Coax modelling question

You should look at the service manual for a Tek 1502. Start at

for photos and the manual plus

The latter is suggested as a replacement for the tunnel diode step source.

That's what I got from a home-etched board / semi rigid mixture:

< >

IIRC it was the CML version.

regards, Gerhard

Thanks, I'd seen many resources including the 1502, but not the Starecki/Misiaszek paper - though they don't give too many details, what they did is close to what I envisaged.

There are some very nice VNA sampling head projects documented on the web, and they'd work well to digitise the output from a TDR.

Though I must say, that DS1021S-25 with its minimum step of 250ps, that's already 7cm, or 5cm on a co-ax... I liked Win's biassed comparator on a ramp better (which is also what Tom McEwan did).

Clifford Heath.

Wouldn't it be easier to use edge-mount SMA connectors and have them poke out of the side of the box? BNC would obviously be easier for students to use, but for a really fast clean signal the SMAs should be a lot better.

John

Student gear has to be super rugged, though. I use metal BNCs a lot.

Cheers

Phil Hobbs

I looked at digital delay lines a couple of years ago, and IIRC there are several with much smaller delay increments. ISTR they are effectively ECL-to-ECL with a chain of internal ECL gates that are switched in/out.

None of them were Maxim, which removes a source of, um, uncertainty.

Nice! What was the substrate and etch resist :)

< > and right or left neighbour.

printed on OKI laser printer to foil, contact copy to

0.8 mm FR4 presensitized material by Bungard between 2 glass plates, 15 min UV from a "cosmetic" UV lamp. Nearly all my test board are 1-sided + GND, so there can go nothing wrong with the alignment. < >

The process is good for 5 mil features and done in an hour if you are used to it. Most of the RF path is UG-141 semi rigid. I also can re-use circuit & layout for more complex boards via the Altium snippet feature.

regards, Gerhard

Yeah I don't really know about fast stuff. I don't really need state of the art.. just a learning experience for students. BNC's are standard.. that's what all the 'scope inputs are going to be. I'm also guessing that if we made it with SMA the user's would stick on a BNC adapter.

George H.

Maybe you can get a case of red freebies at the fine, and very beautiful, golf club and resort (one of the finest in the state of New York) just a stone's throw away?

--
Grizzly H.

Thanks.

The foil is an interesting technique that I don't think is available to me. I've seen people claiming and demonstrating

6 mil (0.15 mm) traces using the bog-standard toner+iron technique: of all places, but with a pleasantly pragmatic/rigorous attitude.

Anyway; good results. My fingers are becoming itchy; too many things to do with my remaining life.

BNCs are fine to at least 3 GHz. A 1 ns TDR is effectively about 350 MHz. The scopes used with a student TDR will have BNCs too.

A 1 ns edge is a foot long at c.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

The ADCMP580 is a nice fast step generator, but the Maxim delay line is horrible. Never Buy Maxim.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

SMAs are typically spec'ed for only 500 mating cycles.

At those speeds, connectors matter. See , where I compared some connectors using a TDR setup with a Tek S-52 pulse generator, S-6 sampler setup.

Jeroen Belleman

Here's a test on some cheap BNCs and adapters. Scope net TDR rise time is under 30 ps.

The discontinuties are mild and only last a couple hundred ps. The 147 ps thru risetime includes coax losses.

The RG58 coax chunks are 51 and 47.5 ohms!

BNC terminators are often terrible.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

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

And that's assuming you use a torque wrench and don't let the cable rotate as you tighten the nut. If SMBs weren't so flimsy, they'd be a good option. JL uses a whole lot of them, but his customers are well-behaved, at least towards instruments. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

That's a very encouraging result Gerhard. I also use single-sided toner transfer with foil, though I need to test the process with my new printer.

I'd love to see a photo showing how you used semi-rigid line with your PCB.

Have you attempted a sampler also?

Clifford Heath.

On Sep 15, 2016, Phil Hobbs wrote (in article):

I agree that SMAs are a bit fragile. Even the stainless steel ones.

TNC connector are student-proof and far more stable than BNCs. And then there are Type N connectors.

Joe Gwinn