[time-nuts] Experience with THS788 from TI?

Ben Gamari bgamari at physics.umass.edu
Thu Mar 22 22:19:44 UTC 2012


Attila Kinali <attila at kinali.ch> writes:

> On Thu, 22 Mar 2012 11:12:40 -0400
> Ben Gamari <bgamari at physics.umass.edu> wrote:
>
>> > If you are really going to build your own design, then i suggest you
>> > read these papers:
>> 
>> Thank you very much for this list. While I have already stumbled upon a
>> few of the FPGA papers, I'm largely ignorant of the other possible
>> approaches. Given the limitations of FPGA TDCs, it will be nice to see
>> what is possible by other means.
>
> I'm quite sure there is much more around. I only dug a little bit a
> weekend or so and got a couple of intersting papers. It is also a 
> good idea, to dig trough old circuit descriptions, form the 70s and older.
> You will find there many forgotten gems, that get increasingly relevant
> when you leave the digital domain, especially when going high speed.
>  
If only there weren't so much to read.

>> That being said, I'm quite keen on bringing up something on the FPGA. I
>> just got the power supplies on the PandaDAQ running last night (QFN is a
>> pain without my shiny new hot air rework station), so it seems that soon
>> enough I'll have a Spartan 6 at my disposal.
>
> *g*
> If you do QFN or any serious SMD stuff, get a Leiter HOT JET S
> with fine nozzles (3mm and 5mm), or anything similar. Normal hot
> air guns don't really work and a complete rework station is way
> too expensive for anything a mere mortal does. On the other hand,
> with such a Hot Jet S (or similar) you can even solder BGAs, reliably.
>
After my first attempt at the QFN with an iron, I broke down and picked
up a rework station for $100. It's not difficult to find them in the $75
range on eBay if you want to go lower. May not be the best station, but
it will serve it's purpose.
>
>> > "Time-Interval Measurements Based on SAW Filter Excitation", by Petr Panek, 2007
>> > "Time interval measurement device based on surface acoustic wave filter
>> > excitation, providing 1ps precision and stability", by Panek and Prochazka, 2007
>> > "Random Erros in Time Interval Measurement Based on SAW Filter Excitation",
>> > by Petr Panek, 2008
>> > A very nice idea on how to use a high frequency startable oscillator with
>> > an ADC as phase detector. Panek claims to get below 1ps with a 200MHz clock
>> > and a 525MHz filter/oscillator. His calculations indicate that the ultimate
>> > limit of resolution is given by the sampling jitter of the ADC and the
>> > frequency and bandwidth (ie the Q of the oscillator). There 
>
> Oops.. "There" is a sentence missing... Interrupts are bad for emails.
>
> There is even a report of a similar design, using an LC tank as resonant
> circuit that got into the <10ps RMS region. See "High frequency, high time
> time resolution time-to-digital converter emplying passive resonant circuits",
> by Ripamonti, Abba and Geracy, 2010
> http://risorse.dei.polimi.it/digital/products/2010/High%20frequency,%20high%20time%20resolution%20time-to-digital%20converter%20employing%20passive%20resonating%20circuits.pdf
>  
After looking at the SAW filter technique a little bit more carefully I
may consider implementing it in the past. It has a nice elegance to
it. That being said, the path of least resistance at this point is to
use the board I have, which lacks the 100MHz ADC. A project for a later
date...

>> > But getting to below that will not be
>> > easy. Mainly due to all those side effect, non-idealities and other
>> > stuff you have to deal with. And be aware, that you are dealing with
>> > an high frequncy/high speed circuit. Crudly said, you are in the
>> > ballpark of a 1/10ps = 100GHz system. Everything has to be right to
>> > get you there.
>> 
>> Sure. This is the real issue. I am a physicist by training, so the
>> basics of high-speed design are largely a mystery to me. From
>> application notes (in particular Jim Williams' old but very readable
>> work) I've gleaned the following,
>> 
>> 1) Keep traces short and well impedence matched
>> 2) Ample bypassing
>> 3) Ground plane is essential
>
> 4) Know where thy return path is!
>
Duly noted.

> I recommend getting the Tietze-Schenk "Halbleiter Schaltungstechnik"
> (resp "Electronic Circuits" in englisch). It's like Horowitz, just
> with more theoretical background and more explenation how to design
> stuff. Ie you get the formulas to calculate what you need if you want
> to go to the limit. It's still very much practical (only as much
> theory as needed) but covers enough of the theory if you want to have
> more than just cookbook examples.
>
Thanks! I'll look into this.

> I'm still looking for a good high speed / high frequency book.
> I've asked for literature in that area a few weeks ago on this
> list, but have not gotten the time yet to read all those books.
> You might want to check the answers too.
>  
>> But beyond these rough guidelines my intuition isn't so well
>> honed. Do you have anything to add?
>
> Unfortunatley not. I'm too young for much experience and don't have
> a big lab to test things. Not to mention the constant lack of time.
> The only thing i have at my disposal are tons of papers i've read
> on train rides... If you've specific questions, i can see whether
> i can dig something up in my collection. But i don't have any general
> "learning" tools yet.
>
>> One final question:
>> In Williams' notes, you often see images of point-to-point wired
>> circuits constructed over a copper clad board ground plane. While in the
>> '90s I can see this being a very reasonable approach, is it still
>> relevant in today's world of surface mount packages and 100GHz
>> bandwidths?  Assuming fly-wiring isn't an option, is there any way to
>> prototype a circuit capable of handling even 50ps signals short of
>> etching a board and hoping for the best?
>
> It's done as well. Just at an other level. You still have these
> copper boards, but you have specialized "pads" for different
> IC packages and you connect them using coax and short copper wires.
> I've never used any of those myself, but i've heard that they are
> good to 1-2GHz at least (mostly due to the good ground). But of
> course you have to know what you are doing when you want do get
> to that performance.
>
> Also forget about 100GHz, that's not something you can do using
> normal of the shelf components. If you get a couple GHz, feel lucky.
> Anything beyond that will not reliably work. The 1/10ps = 100GHz
> was more a figure of speach. Your signals will not have such high
> frequency components. You cannot get slopes that steep to get there
> with normal components and packages. But you have to design your
> circuit thinking you have that frequency because the high frequency
> handling ideas (correct termination, signal path etc) make it easier
> to get where you want to go.
>
Yes, I suppose I should have stopped to think a bit before I
parrotted that number. Fast NIM logic rise times are on the order of
nanoseconds.

Cheers,

- Ben



More information about the time-nuts mailing list