On Tue, 18 Feb 2014 10:30:41 -0600, John S <Sophi.2@invalid.org> wrote:>On 2/18/2014 9:18 AM, John Larkin wrote: >> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> >> wrote: >> >>> On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: >>>> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: >>>> >>> <snip> >>>> This is the classic 2-diode feedback sampler: >>>> >>>> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG >>> Thanks! I didn't really understand what you guys were talking about. >>> >>> So the diodes are kept "almost on" by the feedback? >> No, they are back-biased by a volt or so. The bias tracks the last-sampled >> value. >> >>> Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? >> The signal level info is actually stored in C1 and C2. If the signal (at the >> sampling instant) is equal to the feedback voltage, C1 and C2 receive equal >> amounts of charge and there is no net signal into the amps. If not, one of the >> caps charges more than the other, and you get a signed glitch to work with. The >> voltage glitch at A may be just a few per cent of the difference between the >> signal level and the output... low "sampling efficiency." The feedback loop >> boosts the signal and makes everything linear. >> >> Joerg wants to use a full-bridge sampler with no feedback. >> >> >Hi, John - > >Are those transmission lines where it looks like the pulse generator >output is shorted to ground? > >Thanks, >John SYes. Like this: https://dl.dropboxusercontent.com/u/53724080/Sampling/Sampler1.JPG That was intended to be a dual-channel sampler, but I never built the one on the right. There's a little ceramic pill SRD in the center of the pattern, fed by the twisted pair from below. -- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation
Sampler diodes with more barrier height?
Started by ●February 15, 2014
Reply by ●February 18, 20142014-02-18
Reply by ●February 18, 20142014-02-18
On Tuesday, February 18, 2014 10:18:21 AM UTC-5, John Larkin wrote:> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> > > wrote: > > > > >On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: > > >> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: > >> > ><snip> > >> > >> This is the classic 2-diode feedback sampler: > >> > >> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG > > > >Thanks! I didn't really understand what you guys were talking about. > > > >So the diodes are kept "almost on" by the feedback? > > No, they are back-biased by a volt or so. The bias tracks the last-sampled > value. > > > >Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? > > The signal level info is actually stored in C1 and C2. If the signal (at the > sampling instant) is equal to the feedback voltage, C1 and C2 receive equal > amounts of charge and there is no net signal into the amps. If not, one of > caps charges more than the other, and you get a signed glitch to work with. The > > voltage glitch at A may be just a few per cent of the difference between the > signal level and the output... low "sampling efficiency." The feedback loop > boosts the signal and makes everything linear.Thanks for the description... I clearly have no clue how it works :^) George H.> > > Joerg wants to use a full-bridge sampler with no feedback. > > > > > > -- > > > > John Larkin Highland Technology Inc > > www.highlandtechnology.com jlarkin at highlandtechnology dot com > > > > Precision electronic instrumentation
Reply by ●February 18, 20142014-02-18
On 02/18/2014 01:14 PM, George Herold wrote:> On Tuesday, February 18, 2014 10:18:21 AM UTC-5, John Larkin wrote: >> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> >> >> wrote: >> >> >> >>> On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: >> >>>> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: >>>> >>> <snip> >>>> >>>> This is the classic 2-diode feedback sampler: >>>> >>>> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG >>> >>> Thanks! I didn't really understand what you guys were talking about. >>> >>> So the diodes are kept "almost on" by the feedback? >> >> No, they are back-biased by a volt or so. The bias tracks the last-sampled >> value. >> >> >>> Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? >> >> The signal level info is actually stored in C1 and C2. If the signal (at the >> sampling instant) is equal to the feedback voltage, C1 and C2 receive equal >> amounts of charge and there is no net signal into the amps. If not, one of >> caps charges more than the other, and you get a signed glitch to work with. The >> >> voltage glitch at A may be just a few per cent of the difference between the >> signal level and the output... low "sampling efficiency." The feedback loop >> boosts the signal and makes everything linear. > > Thanks for the description... I clearly have no clue how it works :^) > George H.The Tektronix Concepts sampling book is a really good read. S2 sampler vintage rather than SD-24, but oh well--there's still a lot of good stuff there, including an excellent description of how sampling loops work. http://electrooptical.net/Tektronix/062-1172-00_TekSamplingCircuitsMar70.pdf 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
Reply by ●February 18, 20142014-02-18
On Mon, 17 Feb 2014 16:28:29 -0800, Joerg <invalid@invalid.invalid> wrote:>Phil Hobbs wrote: >> On 2/17/2014 3:34 PM, Joerg wrote: >>> Phil Hobbs wrote: >>>> On 02/17/2014 12:39 PM, Joerg wrote: >>>>> Phil Hobbs wrote: >>>>>> On 02/17/2014 11:25 AM, Phil Hobbs wrote: > >[...] > >>>>>>> >>>>>>> Assuming that one has a sufficiently accurate model of diode >>>>>>> behaviour >>>>>>> under dynamic conditions. Feedback is nice that way. >>>>>>> >>>>>>> A BF908 has a reverse transfer capacitance (G1-D) of typically >>>>>>> 0.03 pF. >>>>>>> The G2-D capacitance would be more of a worry, but one more >>>>>>> cascode >>>>>>> stage should fix that. >>>>>> >>>>>> Or run G2 from the buffer output, which would also fix it. Have to >>>>>> try >>>>>> that. >>>>>> >>>>> >>>>> True, but if you need a buffer or a charge amp with a beefy output you >>>>> might as well use a Schottky diode there, too :-) >>>>> >>>> >>>> It only has to drive a 3-pF capacitance, and it's 3 pF to the input and >>>> not to the sample pulse. Plus when it's off, it's really off. But that >>>> part will run out of steam above a few hundred megahertz. >>>> >>>> My original suggestion was a cascode made out of two pHEMTs, e.g. >>>> SKY65050s. It would have much the same virtues at much lower >>>> capacitance, and has gain up to 12 GHz or so. The Skyworks ones are >>>> much nicer than the Avago ones for this sort of job, because their >>>> output conductance is a lot lower, i.e. they have a really high Early >>>> voltage. >>>> >>> >>> They are nice but most likely they'd have to be adjusted in production >>> for the upper gate bias (where the to-be-sampled signals would go in). >>> The datasheet mentions about 30% total tolrance in the pinchoff. I guess >>> I could automate that. The other thing is whether it'll be stable with >>> one riding on the drain of the other. With GHz devices that can be like >>> trying to balance a garden hose on the tip of a finger. >> >> I've used them as the bottom of a cascode with a BFP640 SiGe:C on top. >> The BFP640 is a 40-GHz device, but works fine with a 5-ohm bead in its >> base lead. I haven't tried a pair of SKY65050s. >> >> One especially nice thing about them is that there's a particular >> voltage (within operating bias) where the gate current goes to zero. I >> don't know how consistent that is from device to device, but in the >> devices I have, the leakage definitely changes sign somewhere in the >> normal bias region. There are lots of things you can do with that. >> > >Probably the leakage is very low to begin with. Unlike the Schottky >diodes with low barriers. > >John: Just to to Skyworks and the are sending me samplers of their >DMJ2824. Around 0.1pF, amazing. Seems those are the only ones with high >Schottky barrier and thus their Vf is more like a silicon diode. Big >downside is that they cost a ton.Ask them, or RFMW (or both!) for the Skyworks diode sample kit. Varicaps, PINs, schottkies. No quads. I also have a dusty loosleaf binder, HP RF/uW Diode Designers Kit, full of data sheets and samples. I could loan you both if you promise to not use *all* of the samples. -- John Larkin Highland Technology, Inc jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Reply by ●February 18, 20142014-02-18
On 2/18/2014 10:41 AM, John Larkin wrote:> On Tue, 18 Feb 2014 10:30:41 -0600, John S <Sophi.2@invalid.org> wrote: > >> On 2/18/2014 9:18 AM, John Larkin wrote: >>> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> >>> wrote: >>> >>>> On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: >>>>> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: >>>>> >>>> <snip> >>>>> This is the classic 2-diode feedback sampler: >>>>> >>>>> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG >>>> Thanks! I didn't really understand what you guys were talking about. >>>> >>>> So the diodes are kept "almost on" by the feedback? >>> No, they are back-biased by a volt or so. The bias tracks the last-sampled >>> value. >>> >>>> Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? >>> The signal level info is actually stored in C1 and C2. If the signal (at the >>> sampling instant) is equal to the feedback voltage, C1 and C2 receive equal >>> amounts of charge and there is no net signal into the amps. If not, one of the >>> caps charges more than the other, and you get a signed glitch to work with. The >>> voltage glitch at A may be just a few per cent of the difference between the >>> signal level and the output... low "sampling efficiency." The feedback loop >>> boosts the signal and makes everything linear. >>> >>> Joerg wants to use a full-bridge sampler with no feedback. >>> >>> >> Hi, John - >> >> Are those transmission lines where it looks like the pulse generator >> output is shorted to ground? >> >> Thanks, >> John S > Yes. Like this: > > https://dl.dropboxusercontent.com/u/53724080/Sampling/Sampler1.JPG > > That was intended to be a dual-channel sampler, but I never built the one on the > right. There's a little ceramic pill SRD in the center of the pattern, fed by > the twisted pair from below.Thanks much, John. Cheers, John S
Reply by ●February 18, 20142014-02-18
John Larkin wrote:> On Tue, 18 Feb 2014 07:58:39 -0800, Joerg <invalid@invalid.invalid> wrote: > >> John Larkin wrote: >>> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> >>> wrote: >>> >>>> On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: >>>>> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: >>>>> >>>> <snip> >>>>> This is the classic 2-diode feedback sampler: >>>>> >>>>> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG >>>> Thanks! I didn't really understand what you guys were talking about. >>>> >>>> So the diodes are kept "almost on" by the feedback? >>> No, they are back-biased by a volt or so. The bias tracks the last-sampled >>> value. >>> >>>> Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? >>> The signal level info is actually stored in C1 and C2. If the signal (at the >>> sampling instant) is equal to the feedback voltage, C1 and C2 receive equal >>> amounts of charge and there is no net signal into the amps. If not, one of the >>> caps charges more than the other, and you get a signed glitch to work with. The >>> voltage glitch at A may be just a few per cent of the difference between the >>> signal level and the output... low "sampling efficiency." The feedback loop >>> boosts the signal and makes everything linear. >>> >>> Joerg wants to use a full-bridge sampler with no feedback. >>> >> Probably just single-diode, the full bridge needs too much "stuff" for >> this design and there is no negative supply. Linearity isn't a concern >> and, provided each delta-V step is at least several millivolts, there >> isn't much to be gained in SNR by a charge accumulation method versus >> just averaging or lowpass-filtering these delta-Vs into an RC or >> something. Of course, then the sampling pulse must come from a very >> clean source. >> >> Now I just need a high Schottky barrier (or at least middle barrier) >> diode that does not leak so much and won't cost an arm and a leg. The >> latter seems to be a problem. Worst case I'll have to live with a >> solution where nearly all the charge leaks out during each cycle and I >> average the resulting "sawteeth". Lowers the SNR but not by much. >> >> I wonder how they do that on the cheap cable testers. > > Cable TDRs are slow, aren't they? > > The Tek field-level TDRs (1500 series) had full-bridge sampler front-ends, with > risetimes in the hundreds of ps. > > http://www.ko4bb.com/Manuals/Tektronix/Tektronix_-_1503C_Metallic_Time_Domain_Reflectometer/TEK%201503C%20Service_EN.pdf >2nsec half sine? <yawn> In my book that ain't real TDR yet :-)> Story: Tek once had a scope assembly plant in Guernsey, Channel Islands, for > european tax reasons. (Channel Islands has a unique legal status and was > occupied by the Nazis in WWII.) Later things changed, so they pulled out. But > they let the islanders start up Polar Instruments > > http://www.polarinstruments.com/ > > Last time I checked, they were still using the 15xx-series TDR boards, which are > actually slow for PCB work. >Not sure if Tek still makes those. But there's whole industries around legacy gear. For example, companies that still make spare parts for the DC-3 (a.k.a. "The Goon"), where many of the still flying aircraft are now around 70 years old. And counting.> Good book: > > https://en.wikipedia.org/wiki/The_Guernsey_Literary_and_Potato_Peel_Pie_Society >:-) -- Regards, Joerg http://www.analogconsultants.com/
Reply by ●February 18, 20142014-02-18
John Larkin wrote:> On Mon, 17 Feb 2014 16:28:29 -0800, Joerg <invalid@invalid.invalid> > wrote: > >> Phil Hobbs wrote: >>> On 2/17/2014 3:34 PM, Joerg wrote: >>>> Phil Hobbs wrote: >>>>> On 02/17/2014 12:39 PM, Joerg wrote: >>>>>> Phil Hobbs wrote: >>>>>>> On 02/17/2014 11:25 AM, Phil Hobbs wrote: >> [...] >> >>>>>>>> Assuming that one has a sufficiently accurate model of diode >>>>>>>> behaviour >>>>>>>> under dynamic conditions. Feedback is nice that way. >>>>>>>> >>>>>>>> A BF908 has a reverse transfer capacitance (G1-D) of typically >>>>>>>> 0.03 pF. >>>>>>>> The G2-D capacitance would be more of a worry, but one more >>>>>>>> cascode >>>>>>>> stage should fix that. >>>>>>> Or run G2 from the buffer output, which would also fix it. Have to >>>>>>> try >>>>>>> that. >>>>>>> >>>>>> True, but if you need a buffer or a charge amp with a beefy output you >>>>>> might as well use a Schottky diode there, too :-) >>>>>> >>>>> It only has to drive a 3-pF capacitance, and it's 3 pF to the input and >>>>> not to the sample pulse. Plus when it's off, it's really off. But that >>>>> part will run out of steam above a few hundred megahertz. >>>>> >>>>> My original suggestion was a cascode made out of two pHEMTs, e.g. >>>>> SKY65050s. It would have much the same virtues at much lower >>>>> capacitance, and has gain up to 12 GHz or so. The Skyworks ones are >>>>> much nicer than the Avago ones for this sort of job, because their >>>>> output conductance is a lot lower, i.e. they have a really high Early >>>>> voltage. >>>>> >>>> They are nice but most likely they'd have to be adjusted in production >>>> for the upper gate bias (where the to-be-sampled signals would go in). >>>> The datasheet mentions about 30% total tolrance in the pinchoff. I guess >>>> I could automate that. The other thing is whether it'll be stable with >>>> one riding on the drain of the other. With GHz devices that can be like >>>> trying to balance a garden hose on the tip of a finger. >>> I've used them as the bottom of a cascode with a BFP640 SiGe:C on top. >>> The BFP640 is a 40-GHz device, but works fine with a 5-ohm bead in its >>> base lead. I haven't tried a pair of SKY65050s. >>> >>> One especially nice thing about them is that there's a particular >>> voltage (within operating bias) where the gate current goes to zero. I >>> don't know how consistent that is from device to device, but in the >>> devices I have, the leakage definitely changes sign somewhere in the >>> normal bias region. There are lots of things you can do with that. >>> >> Probably the leakage is very low to begin with. Unlike the Schottky >> diodes with low barriers. >> >> John: Just to to Skyworks and the are sending me samplers of their >> DMJ2824. Around 0.1pF, amazing. Seems those are the only ones with high >> Schottky barrier and thus their Vf is more like a silicon diode. Big >> downside is that they cost a ton. > > Ask them, or RFMW (or both!) for the Skyworks diode sample kit. > Varicaps, PINs, schottkies. No quads. > > I also have a dusty loosleaf binder, HP RF/uW Diode Designers Kit, > full of data sheets and samples. > > I could loan you both if you promise to not use *all* of the samples. >Oh, I would promise. But I don't need a whole kit right now, essentially I just have to qualify one good diode type and design that in. From the discussion with their engineer yesterday it pretty much boils down to the DMJ series. There seem to be no more high barrier diodes available in coach class these days. Unless they are at other mfgs such as Avago, which unfortunately is not as responsive as Skyworks. -- Regards, Joerg http://www.analogconsultants.com/
Reply by ●February 18, 20142014-02-18
On Tue, 18 Feb 2014 11:59:02 -0800, Joerg <invalid@invalid.invalid> wrote:>John Larkin wrote: >> On Tue, 18 Feb 2014 07:58:39 -0800, Joerg <invalid@invalid.invalid> wrote: >> >>> John Larkin wrote: >>>> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> >>>> wrote: >>>> >>>>> On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: >>>>>> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: >>>>>> >>>>> <snip> >>>>>> This is the classic 2-diode feedback sampler: >>>>>> >>>>>> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG >>>>> Thanks! I didn't really understand what you guys were talking about. >>>>> >>>>> So the diodes are kept "almost on" by the feedback? >>>> No, they are back-biased by a volt or so. The bias tracks the last-sampled >>>> value. >>>> >>>>> Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? >>>> The signal level info is actually stored in C1 and C2. If the signal (at the >>>> sampling instant) is equal to the feedback voltage, C1 and C2 receive equal >>>> amounts of charge and there is no net signal into the amps. If not, one of the >>>> caps charges more than the other, and you get a signed glitch to work with. The >>>> voltage glitch at A may be just a few per cent of the difference between the >>>> signal level and the output... low "sampling efficiency." The feedback loop >>>> boosts the signal and makes everything linear. >>>> >>>> Joerg wants to use a full-bridge sampler with no feedback. >>>> >>> Probably just single-diode, the full bridge needs too much "stuff" for >>> this design and there is no negative supply. Linearity isn't a concern >>> and, provided each delta-V step is at least several millivolts, there >>> isn't much to be gained in SNR by a charge accumulation method versus >>> just averaging or lowpass-filtering these delta-Vs into an RC or >>> something. Of course, then the sampling pulse must come from a very >>> clean source. >>> >>> Now I just need a high Schottky barrier (or at least middle barrier) >>> diode that does not leak so much and won't cost an arm and a leg. The >>> latter seems to be a problem. Worst case I'll have to live with a >>> solution where nearly all the charge leaks out during each cycle and I >>> average the resulting "sawteeth". Lowers the SNR but not by much. >>> >>> I wonder how they do that on the cheap cable testers. >> >> Cable TDRs are slow, aren't they? >> >> The Tek field-level TDRs (1500 series) had full-bridge sampler front-ends, with >> risetimes in the hundreds of ps. >> >> http://www.ko4bb.com/Manuals/Tektronix/Tektronix_-_1503C_Metallic_Time_Domain_Reflectometer/TEK%201503C%20Service_EN.pdf >> > >2nsec half sine? <yawn> > >In my book that ain't real TDR yet :-) > > >> Story: Tek once had a scope assembly plant in Guernsey, Channel Islands, for >> european tax reasons. (Channel Islands has a unique legal status and was >> occupied by the Nazis in WWII.) Later things changed, so they pulled out. But >> they let the islanders start up Polar Instruments >> >> http://www.polarinstruments.com/ >> >> Last time I checked, they were still using the 15xx-series TDR boards, which are >> actually slow for PCB work. >> > >Not sure if Tek still makes those. But there's whole industries around >legacy gear. For example, companies that still make spare parts for the >DC-3 (a.k.a. "The Goon"), where many of the still flying aircraft are >now around 70 years old. And counting. > > >> Good book: >> >> https://en.wikipedia.org/wiki/The_Guernsey_Literary_and_Potato_Peel_Pie_Society >> > >:-)There is, in theory, some B-52 pilot flying the same plane that his grandfather flew. Those planes are scheduled to be retired when they are 80 years old. My first airplane flight was in the mil version of the DC-3, a horrible noisy sweat box with no interior and web seats. It got us from New Orleans to Charleston in 14 hours. (That's when I was in the Navy for a week.) -- John Larkin Highland Technology, Inc jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Reply by ●February 18, 20142014-02-18
John Larkin wrote:> On Tue, 18 Feb 2014 11:59:02 -0800, Joerg <invalid@invalid.invalid> > wrote: > >> John Larkin wrote: >>> On Tue, 18 Feb 2014 07:58:39 -0800, Joerg <invalid@invalid.invalid> wrote: >>> >>>> John Larkin wrote: >>>>> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> >>>>> wrote: >>>>> >>>>>> On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: >>>>>>> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: >>>>>>> >>>>>> <snip> >>>>>>> This is the classic 2-diode feedback sampler: >>>>>>> >>>>>>> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG >>>>>> Thanks! I didn't really understand what you guys were talking about. >>>>>> >>>>>> So the diodes are kept "almost on" by the feedback? >>>>> No, they are back-biased by a volt or so. The bias tracks the last-sampled >>>>> value. >>>>> >>>>>> Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? >>>>> The signal level info is actually stored in C1 and C2. If the signal (at the >>>>> sampling instant) is equal to the feedback voltage, C1 and C2 receive equal >>>>> amounts of charge and there is no net signal into the amps. If not, one of the >>>>> caps charges more than the other, and you get a signed glitch to work with. The >>>>> voltage glitch at A may be just a few per cent of the difference between the >>>>> signal level and the output... low "sampling efficiency." The feedback loop >>>>> boosts the signal and makes everything linear. >>>>> >>>>> Joerg wants to use a full-bridge sampler with no feedback. >>>>> >>>> Probably just single-diode, the full bridge needs too much "stuff" for >>>> this design and there is no negative supply. Linearity isn't a concern >>>> and, provided each delta-V step is at least several millivolts, there >>>> isn't much to be gained in SNR by a charge accumulation method versus >>>> just averaging or lowpass-filtering these delta-Vs into an RC or >>>> something. Of course, then the sampling pulse must come from a very >>>> clean source. >>>> >>>> Now I just need a high Schottky barrier (or at least middle barrier) >>>> diode that does not leak so much and won't cost an arm and a leg. The >>>> latter seems to be a problem. Worst case I'll have to live with a >>>> solution where nearly all the charge leaks out during each cycle and I >>>> average the resulting "sawteeth". Lowers the SNR but not by much. >>>> >>>> I wonder how they do that on the cheap cable testers. >>> Cable TDRs are slow, aren't they? >>> >>> The Tek field-level TDRs (1500 series) had full-bridge sampler front-ends, with >>> risetimes in the hundreds of ps. >>> >>> http://www.ko4bb.com/Manuals/Tektronix/Tektronix_-_1503C_Metallic_Time_Domain_Reflectometer/TEK%201503C%20Service_EN.pdf >>> >> 2nsec half sine? <yawn> >> >> In my book that ain't real TDR yet :-) >> >> >>> Story: Tek once had a scope assembly plant in Guernsey, Channel Islands, for >>> european tax reasons. (Channel Islands has a unique legal status and was >>> occupied by the Nazis in WWII.) Later things changed, so they pulled out. But >>> they let the islanders start up Polar Instruments >>> >>> http://www.polarinstruments.com/ >>> >>> Last time I checked, they were still using the 15xx-series TDR boards, which are >>> actually slow for PCB work. >>> >> Not sure if Tek still makes those. But there's whole industries around >> legacy gear. For example, companies that still make spare parts for the >> DC-3 (a.k.a. "The Goon"), where many of the still flying aircraft are >> now around 70 years old. And counting. >> >> >>> Good book: >>> >>> https://en.wikipedia.org/wiki/The_Guernsey_Literary_and_Potato_Peel_Pie_Society >>> >> :-) > > There is, in theory, some B-52 pilot flying the same plane that his > grandfather flew. Those planes are scheduled to be retired when they > are 80 years old. >Sometimes I do designs for aerospace. That's a whole different world. A lot of stuff is expected to last 30 years, 30-40% of which is spent in the air.> My first airplane flight was in the mil version of the DC-3, a > horrible noisy sweat box with no interior and web seats. It got us > from New Orleans to Charleston in 14 hours. (That's when I was in the > Navy for a week.) >A week? Did you quit after that experience? :-) I like old airplanes. My best flight was in a vintage Boeing Stearman but with a more powerful engine than standard. Open cockpit, goggles, leather cap. The pilot asked whether I was squeamish about doing "some stuff". I sad no, that I had parachuted, so no worries. "Alright then! Let's go south of the highway where there's no houses, in case we screw up". I am glad my wife did not hear this. The best part was when he said "Your aircraft ... take us back to the airport", let go of stick and rudder and leaned back. Afterwards my T-shirt had a distinct oil smell from all the splatters that flew past the postcard sized "windshield". I asked my wife not to wash it for a few days. -- Regards, Joerg http://www.analogconsultants.com/
Reply by ●February 18, 20142014-02-18
Den tirsdag den 18. februar 2014 21.54.21 UTC+1 skrev Joerg:> John Larkin wrote: > > > On Tue, 18 Feb 2014 11:59:02 -0800, Joerg <invalid@invalid.invalid> > > > wrote: > > > > > >> John Larkin wrote: > > >>> On Tue, 18 Feb 2014 07:58:39 -0800, Joerg <invalid@invalid.invalid> wrote: > > >>> > > >>>> John Larkin wrote: > > >>>>> On Tue, 18 Feb 2014 06:41:56 -0800 (PST), George Herold <gherold@teachspin.com> > > >>>>> wrote: > > >>>>> > > >>>>>> On Sunday, February 16, 2014 7:07:03 PM UTC-5, John Larkin wrote: > > >>>>>>> On Sun, 16 Feb 2014 15:38:43 -0800, Joerg <invalid@invalid.invalid> wrote: > > >>>>>>> > > >>>>>> <snip> > > >>>>>>> This is the classic 2-diode feedback sampler: > > >>>>>>> > > >>>>>>> https://dl.dropboxusercontent.com/u/53724080/Sampling/2-diode-sampler.JPG > > >>>>>> Thanks! I didn't really understand what you guys were talking about. > > >>>>>> > > >>>>>> So the diodes are kept "almost on" by the feedback? > > >>>>> No, they are back-biased by a volt or so. The bias tracks the last-sampled > > >>>>> value. > > >>>>> > > >>>>>> Then with the sampling pulse they (the diodes) are like switches and transmit the input voltage to the caps? > > >>>>> The signal level info is actually stored in C1 and C2. If the signal (at the > > >>>>> sampling instant) is equal to the feedback voltage, C1 and C2 receive equal > > >>>>> amounts of charge and there is no net signal into the amps. If not, one of the > > >>>>> caps charges more than the other, and you get a signed glitch to work with. The > > >>>>> voltage glitch at A may be just a few per cent of the difference between the > > >>>>> signal level and the output... low "sampling efficiency." The feedback loop > > >>>>> boosts the signal and makes everything linear. > > >>>>> > > >>>>> Joerg wants to use a full-bridge sampler with no feedback. > > >>>>> > > >>>> Probably just single-diode, the full bridge needs too much "stuff" for > > >>>> this design and there is no negative supply. Linearity isn't a concern > > >>>> and, provided each delta-V step is at least several millivolts, there > > >>>> isn't much to be gained in SNR by a charge accumulation method versus > > >>>> just averaging or lowpass-filtering these delta-Vs into an RC or > > >>>> something. Of course, then the sampling pulse must come from a very > > >>>> clean source. > > >>>> > > >>>> Now I just need a high Schottky barrier (or at least middle barrier) > > >>>> diode that does not leak so much and won't cost an arm and a leg. The > > >>>> latter seems to be a problem. Worst case I'll have to live with a > > >>>> solution where nearly all the charge leaks out during each cycle and I > > >>>> average the resulting "sawteeth". Lowers the SNR but not by much. > > >>>> > > >>>> I wonder how they do that on the cheap cable testers. > > >>> Cable TDRs are slow, aren't they? > > >>> > > >>> The Tek field-level TDRs (1500 series) had full-bridge sampler front-ends, with > > >>> risetimes in the hundreds of ps. > > >>> > > >>> http://www.ko4bb.com/Manuals/Tektronix/Tektronix_-_1503C_Metallic_Time_Domain_Reflectometer/TEK%201503C%20Service_EN.pdf > > >>> > > >> 2nsec half sine? <yawn> > > >> > > >> In my book that ain't real TDR yet :-) > > >> > > >> > > >>> Story: Tek once had a scope assembly plant in Guernsey, Channel Islands, for > > >>> european tax reasons. (Channel Islands has a unique legal status and was > > >>> occupied by the Nazis in WWII.) Later things changed, so they pulled out. But > > >>> they let the islanders start up Polar Instruments > > >>> > > >>> http://www.polarinstruments.com/ > > >>> > > >>> Last time I checked, they were still using the 15xx-series TDR boards, which are > > >>> actually slow for PCB work. > > >>> > > >> Not sure if Tek still makes those. But there's whole industries around > > >> legacy gear. For example, companies that still make spare parts for the > > >> DC-3 (a.k.a. "The Goon"), where many of the still flying aircraft are > > >> now around 70 years old. And counting. > > >> > > >> > > >>> Good book: > > >>> > > >>> https://en.wikipedia.org/wiki/The_Guernsey_Literary_and_Potato_Peel_Pie_Society > > >>> > > >> :-) > > > > > > There is, in theory, some B-52 pilot flying the same plane that his > > > grandfather flew. Those planes are scheduled to be retired when they > > > are 80 years old. > > > > > > > Sometimes I do designs for aerospace. That's a whole different world. A > > lot of stuff is expected to last 30 years, 30-40% of which is spent in > > the air. > > > > > > > My first airplane flight was in the mil version of the DC-3, a > > > horrible noisy sweat box with no interior and web seats. It got us > > > from New Orleans to Charleston in 14 hours. (That's when I was in the > > > Navy for a week.) > > > > > > > A week? Did you quit after that experience? :-) > > > > I like old airplanes. My best flight was in a vintage Boeing Stearman > > but with a more powerful engine than standard. Open cockpit, goggles, > > leather cap. The pilot asked whether I was squeamish about doing "some > > stuff". I sad no, that I had parachuted, so no worries. "Alright then! > > Let's go south of the highway where there's no houses, in case we screw > > up". I am glad my wife did not hear this. The best part was when he said > > "Your aircraft ... take us back to the airport", let go of stick and > > rudder and leaned back. Afterwards my T-shirt had a distinct oil smell > > from all the splatters that flew past the postcard sized "windshield". I > > asked my wife not to wash it for a few days.seem to remember stories of old WWI pilots all having diarrhea from the total-loss oiling system with castor oil :P -Lasse
