Re: OT: High Frequency coax?? Copper tubing??

GerryR
 

Dennis,
Thank you again, but just for the record, I was not referring to Leo's design. When I went to his website, I realized that his design was much different than the tunnel diode pulser. It was actually a designer (I don't know his name) that uses tunnel diodes (I believe from a Russian source) in his pulser, so I assumed it was a copy or similar design of the old Tek pulser. These subtleties in the designs, which make all the difference in performance, are rarely explained in detail. I realize that certain aspects of a design must remain proprietary for marketing reasons, but as "they" say, the devil is in the details, and that's what interests me. Thank you for your explanation; it will be added to my file of notes on the Tek pulser.
Best regards,
GerryR
KK4GER

----- Original Message -----
From: "Dennis Tillman W7PF" <@Dennis_Tillman_W7PF>
To: <TekScopes@groups.io>
Sent: Sunday, November 03, 2019 7:40 PM
Subject: Re: [TekScopes] OT: High Frequency coax?? Copper tubing??


Hi Gerry,
I'm going to go out on a limb and guessing you are talking about the Leo Bodnar Pulser and the Tektronix 067-0681-01 Calibration Fixture Tunnel Diode Pulser designed by a legendary engineer at Tektronix, John Addis, who spent much of his career understanding and designing high speed vertical amplifiers.

Making a fast rise pulse generator is actually pretty simple. Controlling the aberrations is where the skill comes in. I introduced Leo to John because I thought Leo might benefit from John's experience particularly when it came to reducing aberrations on the leading and trailing edges. They did share many emails discussing circuit changes that would improve the performance of Leo's Pulser. Some were easy to incorporate into the design. Some changes resulted in tradeoffs that would require building many actual prototypes to evaluate which would be very time consuming. Other changes would result in increases in cost that might have an adverse impact on sales.

John's tunnel diode design incorporates some very subtle design features that aren't obvious such as three 1K 1/8W resistors in series (this eliminates the inductance and capacitance that a 3K 1/4W resistor has) which further reduces the aberrations. The BNC output connector is special as well. Nearly all BNC connectors are assumed to be 50 ohms but they are actually 51 to 52 ohms and poorly specified by their manufacturers. Tek made actual BNCs that are 50.0 ohms for their TD Pulser to reduce aberrations caused by the impedance mismatch. Tek placed the TD right between the center pin and the case of the BNC to further reduce aberrations. On several occasions manufacturing actually relaxed John's requirement on keeping the TD leads as short as possible because it was too time consuming. The result was an increase in aberrations that John found out about. He quickly put a stop to it.

Leo took an entire different approach which takes advantage of readily available semiconductor devices made today. If I am not mistaken Leo used extremely fast current mode optical driver ICs. They have their own advantages and have been steadily improving over time. It is entirely possible they will one day outperform the aberration specs of the TD Pulser. The BNCs that are sold today are not capable of more than a few GHz and they are still poorly specified. Today's SMA connectors are routinely capable of
14GHz and they are well specified. The capabilities of sampling scopes have
improved dramatically since John designed the TD Pulser. Leo has the benefit of being able to take advantage of those scopes as well as high speed pulse circuit design software that didn't exist when John designed his TD Pulser.

It is no accident that the fastest Tek sampling heads (S-4 and S-6) have SMA connectors on them. I urged Leo to replace the BNC on his original design with an SMA connector. I was aware it would result in an inconvenience for some people but I felt Leo's goal should always be to create a reasonably priced pulse generator with the fastest possible rise time and the minimum aberrations. Buyers can always attach an inexpensive SMA to BNC adapter to it if convenience is more important than speed.

Dennis Tillman W7PF

-----Original Message-----
From: GerryR
Sent: Sunday, November 03, 2019 2:22 PM

Thanks Dennis,
After I asked the question I did look up the cable and found out its characteristics. I've spent most of my career (now retired) in industrial electronics / control systems, so some RF stuff still puzzles me.
For instance, recently I inquired about a tunnel diode pulser from a person who makes them. He stated that he offered them with SMA connectors; I wanted BNC connectors. He told me that he couldn't get the response with the BNC's that he can get with the SMA's, yet Tektronix offered (at one
time) a tunnel diode pulser that had a better response than what he offered, with BNC connectors. Why? Thus, my question as to why the semi-rigid cable is so much better at the higher frequencies? Just curious. Thanks, again; I didn't mean for anyone to do my homework for me.
Best regards,
GerryR
KK4GER

----- Original Message -----
From: "Dennis Tillman W7PF" <@Dennis_Tillman_W7PF>
To: <TekScopes@groups.io>
Sent: Sunday, November 03, 2019 4:54 PM
Subject: Re: [TekScopes] OT: High Frequency coax?? Copper tubing??

Hi Gerry,
Every general class ham should have a coax chart at your fingertips
somewhere. Personally I prefer the quickly accessible Wikipedia article on
Coaxial Cable at https://en.wikipedia.org/wiki/Coaxial_cable. It contains a
very detailed chart listing a very large number of coax types and all of the
important parameters anyone would need to compare them all.

It says the insulator / dielectric for RG-402 and RG-405 is PTFE (aka
Teflon).

Dennis Tillman W7PF

-----Original Message-----
From: GerryR
Sent: Sunday, November 03, 2019 11:15 AM

I'm not at all familiar with this type of coax. What is the core material,
and what makes it so good at the higher frequencies?
GerryR
KK4GER

----- Original Message -----
From: "Dennis Tillman W7PF" <@Dennis_Tillman_W7PF>
Sent: Sunday, November 03, 2019 1:46 PM

These two sizes of Semi-Rigid Copper Coax cable are very common:
RG-402 which is 3.58mm OD (0.141" OD)
RG-405 which is 2.20mm OD (0.0865" OD)

Working with it is very different compared to typical coax we are all used
to. Tools to cut it properly and to bend it properly are expensive. Using
common tools to cut it like a pipe cutter, and to bend it like an
inexpensive bending jig, will almost certainly affect the high frequency
performance of whatever you make.

Dennis Tillman W7PF

-----Original Message-----
From: Reginald Beardsley
Sent: Sunday, November 03, 2019 9:01 AM

Here's a representative link for the information you're looking for:

http://www.sumlocable.com/sub_product/rg401-rg402rg405-semi-rigid-43.html

There's lots more. This was simply the first one google tossed up. It is
great stuff provided you don't want to bend it more than once. Though
rebending used cables is OK. The solder filled braid "RG402" has become my
standard cable for anything above 1 GHz. I've made a couple of H field
probes from a used piece of high grade stuff.

Connector quality on Chinese ebay jumpers is rather variable. I've been
documenting my efforts in this thread:
https://www.eevblog.com/forum/rf-microwave/testing-rf-connectors-and-cables/

I will be supplementing the TDR results with 8753B/85046A testing soon now
that I have an APC-7 cal kit and adapters.
Have Fun!
Reg




--
Dennis Tillman W7PF
TekScopes Moderator

Join TekScopes@groups.io to automatically receive all group messages.