hsc@lanl.ARPA (07/18/85)
I am in need of a flash ADC with about a 100 msps conversion rate and eight bit minimum resolution. A converter with a low power drain would be nice also, since the project this ADC will be used in will use about 1200 of them. Any ideas?? Thanks in advance, Scott (the sorcerer's apprentice) Chesney hsc@lanl.ARPA
crs@lanl.ARPA (07/19/85)
> > I am in need of a flash ADC with about a 100 msps conversion rate and > eight bit minimum resolution. A converter with a low power drain would > be nice also, since the project this ADC will be used in will use about > 1200 of them. Any ideas?? > > Thanks in advance, > > Scott (the sorcerer's apprentice) Chesney > hsc@lanl.ARPA I apologize for not replying by mail but my success rate has not been that good lately by that route. I just noticed such a device yesterday in the 15 July issue of _Electronic_Products_ in the "EP/IC Update section (p 91). The one mentioned is the Siemens SDA 8010 (contact Janet Hartley 201-321-8788) 186 Wood Ave. S., Iselin, NJ 08830. Oops! Company's full name is Siemens Communications System Inc. I'm not personally familiar with the product but here are the summary specifications from the magazine: 8 bits. Sampling frequency of 100 MHz; signal bandwidth of 50 MHz; 1.3-W power dissipation; +-1/2 lsb linearity. ECL compatible. 24 pin ceramic DIP. You may also want to enquire of RCA and TRW Semiconductor; I know that they make flash adc's. -- All opinions are mine alone... Charlie Sorsby ...!{cmcl2,ihnp4,...}!lanl!crs crs@lanl.arpa
dsi@unccvax.UUCP (Dataspan Inc) (07/20/85)
You don't want much, do you? 100 mHz, 8 bits, and low power consumption?
RCA's SOS (silicon on sapphire) 8 bit converter can barely hit 14.318 mc;
in fact, this product may be stillborn.
The TRW TDC1007 (and kin) aren't even fast enough. Do what we do. . .
get a bunch of TDC1025's (which are 50 mHz conversion rate) and drive them
with opposite phase clocks. The aperture time of a TDC1025 is 40 picoseconds,
which is excellent; worst case time multiplexing error is still about 15 times
less than the sampling frequency required.
YOU NEED A SAMPLE AND HOLD AMPLIFIER IN FRONT OF THIS THING TO MAKE IT
WORK RIGHT. Even at 50 mHz clock frequency, the full power guaranteed conversion
frequency is only 12.5 mHz. However, you can probably approach theoretical
limits with an Analog Devices HTS-0010 S/H amplifier; it has a 5 ps aperture
error.
Have fun designing the driver amplifier for two A/D converters; the
input impedance is highly flaky on these converters. Those high frequencies
will bring a new meaning to the term 'thermal tail'... about power consumption,
well, two TDC1025's eat a bunch of power. Be prepared to dissipate about
8 watts of heat from the cases.
Another alternative (for the strong-stomached) is to use 4 TDC1029
6 bit converters in the appropriate arrangement. They do 100 mHz and are
guaranteed to do full power quantization at 50 mHz.
Finally, Sony supposedly has a 100 mHz 8-bit A/D in a single package,
but I've yet to see a data sheet.
I wish you lots of luck. Even Computer Labs (Analog Devices) hasn't
taken this one on (commercially) yet. I hope you also have a very big
budget for all the neat test equipment you are going to have to build and
buy.
David Anthony
Chief Development Engineer
DataSpan, Inc (the Southeast's largest user of TRW A/D converters!)
.mike@amdcad.UUCP (Mike Parker) (07/23/85)
In article <28485@lanl.ARPA>, crs@lanl.ARPA writes: > > You may also want to enquire of RCA and TRW Semiconductor; I know that > they make flash adc's. > -- Also try Telmos in Sunnyvale, I doubt that they can do 100Mhz but they might and power consumption would be lower. Mike