Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!eternal-september.org!.POSTED!not-for-mail
From: Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net>
Newsgroups: sci.electronics.design
Subject: Re: like butta
Date: Mon, 27 Jan 2025 00:15:30 -0000 (UTC)
Organization: A noiseless patient Spider
Lines: 78
Message-ID: <vn6j72$96vo$1@dont-email.me>
References: <vmsds7$1bj14$1@dont-email.me>
 <3ch3pjpt1aslfleh2dfu99646aaf9ca4s1@4ax.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Injection-Date: Mon, 27 Jan 2025 01:15:30 +0100 (CET)
Injection-Info: dont-email.me; posting-host="009e42b1ea25dd91ec4174700914c699";
	logging-data="302072"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/6ghwg3Xyl2w/JUV90hUZ7"
User-Agent: NewsTap/5.5 (iPhone/iPod Touch)
Cancel-Lock: sha1:tTSgkwCxO3HZXOoB/ZU1xjPWhhA=
	sha1:/gLne4Ghg43/ptJ6OABVZ5V7p9w=
Bytes: 4155

john larkin <JL@gct.com> wrote:
> On Thu, 23 Jan 2025 03:43:03 -0000 (UTC), Phil Hobbs
> <pcdhSpamMeSenseless@electrooptical.net> wrote:
> 
>> Simon and I are working on a TDR based soil moisture sensor for
>> agriculture.  Like many such things, it uses two parallel tines made of
>> 18/8 stainless, that form a balanced transmission line. 
>> 
>> Ours has a slide hammer for pounding it into really difficult soil, e.g.
>> hardpan. The measured shock from that is around 1.6E5 m/s**2, i.e. 16000
>> gees, and over its lifetime it might see around 1E6 blows. Challenging. 
>> 
>> Doing <200 ps TDR on a balanced line obviously needs a very wideband balun.
>> We’re using a plain ferrite design based on Ferroxcube 61 sleeves on 1.25mm
>> coax. The resulting 50-? differential mode goes into a machined tapered
>> structure that maintains 50 ohms while spreading out the mode to match the
>> 25-mm tine spacing. 
> 
> We've got excellent results with micro-coax on pot cores.
> 
> https://www.dropbox.com/scl/fi/o2r1m53es9rb3e7nbsnsr/Pot_Core_TXline.JPG?rlkey=t1g3rp0erz72tqchb99fupfu7&raw=1
> 
> https://www.dropbox.com/scl/fi/ie7gzmdwuw4gqgy1pxpjl/TX_1.jpg?rlkey=xllwjn2cg0a0t3yjh90om9ap1&raw=1

I have that one in my tricks file, for sure. 

We’re using two ferrite sleeves as a 1:1 balun, which is the sideways
version. 
> 
>> 
>> For survivability, the whole thing is being potted in very hard epoxy with
>> a dielectric constant of 3.5ish. This means that it’s hard to iterate—we
>> get one try per apparatus. 
>> 
>> Sooooo, I tried making a soft material with that epsilon, which turns out
>> to be nontrivial. My initial thought was to use alumina lapping powder
>> (9.3) for filler and vaseline (2.0) for the matrix, but it turned into
>> clumps by the time it got to 3.0. 
>> 
>> Casting about for alternatives, I remembered the butter in the lab fridge,
>> so I measured that: 3.8, better than good enough. 
>> 
>> We were able to verify the design without doing anything irrevocable, and
>> there was some left over for the bagels. 
>> 
>> Fun. 
>> 
>> Cheers 
>> 
>> Phil Hobbs 
> 
> Are you doing a reverse convolution to beautify the step waveform? You
> can tolerate a really ugly TDR if you can make a FIR filter to pound
> it flat.

The step is actually very nice-looking—clean edges, flat pulse tops, no
worries.  We did have to make a couple of layout adjustments to get there,
but the result shows TDR edges surprisingly similar to the SD-24’s. 

Ours is much less fancy, and needs a few pulses per delay value to get good
convergence in the sampling loop, but in a not-too-scientific comparison on
a 24-inch RG-188 cable, both the SD-24 and our gizmo showed 60-ps TDR edges
(10-90%).  Of course the SD-24 is more like 30 ps on a less lossy cable,
but for $16 per completed board, including MCU and data converters, I’m
pretty happy with it. 

Cheers 

Phil Hobbs 
> 
> 



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