Topics

Fast Rise/Fall Time Pulser

Larry McDavid
 

There has been discussion recently about generating fast rise/fall time pulses by various means. Here is a link to a Leo Bodnar device that will provide <40 ps rise/fall time, 50% duty cycle, 10 MHz pulses:

http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=124&products_id=295

This is built on a small PWB with integral BNC connector, powered by 5 vdc through a USB B connector; a trigger output is provided. Price is US$68.

I've ordered one of these for testing.

Is anyone here using one of these?

--
Best wishes,

Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)

TomC
 

I have one. It works well. Besides using the rise and fall times displayed on a scope to get an idea of the bandwidth, it is also interesting to see what the square wave looks like on different scopes and plugins.

Tom

On 10/10/2017 12:19 PM, Larry McDavid lmcdavid@... [TekScopes] wrote:
There has been discussion recently about generating fast rise/fall time
pulses by various means. Here is a link to a Leo Bodnar device that will
provide <40 ps rise/fall time, 50% duty cycle, 10 MHz pulses:
http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=124&products_id=295
This is built on a small PWB with integral BNC connector, powered by 5
vdc through a USB B connector; a trigger output is provided. Price is US$68.
I've ordered one of these for testing.
Is anyone here using one of these?
--
Best wishes,
Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)

Larry McDavid
 

Mine arrived today. Alas, it uses a USB B connector instead of a more common USB mini or micro; I gotta get a cable.

Mine has the comparator IC marked only, "AJK BAA." Yes, this is an Analog Devices ADCMP572 or ADCMP573 Ultrafast SiGe comparator. I'm not sure which output configuration is used yet, but likely is the 572 with CML output drivers.

The controller IC is a 18F25K50 PIC with integral USB. Level programming software is available on the Leo Bodnar website for download.

The board was supplied with a test result data sheet from a Tektronix CSA803A (website picture shows a SD30 40 GHz sampling head used) showing the pulser rise time as 38.91 ps and the fall time as 33.20 ps with about 950 mv P-P.

It's a lovely looking board; lots of thought went into the layout. I can't tell if it is multi-layer. Now for that B cable...

Larry

On 10/10/2017 12:19 PM, Larry McDavid lmcdavid@... [TekScopes] wrote:
There has been discussion recently about generating fast rise/fall time
pulses by various means. Here is a link to a Leo Bodnar device that will
provide <40 ps rise/fall time, 50% duty cycle, 10 MHz pulses:
http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=124&products_id=295
This is built on a small PWB with integral BNC connector, powered by 5
vdc through a USB B connector; a trigger output is provided. Price is US$68.
I've ordered one of these for testing.
Is anyone here using one of these?
--
Best wishes,
Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)
--
Best wishes,

Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)

David DiGiacomo
 

On Tue, Oct 10, 2017 at 6:09 PM, Larry McDavid <lmcdavid@...> wrote:
Mine has the comparator IC marked only, "AJK BAA." Yes, this is an
Analog Devices ADCMP572 or ADCMP573 Ultrafast SiGe comparator. I'm not
sure which output configuration is used yet, but likely is the 572 with
CML output drivers.
The prototype used the ADCMP572, but I believe the current version
uses the (much cheaper) MAX3949.

mosaicmerc
 

Sounds like a decent device....

I did acquire a Tektronix S-52 calibrated at 17pS for about $125 a few months ago. I also have those ADCMP ICs (still in vacuum seal) mentioned as I was planning to design my own pulser but I am  a sucker for Tek gear so.....the S-52 was just too pretty ( not a scratch) to resist.

It's just something in me since engineering undergrad days reading all about fine Tektronix scopes on the full page ads inside Radio-Electronics. The scopes were the price of a new car or two back then. I had to own them.......but I don't suppose I need to explain the Tek bug in this forum.

Ancel

 

Hi Larry,

The designer of the Tek 067-0681-01 Tunnel Diode Pulser was impressed with the simplicity of this design and the fast edges. The overshoot (10%) and the undershoot (5%) were not very impressive. He attributes that to the BNC connector. The good news is that the designer may be able to improve the rise and fall times as well as reduce the aberrations on the edges.

His exact comments were:
" Not bad transient response given the speed and the fact that the output is a BNC connector. +1V and -1V are large amplitudes for this speed. Maybe he doesn't know that the BNC is 52 ohms! That would account for some of the aberrations on the front corner. In any event, that is fast enough for most TDR stuff."

He discovered, when he was working on the 067-0681-01, that what everyone assumes is a 50 ohm BNC is not what it appears. He measured the dimensions of the inner diameter, the outer diameter, and the dielectric to prove his point.

The odd thing is that the designer chose to use a BNC connector at all since it tops out at a few GHz. An SMA connector tops out above 18GHz. SMA would not have been any more money, it would have better edge times, and it would have true 50 ohm impedance which would probably reduce the aberrations on the edges. For buyers who did not have an SMA to BNC adapter one could be provided at a modest additional cost.

Dennis Tillman W7PF

-----Original Message-----
From: TekScopes@... [mailto:TekScopes@...]
Sent: Tuesday, October 10, 2017 5:10 PM
To: TekScopes@...
Subject: Re: [TekScopes] Fast Rise/Fall Time Pulser

Mine arrived today. Alas, it uses a USB B connector instead of a more common USB mini or micro; I gotta get a cable.

Mine has the comparator IC marked only, "AJK BAA." Yes, this is an Analog Devices ADCMP572 or ADCMP573 Ultrafast SiGe comparator. I'm not sure which output configuration is used yet, but likely is the 572 with CML output drivers.

The controller IC is a 18F25K50 PIC with integral USB. Level programming software is available on the Leo Bodnar website for download.

The board was supplied with a test result data sheet from a Tektronix CSA803A (website picture shows a SD30 40 GHz sampling head used) showing the pulser rise time as 38.91 ps and the fall time as 33.20 ps with about 950 mv P-P.

It's a lovely looking board; lots of thought went into the layout. I can't tell if it is multi-layer. Now for that B cable...

Larry


On 10/10/2017 12:19 PM, Larry McDavid lmcdavid@... [TekScopes] wrote:
There has been discussion recently about generating fast rise/fall
time pulses by various means. Here is a link to a Leo Bodnar device
that will provide <40 ps rise/fall time, 50% duty cycle, 10 MHz pulses:

http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=1
24&products_id=295

This is built on a small PWB with integral BNC connector, powered by 5
vdc through a USB B connector; a trigger output is provided. Price is US$68.

I've ordered one of these for testing.

Is anyone here using one of these?
--
Best wishes,

Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)
------------------------------------
Posted by: Larry McDavid <lmcdavid@...>
------------------------------------

Larry McDavid
 

Neither the Analog Devices nor the Maxim part (see below) data sheets suggest the markings actually seen on the device used on the Leo Bodnar pulser I received. As recently as July 8, 2017 Leo Bodnar described the device as "based on 572 comparator from AD" in a posting on EEVblog.

The Maxim MAX3949 is a laser diode driver and is also a SiGe device; its data sheet gives Modulation Current rise/fall times of 22 to 36 ps, so this is also a very fast device. Both devices are in similar 16-pin, 3 mm square packages.

Beats me; anyone who knows, please reply. Can anyone explain the "AJK BAA" markings I see on my pulser IC?

I'm not sure I care beyond curiosity what the IC really is. I do care about the rise/fall times. I see no reason not to trust the data that were supplied with my device.

One poster commented on the high pulse voltage of ±1 volt; that is not correct. The Leo Bodnar pulser output is about 0.95 v P-P.

Larry


On 10/10/2017 5:59 PM, David DiGiacomo telists@... [TekScopes] wrote:
On Tue, Oct 10, 2017 at 6:09 PM, Larry McDavid <lmcdavid@...> wrote:
> Mine has the comparator IC marked only, "AJK BAA." Yes, this is an
> Analog Devices ADCMP572 or ADCMP573 Ultrafast SiGe comparator. I'm not
> sure which output configuration is used yet, but likely is the 572 with
> CML output drivers.
The prototype used the ADCMP572, but I believe the current version
uses the (much cheaper) MAX3949.
...
--
Best wishes,

Larry McDavid W6FUB
Anaheim, California (SE of Los Angeles, near Disneyland)

Craig Sawyers <c.sawyers@...>
 

It's just something in me since engineering undergrad days reading all about fine Tektronix scopes
on
the full page ads inside Radio-Electronics. The scopes were the price of a new car or two back then.
I
had to own them.......but I don't suppose I need to explain the Tek bug in this forum.
That is me too. I used Tek scopes and transient digitizers when at University back in the '70's when
they cost a house mortgage. So I just has to own them!

Craig

Craig Sawyers <c.sawyers@...>
 

His exact comments were:
" Not bad transient response given the speed and the fact that the output is a BNC connector. +1V
and -1V are large amplitudes for this speed. Maybe he doesn't know that the BNC is 52 ohms! That
would account for some of the aberrations on the front corner. In any event, that is fast enough
for
most TDR stuff."

He discovered, when he was working on the 067-0681-01, that what everyone assumes is a 50 ohm
BNC is not what it appears. He measured the dimensions of the inner diameter, the outer diameter,
and the dielectric to prove his point.
That is indeed interesting Dennis. I certainly knew that Tek developed their own BNC connectors for
the 7000-series plugins to overcome the wonky impedance of off-the-shelf BNC's.

Craig

 

Hi Craig,

If you open up 067-0681-01 you can see another trick he developed. The
Tunnel Diode rests inside a small cavity in the Teflon dielectric so it is
physically as close to the output pin of the BNC as possible. He also used
three 1K 1/8 watt resistors in series instead of one 3K 1/4W resistor to
bias the TD. The reason is that the physical size of the 1/8W lead
inductance and body capacitance balance out when the resistor is ~1K so this
reduces some of the aberrations on the edges of the pulses.

I learn a lot every time he explains one of his designs to me.

Dennis Tillman W7PF

-----Original Message-----
Sent: Tuesday, October 10, 2017 11:53 PM
Subject: RE: [TekScopes] Fast Rise/Fall Time Pulser

His exact comments were:
" Not bad transient response given the speed and the fact that the
output is a BNC connector. +1V and -1V are large amplitudes for this
speed. Maybe he doesn't know that the BNC is 52 ohms! That would
account for some of the aberrations on the front corner. In any
event, that is fast enough for most TDR stuff."

He discovered, when he was working on the 067-0681-01, that what
everyone assumes is a 50 ohm BNC is not what it appears. He measured
the dimensions of the inner diameter, the outer diameter, and the
dielectric to prove his point.

That is indeed interesting Dennis. I certainly knew that Tek developed their
own BNC connectors for the 7000-series plugins to overcome the wonky
impedance of off-the-shelf BNC's.

Craig
------------------------------------
Posted by: "Craig Sawyers" <c.sawyers@...>
------------------------------------

 

The Tunnel Diode rests inside a small cavity in the Teflon dielectric
I own three of these and I have seen several other and none of them
have the tunnel diode inside the BNC connector. Wonder if that would
even be possible since one end of the diode goes to GND and the other
to the end of the three 1k resistors you mention. Inside the Teflon
dielectric in the BNC is a 43 ohm resistor. The tunnel diode is soldered
between a GND pin on the BNC connector and the junction of the 43 ohm
and the last 1k, close to but outside the BNC.

/Håkan

Erik Hultine
 

I also own three and they are as Håkan describes.
 
-Erik
 
“The snake which cannot cast its skin has to die.”- Friedrich Nietzsche

-----Original Message-----
From: "hahi@... [TekScopes]" [TekScopes@...]
Date: 11/10/2017 04:38
To: TekScopes@...
Subject: RE: [TekScopes] Fast Rise/Fall Time Pulser

 > The Tunnel Diode rests inside a small cavity in the Teflon dielectric

I own three of these and I have seen several other and none of them
have the tunnel diode inside the BNC connector. Wonder if that would
even be possible since one end of the diode goes to GND and the other
to the end of the three 1k resistors you mention. Inside the Teflon
dielectric in the BNC is a 43 ohm resistor. The tunnel diode is soldered
between a GND pin on the BNC connector and the junction of the 43 ohm
and the last 1k, close to but outside the BNC.

/Håkan

[Non-text portions of this message have been removed]
 
 

[Non-text portions of this message have been removed]

Eino Väänänen
 

I have had 067-0554-00 for a long time. I tried it with PG506 set to 100 V square wave. The amplitude was 60 mV and rise time was about 700 ns. Maybe there is something wrong? The pulse could be seen also with smaller amplitudes than 100 V !?

Chuck Harris
 

The 100V squarewave is mostly just the power supply for
the tunnel diode, and it also sets the repetition rate
for the tunnel diode pulse.

There is a control on the pulser that sets the current
into the diode, that must be set to the point where the
pulse just starts to appear.

The amplitude is set by the tunnel diode, and the resistors
around it.

How are you measuring the rise time? If you are measuring
it on a scope with about 20MHz bandwith, it will go slow
like that. (think BWL position)

-Chuck Harris

einovaan@... [TekScopes] wrote:

I have had 067-0554-00 for a long time. I tried it with PG506 set to 100 V square wave. The amplitude was 60 mV and rise time was about 700 ns. Maybe there is something wrong? The pulse could be seen also with smaller amplitudes than 100 V !?




David DiGiacomo
 

On Tue, Oct 10, 2017 at 11:01 PM, Larry McDavid <lmcdavid@...> wrote:
Neither the Analog Devices nor the Maxim part (see below) data sheets
suggest the markings actually seen on the device used on the Leo Bodnar
pulser I received. As recently as July 8, 2017 Leo Bodnar described the
device as "based on 572 comparator from AD" in a posting on EEVblog.
The Maxim datasheet doesn't say anything about the product marking.

Here's Leo's later post:
https://www.eevblog.com/forum/projects/yet-another-fast-edge-pulse-generator/msg1275011/#msg1275011

One poster commented on the high pulse voltage of ±1 volt; that is not
correct. The Leo Bodnar pulser output is about 0.95 v P-P.
It's adjustable up to 1.2VPP into 50ohms, or 2.4VPP open circuit.

Product page: http://www.leobodnar.com/shop/index.php?main_page=product_info&cPath=124&products_id=295

Dave Wise
 

What frequency is the PG506? The pulser cannot cycle above a few hundred kHz.

Dave Wise

-----Original Message-----
From: TekScopes@... [mailto:TekScopes@...]
Sent: Wednesday, October 11, 2017 8:27 AM
To: TekScopes@...
Subject: Re: [TekScopes] Fast Rise/Fall Time Pulser

The 100V squarewave is mostly just the power supply for the tunnel diode, and it also sets the repetition rate for the tunnel diode pulse.

There is a control on the pulser that sets the current into the diode, that must be set to the point where the pulse just starts to appear.

The amplitude is set by the tunnel diode, and the resistors around it.

How are you measuring the rise time? If you are measuring it on a scope with about 20MHz bandwith, it will go slow like that. (think BWL position)

-Chuck Harris

einovaan@... [TekScopes] wrote:
I have had 067-0554-00 for a long time. I tried it with PG506 set to 100 V square wave. The amplitude was 60 mV and rise time was about 700 ns. Maybe there is something wrong? The pulse could be seen also with smaller amplitudes than 100 V !?

[Non-text portions of this message have been removed]



------------------------------------
Posted by: Chuck Harris <cfharris@...>
------------------------------------

Eino Väänänen
 

Thanks for comments Chuck and David!


PG506 frequency is 1 kHz. BW of the scope is 500 MHz and I think the frequency is same 1 kHz to the scope input. Tunnel diode has not changed it. The amplitude 60 mV also looks too small. Maybe 100 V is not large enough or there is some faulty part. As I told there comes same 1 kHz frequency to the scope input also when using voltages 5 ... 100 V with smaller amplitude.


Eino

Bert Haskins
 

On 10/11/2017 11:27 AM, Chuck Harris cfharris@... [TekScopes] wrote:

The 100V squarewave is mostly just the power supply for
the tunnel diode, and it also sets the repetition rate
for the tunnel diode pulse.

There is a control on the pulser that sets the current
into the diode, that must be set to the point where the
pulse just starts to appear.

The amplitude is set by the tunnel diode, and the resistors
around it.

How are you measuring the rise time? If you are measuring
it on a scope with about 20MHz bandwith, it will go slow
like that. (think BWL position)

-Chuck Harris

einovaan@... [TekScopes] wrote:
I have had 067-0554-00 for a long time. I tried it with PG506 set to
100 V square wave. The amplitude was 60 mV and rise time was about 700
ns. Maybe there is something wrong? The pulse could be seen also with
smaller amplitudes than 100 V !?





------------------------------------------------------------------------
Posted by: Chuck Harris <cfharris@...>
Why the need for the 100 v pulse?
If it's for constant current why not just use a pair of transistors to
get the job done.

Bert










Chuck Harris
 

1KHz is too low, set your PG506 to 100KHz, High voltage pulse,
and the pulse amplitude control fully clockwise.

Put the scope in its highest sweep speed, AC trigger,
NORM trigger mode, and "+" (leading edge).

The vertical amplifier must be in 50 ohm DC mode, if it
can't, you must use a 50 ohm terminator feed through.

The vertical amplifier must be in 10mv position, where the
pulser should be able to make 5 divisions of amplitude
peak-to-peak.

Turn the tunnel diode pulser's control fully CCW, and then
slowly turn it clockwise until the scope starts to trigger.

You may have to adjust trigger level, horizontal position,
and trigger hold off to get the trace you desire.

-Chuck Harris



einovaan@... [TekScopes] wrote:

Thanks for comments Chuck and David!


PG506 frequency is 1 kHz. BW of the scope is 500 MHz and I think the frequency is same 1 kHz to the scope input. Tunnel diode has not changed it. The amplitude 60 mV also looks too small. Maybe 100 V is not large enough or there is some faulty part. As I told there comes same 1 kHz frequency to the scope input also when using voltages 5 ... 100 V with smaller amplitude.


Eino








------------------------------------
Posted by: einovaan@...
------------------------------------


------------------------------------

Yahoo Groups Links



Dave Wise
 

This pulser was designed to operate off the mainframe's Calibrator output, set to 100V, or a 106 Square Wave Generator. I seem to recall that a modification was needed before it would work with the PG506.

Dave Wise

-----Original Message-----
From: TekScopes@... [mailto:TekScopes@...]
Sent: Wednesday, October 11, 2017 10:18 AM
To: TekScopes@...
Subject: Re: [TekScopes] Fast Rise/Fall Time Pulser



On 10/11/2017 11:27 AM, Chuck Harris cfharris@... [TekScopes] wrote:

The 100V squarewave is mostly just the power supply for the tunnel
diode, and it also sets the repetition rate for the tunnel diode
pulse.

There is a control on the pulser that sets the current into the diode,
that must be set to the point where the pulse just starts to appear.

The amplitude is set by the tunnel diode, and the resistors around it.

How are you measuring the rise time? If you are measuring it on a
scope with about 20MHz bandwith, it will go slow like that. (think BWL
position)

-Chuck Harris

einovaan@... [TekScopes] wrote:
I have had 067-0554-00 for a long time. I tried it with PG506 set to
100 V square wave. The amplitude was 60 mV and rise time was about 700
ns. Maybe there is something wrong? The pulse could be seen also with
smaller amplitudes than 100 V !?





----------------------------------------------------------------------
-- Posted by: Chuck Harris <cfharris@...>
Why the need for the 100 v pulse?
If it's for constant current why not just use a pair of transistors to get the job done.

Bert