Rectify 500MHz Small Signal
in [Design]
|
From: Joerg on 30 Sep 2009 14:09 John Larkin wrote: > On Wed, 30 Sep 2009 10:24:56 -0700, Joerg <invalid(a)invalid.invalid> > wrote: >>> That stuff works, but it's not as quantitative as the later gear. And >>> TDR is a fabulous thing to have, and the TDR on the 7-series stuff is >>> really mediocre. >>> >>> Isn't this beautiful? >>> >>> ftp://jjlarkin.lmi.net/Z250A.jpg >>> ftp://jjlarkin.lmi.net/Z250_TDR.jpg >>> >>> (TDR of the test trace, J28 to J29) >>> >> That is indeed beautiful. Although, to be honest, I never really had an >> urgent need for TDR. Occasionally I resonate stuff out with a <gasp> dip >> meter. But even there it has happened that I didn't use for a while and >> the battery oozed out. I also have this as a backup: >> >> http://oak.cats.ohiou.edu/~postr/bapix/Dip_59.htm >> >> Knock on wood that the tubes don't go out on me. Of course, when things >> resonate above 400MHz I am up the creek unless I kludge something. > > Trace impedances matter a lot to us. Board houses sometimes take > liberties with multilayer stackups and copper thickness, so we try to > build in test traces for both impedance and resistance when it > matters. That test trace in the pics takes a tour of all 4 board > layers, so we can see all the impedances. It's even more fun on > 8-layer boards. It's a good check for our math, too. > Ok, I understand that. But since you know the geometries from the Gerbers can't you just measure total resistance and capacitance to see if things are within ballpark? > It's also cool to add SMAs to access power pour layers and TDR them. > That blows away all sorts of cockeyed theories about bypassing > high-speed stuff. > Those theories sure are weird :-) > Transitions, like pcb-to-connector, are also interesting to TDR. Leads > to x-acto slashing sometimes. > That is next to impossible to do without TDR. >>>>> There are some little USB samplers, but they're very expensive. >>>>> >>>> Probably not much of a market anymore and the TDR guys in the field >>>> dont' need a precision under a foot to figure where they have to drop >>>> the bucket of their Kubota. >>> Too much cheap surplus stuff on ebay, too. I'd love to do a cheapish >>> USB TDR, but there's probably no good market. >>> >> There probably isn't much of a TDR market. But I believe there would be >> a market for a sampling scope app where people can check (repetitive) >> timings to within several picoseconds. With comm data rates almost in >> the two-digit GHz-range that market should be growing. > > One market for a small, affordable USB TDR would be to sell to board > houses. The Polar Instruments stuff (from the Channel Islands, another > story) is expensive and very dated. > But it sure is beautiful there: http://lisamm.files.wordpress.com/2009/08/guernsey.jpg I could imagine living there, we are not the kind of folkss who'd get cabin fever because of the confines of a smallish island. As long as there are some nice pubs. -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: Nico Coesel on 30 Sep 2009 14:22 John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >On Wed, 30 Sep 2009 16:29:03 GMT, nico(a)puntnl.niks (Nico Coesel) >wrote: > >>John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >> >>>On Tue, 29 Sep 2009 12:53:36 -0700, Joerg <invalid(a)invalid.invalid> >>>wrote: >>> >>> >>>There are some little USB samplers, but they're very expensive. >> >>What amount do you consider expensive? > >$12K is expensive. Expensive indeed. Any idea what makes these buggers so expensive? -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... "If it doesn't fit, use a bigger hammer!" --------------------------------------------------------------
From: Joerg on 30 Sep 2009 14:30 Nico Coesel wrote: > John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: > >> On Wed, 30 Sep 2009 16:29:03 GMT, nico(a)puntnl.niks (Nico Coesel) >> wrote: >> >>> John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >>> >>>> On Tue, 29 Sep 2009 12:53:36 -0700, Joerg <invalid(a)invalid.invalid> >>>> wrote: >>>> >>>> >>>> There are some little USB samplers, but they're very expensive. >>> What amount do you consider expensive? >> $12K is expensive. > > Expensive indeed. Any idea what makes these buggers so expensive? > The usual. High NRE costs, low sales volume, combined with the desire to make a living. You wouldn't believe what I've seen in some electro-optical gear. A couple of <$1 photodiodes, some MMICs or RF amps in the same price range, a couple BNC connectors, a (chintzy and cheap) PVC plastic enclosure, price tags between $500 and $1000. For that money they didn't even include a TVS to guard against the common ESD or "whoops, got the power reversed" accidents in a lab. That would have pushed the cost by at least 50c ... -- Regards, Joerg http://www.analogconsultants.com/ "gmail" domain blocked because of excessive spam. Use another domain or send PM.
From: John Larkin on 30 Sep 2009 15:07 On Wed, 30 Sep 2009 11:09:41 -0700, Joerg <invalid(a)invalid.invalid> wrote: >John Larkin wrote: >> On Wed, 30 Sep 2009 10:24:56 -0700, Joerg <invalid(a)invalid.invalid> >> wrote: >>>> That stuff works, but it's not as quantitative as the later gear. And >>>> TDR is a fabulous thing to have, and the TDR on the 7-series stuff is >>>> really mediocre. >>>> >>>> Isn't this beautiful? >>>> >>>> ftp://jjlarkin.lmi.net/Z250A.jpg >>>> ftp://jjlarkin.lmi.net/Z250_TDR.jpg >>>> >>>> (TDR of the test trace, J28 to J29) >>>> >>> That is indeed beautiful. Although, to be honest, I never really had an >>> urgent need for TDR. Occasionally I resonate stuff out with a <gasp> dip >>> meter. But even there it has happened that I didn't use for a while and >>> the battery oozed out. I also have this as a backup: >>> >>> http://oak.cats.ohiou.edu/~postr/bapix/Dip_59.htm >>> >>> Knock on wood that the tubes don't go out on me. Of course, when things >>> resonate above 400MHz I am up the creek unless I kludge something. >> >> Trace impedances matter a lot to us. Board houses sometimes take >> liberties with multilayer stackups and copper thickness, so we try to >> build in test traces for both impedance and resistance when it >> matters. That test trace in the pics takes a tour of all 4 board >> layers, so we can see all the impedances. It's even more fun on >> 8-layer boards. It's a good check for our math, too. >> > >Ok, I understand that. But since you know the geometries from the >Gerbers can't you just measure total resistance and capacitance to see >if things are within ballpark? Total C would be a very crude way to guess trace impedances. Too ambiguous. > > >> It's also cool to add SMAs to access power pour layers and TDR them. >> That blows away all sorts of cockeyed theories about bypassing >> high-speed stuff. >> > >Those theories sure are weird :-) > > >> Transitions, like pcb-to-connector, are also interesting to TDR. Leads >> to x-acto slashing sometimes. >> > >That is next to impossible to do without TDR. > > >>>>>> There are some little USB samplers, but they're very expensive. >>>>>> >>>>> Probably not much of a market anymore and the TDR guys in the field >>>>> dont' need a precision under a foot to figure where they have to drop >>>>> the bucket of their Kubota. >>>> Too much cheap surplus stuff on ebay, too. I'd love to do a cheapish >>>> USB TDR, but there's probably no good market. >>>> >>> There probably isn't much of a TDR market. But I believe there would be >>> a market for a sampling scope app where people can check (repetitive) >>> timings to within several picoseconds. With comm data rates almost in >>> the two-digit GHz-range that market should be growing. >> >> One market for a small, affordable USB TDR would be to sell to board >> houses. The Polar Instruments stuff (from the Channel Islands, another >> story) is expensive and very dated. >> > >But it sure is beautiful there: > >http://lisamm.files.wordpress.com/2009/08/guernsey.jpg Yup, very nice. It was occupied by the Nazis during the war, and there was a good PBS series dramatizing that strange story. > >I could imagine living there, we are not the kind of folkss who'd get >cabin fever because of the confines of a smallish island. As long as >there are some nice pubs. Tek used to assemble scopes there, for tax reasons. After that stopped making sense, they licensed some of their TDR technology to Polar so that the jobs wouldn't disappear. But it's been a while and the technology is getting rusty. John
From: John Larkin on 30 Sep 2009 15:10
On Wed, 30 Sep 2009 11:30:28 -0700, Joerg <invalid(a)invalid.invalid> wrote: >Nico Coesel wrote: >> John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >> >>> On Wed, 30 Sep 2009 16:29:03 GMT, nico(a)puntnl.niks (Nico Coesel) >>> wrote: >>> >>>> John Larkin <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote: >>>> >>>>> On Tue, 29 Sep 2009 12:53:36 -0700, Joerg <invalid(a)invalid.invalid> >>>>> wrote: >>>>> >>>>> >>>>> There are some little USB samplers, but they're very expensive. >>>> What amount do you consider expensive? >>> $12K is expensive. >> >> Expensive indeed. Any idea what makes these buggers so expensive? >> > >The usual. High NRE costs, low sales volume, combined with the desire to >make a living. > >You wouldn't believe what I've seen in some electro-optical gear. A >couple of <$1 photodiodes, some MMICs or RF amps in the same price >range, a couple BNC connectors, a (chintzy and cheap) PVC plastic >enclosure, price tags between $500 and $1000. For that money they didn't >even include a TVS to guard against the common ESD or "whoops, got the >power reversed" accidents in a lab. That would have pushed the cost by >at least 50c ... We, at least, use metal: http://www.highlandtechnology.com/DSS/J740DS.html John |