OSC and unity gain buffer
in [Design]
|
Prev: A/D converter
Next: power flap detector
From: Joel Koltner on 10 Apr 2008 12:43 "eeaj2002" <eeaj2002(a)gmail.com> wrote in message news:1173a51d-1a2b-4f72-9685-8e821a8a05df(a)q27g2000prf.googlegroups.com... "Thank you for your hint but this is not a homework problem. I was just trying to learn about opamp myself." OK... besides the usual suspects such as The Art of Electronics, you might want to read something like, "Op Amps for Everyone" by Ron Mancini. It's available as a printed book from Amazon, although it's also available freely for download at http://focus.ti.com/lit/an/slod006b/slod006b.pdf . "Intuitive IC Op Amps" by Frederiksen is also quite good, although it's out of print, having been published in 1984. (I keep meaning to scan it in one of these days and hope that the fine folks at National Semiconductor don't mind if I post it...) "Are you saying that its slew rate must be greater than 1.4V/us?" Yep! The "hint" was to suggest how you get the necessary slew rate... at 1V, 1Hz you need an op-amp with a 6.28 (2*pi) V/s slew rate, so at 3V, 1MHz you're up to 2*pi*3*1e6 = 18.8V/us -- well beyond what your LMC6484 can achieve. ---Joel
From: Tim Wescott on 10 Apr 2008 12:46 eeaj2002 wrote: > > Hi all, > > I need help with my circuit please. I have a one MHz osc fed into the > positive input of the opamp LMC6484 and the negative terminal is > connected to the output (unity gain buffer). At the positive input > terminal I have the 1MHz. from the oscillator and I should see one MHz > at the output as well but I see distorted sine wave signal. If the > gain bandwidth of the opamp is 1.5MHz. should not I see one MHz. at > the output of the opamp as well. I did run the oscillator at 3.0 > Volts and the opamp at 15V. > > Thanks for you help in advance, > > John. You really don't want to run an op-amp right up against it's gain-bandwidth product. The design philosophy of an op-amp is that you use a crappy amplifier with tons of excess gain, and fix all of the amplifier problems with tons of feedback. When you get close to the loop bandwidth, you no longer have tons of excess gain with which to generate tons of feedback, and all you're left with is a crappy amplifier. I'd be looking for an op-amp with a G-B product of 10MHz or more (and 10x is still wimpy), or if my circuit only needed to be AC coupled I'd seriously consider using an emitter- or source-follower to buffer my oscillator. -- Tim Wescott Wescott Design Services http://www.wescottdesign.com Do you need to implement control loops in software? "Applied Control Theory for Embedded Systems" gives you just what it says. See details at http://www.wescottdesign.com/actfes/actfes.html
From: Joel Koltner on 10 Apr 2008 12:57 "Tim Wescott" <tim(a)seemywebsite.com> wrote in message news:6rydnbZMeMTA2WPanZ2dnUVZ_g2dnZ2d(a)web-ster.com... > The design philosophy of an op-amp is that you use a crappy amplifier with > tons of excess gain, and fix all of the amplifier problems with tons of > feedback. And here I was about to give you $100k to build me a class A, zero-feedback, tube-based audio amplifier because it will OBVIOUSLY sound so much better than a traditional design with your approach. :-)
From: eeaj2002 on 10 Apr 2008 14:58 On Apr 10, 9:57 am, "Joel Koltner" <zapwireDASHgro...(a)yahoo.com> wrote: > "Tim Wescott" <t...(a)seemywebsite.com> wrote in message > > news:6rydnbZMeMTA2WPanZ2dnUVZ_g2dnZ2d(a)web-ster.com... > > > The design philosophy of an op-amp is that you use a crappy amplifier with > > tons of excess gain, and fix all of the amplifier problems with tons of > > feedback. > > And here I was about to give you $100k to build me a class A, zero-feedback, > tube-based audio amplifier because it will OBVIOUSLY sound so much better than > a traditional design with your approach. :-) Thank you all for your help. I need further help please. I wrote a simple program that generate a monochrome NTSC signal. I connected this signal to one of the eye glasses which take the NTSC signal but nothing was displayed. When I connected my signal to the DVD player and I fed the output of DVD player into the eye glasses then it displayed fine. I am assuming the reason for not displaying is the color burst that is missing (When I compared my signal and the output of the DVD player signal, the color burst was missing). This why I was trying to see if I can sum the 3.472MHz signal with my NTSC signal and see if I can display my signal on the eye glasses. By the way the eye glasses have VGA color display that takes the NTSC or PAL signal. Thank you all, John.
From: Joel Koltner on 10 Apr 2008 16:56
"eeaj2002" <eeaj2002(a)gmail.com> wrote in message news:2f48edc0-cacb-4c94-b704-8a5572a4d93b(a)q27g2000prf.googlegroups.com... > I wrote a simple program that generate a monochrome NTSC signal. I > connected this signal to one of the eye glasses which take the NTSC > signal but nothing was displayed. When I connected my signal to the > DVD player and I fed the output of DVD player into the eye glasses > then it displayed fine. I am assuming the reason for not displaying > is the color burst that is missing This is highly unlikely to be the problem; I don't think I've ever encountered a monitor that couldn't handle a regular old (RS-170) black and white signal. > This > why I was trying to see if I can sum the 3.472MHz signal with my NTSC > signal and see if I can display my signal on the eye glasses. This won't work... 1) The colorburst frequency is 3.579545MHz or somesuch -- but definitely not 3.472MHz. 2) It's a color *burst*, not a color *sub-carrier*. Take a look at the timing of a color video signal and you'll see it's only present during one of the "porch" after the sync pulse on each line. (Perhaps you know to do this and I just misunderstood what you wrote?) Have you tried feeding your signal into a regular old TV with a composite video input? Can you take a picture of a 'scope shot to show us what your timing looks like? Last time I was writing software to directly generate video, getting all the timing correct took plenty of time. To begin with you might want to just generate lines of video (containing horizontal sync) but skip vertical sync -- if you generate each line the same, the fact that the frame is "rolling" won't matter. Note that unfortunately the newer and hence "smarter" a TV is, the less useful it'll tend to be for debugging since it'll want to, e.g., just give you a blue screen when it doesn't "like" your video signal. ---Joel |