products/ltc6905.html

The lowest pair is 17 and 34 MHz, which we can probably handle at the
receiver.

Maxim and Silabs make silicon oscillators too.



--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

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