开云体育

On 08/07/18 04:53, Ed Breya via Groups.Io wrote:

If you are concerned about setting or controlling the frequency, you can

drive the output stage instead from another with XO or a generator. My
preferred 74AC driver for fast, strong pulsing is the 74AC541, with eight
amplifiers that can be paralleled. Also, the pinout is excellent for this
sort of thing, layout-wise - all ins on one side, and all outs on the other.
My preference is 74LVC1G*, which have ~7ohms output impedance, 32mA
drive, and
work from ~1V to 5V albeit with different risetimes.

If you add a 143ohm series resistance and put three of those in series
(and
decouple them very well), then you get a clean 50ohm output. I can't
measure the
risetime; it is certainly sub nanosecond, another measure gives ~600ps,
and I've
seen a (poorly controlled) measurement of 300ps on a 1GHz scope.

From: "David C. Partridge" <david.partridge@...>

A great idea - Hobbycraft has 1/8" hard wood dowels

From: "David C. Partridge" <david.partridge@...>

Problem is the UK suppliers mostly have 3mm not 3.2mm

From: "Bruce Griffiths" <bruce.griffiths@...>
To: [email protected]
Sent: Saturday, July 7, 2018 6:12:50 PM
Subject: Re: [TekScopes] Kludging together a TDR or similar?

At low frequencies where propagation delay can be neglected the frequency
will be k/RC where in this case
where the output resistance of the 74AC14 can also be neglected:
k ~ 1/(ln(Vt+/Vt-)*((Vcc-Vt-)/(Vcc-Vt+)))
where
Vt+ is the upper input switching threshold of the 74AC14
Vt- is the lower input switching threshold of the 74AC14
Vcc (5V in this case) is the power supply voltage for the 74AC14
Bruce

On 08 July 2018 at 06:49 n4buq <n4buq@...> wrote:


No, I'm not overly concerned about the pulse width. I was just wondering
why there is a difference between the measured width (on the scope) vs the
fequency I'd expected with 1/RC. According to Bruce, there's more to it
than that and the hysteresis of the Schmitt trigger has to be taken into
consideration.

Thanks,
Barry - N4BUQ

----- Original Message -----

From: "Michael A. Terrell" <mike.terrell@...>
To: [email protected]
Sent: Saturday, July 7, 2018 9:14:23 AM
Subject: Re: [TekScopes] Kludging together a TDR or similar?

Both parts are 5% but that isn't important.

The time required for a pulse to travel to the discontinuity isn't
affected
by the error in the pulse width. If it bothers you, use a crystal time
base
instead of a simple one gate oscillator.

I have another schematic with multiple pulse widths that you can select
from,
but like this design, it is built with 5% parts. That one also allows you
to
select 50, 75 or 100 ohm output impedance by adding a 22 ohm, or a 50 ohm
resistor in the output path.

Even temperature has some effect on the frequency, along with the
propagation
delay on the gate.


Michael A. Terrell


-----Original Message-----

From: n4buq <n4buq@...>
Sent: Jul 7, 2018 9:03 AM
To: [email protected]
Subject: Re: [TekScopes] Kludging together a TDR or similar?

Looking at that schematic, it appears the frequency is determined by C3
and
R1. If I'm not mistaken, the exact values of those components would
result
in a frequency of 3.129 kHz; however, the oscilloscope images appear to
show 4.81 kHz. Is that difference due to component tolerances or is
something else at work here?

Thanks,
Barry - N4BUQ

On 08 July 2018 at 06:49 n4buq <n4buq@...> wrote:


No, I'm not overly concerned about the pulse width. I was just wondering why there is a difference between the measured width (on the scope) vs the fequency I'd expected with 1/RC. According to Bruce, there's more to it than that and the hysteresis of the Schmitt trigger has to be taken into consideration.

Thanks,
Barry - N4BUQ

----- Original Message -----

From: "Michael A. Terrell" <mike.terrell@...>
To: [email protected]
Sent: Saturday, July 7, 2018 9:14:23 AM
Subject: Re: [TekScopes] Kludging together a TDR or similar?

Both parts are 5% but that isn't important.

The time required for a pulse to travel to the discontinuity isn't affected
by the error in the pulse width. If it bothers you, use a crystal time base
instead of a simple one gate oscillator.

I have another schematic with multiple pulse widths that you can select from,
but like this design, it is built with 5% parts. That one also allows you to
select 50, 75 or 100 ohm output impedance by adding a 22 ohm, or a 50 ohm
resistor in the output path.

Even temperature has some effect on the frequency, along with the propagation
delay on the gate.


Michael A. Terrell


-----Original Message-----

From: n4buq <n4buq@...>
Sent: Jul 7, 2018 9:03 AM
To: [email protected]
Subject: Re: [TekScopes] Kludging together a TDR or similar?

Looking at that schematic, it appears the frequency is determined by C3 and
R1. If I'm not mistaken, the exact values of those components would result
in a frequency of 3.129 kHz; however, the oscilloscope images appear to
show 4.81 kHz. Is that difference due to component tolerances or is
something else at work here?

Thanks,
Barry - N4BUQ

From: "David C. Partridge" <david.partridge@...>

Problem is the UK suppliers mostly have 3mm not 3.2mm

I'm looking for a UK source of 1/8" (0.125" or 3.2mm) diameter fibreglass
rods (or other hard plastic) such as is used in many 7000 series plugins.

For the specific application (the rod down the middle of plastic expansion
rivet) a 3mm rod won't do.

I only want a small quantity, so any suppliers with large minimum orders
would be a problem.

Thanks
David

From: n4buq <n4buq@...>

No, I'm not overly concerned about the pulse width. I was just wondering why there is a difference between the measured width (on the scope) vs the frequency I'd expected with 1/RC. According to Bruce, there's more to it than that and the hysteresis of the Schmitt trigger has to be taken into consideration.

Thanks,
Barry - N4BUQ

From: "Michael A. Terrell" <mike.terrell@...>
To: [email protected]
Sent: Saturday, July 7, 2018 9:14:23 AM
Subject: Re: [TekScopes] Kludging together a TDR or similar?

Both parts are 5% but that isn't important.

The time required for a pulse to travel to the discontinuity isn't affected
by the error in the pulse width. If it bothers you, use a crystal time base
instead of a simple one gate oscillator.

I have another schematic with multiple pulse widths that you can select from,
but like this design, it is built with 5% parts. That one also allows you to
select 50, 75 or 100 ohm output impedance by adding a 22 ohm, or a 50 ohm
resistor in the output path.

Even temperature has some effect on the frequency, along with the propagation
delay on the gate.


Michael A. Terrell


-----Original Message-----

From: n4buq <n4buq@...>
Sent: Jul 7, 2018 9:03 AM
To: [email protected]
Subject: Re: [TekScopes] Kludging together a TDR or similar?

Looking at that schematic, it appears the frequency is determined by C3 and
R1. If I'm not mistaken, the exact values of those components would result
in a frequency of 3.129 kHz; however, the oscilloscope images appear to
show 4.81 kHz. Is that difference due to component tolerances or is
something else at work here?

Thanks,
Barry - N4BUQ

I'm looking for a UK source of 1/8" (0.125" or 3.2mm) diameter fibreglass
rods (or other hard plastic) such as is used in many 7000 series plugins.

For the specific application (the rod down the middle of plastic expansion
rivet) a 3mm rod won't do.

I only want a small quantity, so any suppliers with large minimum orders
would be a problem.

Thanks
David