Instrumentation op-amp for DC-coupling to audio input?
in [Design]
|
From: Lostgallifreyan on 14 Jul 2010 13:08 tony sayer <tony(a)bancom.co.uk> wrote in news:tFmovibeqdPMFwmL(a)bancom.co.uk: > In article <Xns9DB5A2B3F13DDzoodlewurdle(a)216.196.109.145>, > Lostgallifreyan <no-one(a)nowhere.net> scribeth thus >>George Herold <gherold(a)teachspin.com> wrote in >>news:11271e1e-5ded-4bee-9852- 9152ec537b42(a)j13g2000yqj.googlegroups.com: >> >>>> further. :) I doubt that any input I'll put into this system will >>>> have mu >>> ch >>>> signal above audio band. >>> >>> Oh, your opamps and resistors will have noise that goes out to MHz and >>> beyond. And will only be fall off when you reach the gain BW limit of >>> the opamp. >>> >> >>Yes, but I won't be adding to it in any obvious or fixable way, and the >>systems I'm adapting a DC coupler to will already have filtering for >>their signal converters unless they weren't designed well. >> >>> The TI website is now working for me. The 'new' audio opamp is the >>> OPA1641, 1642 and 1644. The 'audio' specs on these look pretty good. >>> (You should at least give them a glance.) The only down side is the >>> DC performance. The offset voltage is 0.5mV or 1 mV. >>> >>> George H. >>> >> >>Thanks. I'll look at them, and if I can afford to, buy some at some >>point to get used to them. I'm staying with OPA2277 for now though, as >>DC is what I want, I just wanted to be sure it didn't let me down for >>AC, and my slew rate and sample rate calculations suggest it won't. I'm >>not building high gain mic preamps with them.. > > Is it possible to say more precisely what you are doing with this > application at all JOOI?... No. Without disclosing a diagram I have no right to pass on, there isn't. In the posts here I described it in a lot of detail too, so I won't repeat myself.
From: tony sayer on 15 Jul 2010 05:28 In article <Xns9DB5B87CA4FC1zoodlewurdle(a)216.196.109.145>, Lostgallifreyan <no-one(a)nowhere.net> scribeth thus >tony sayer <tony(a)bancom.co.uk> wrote in >news:tFmovibeqdPMFwmL(a)bancom.co.uk: > >> In article <Xns9DB5A2B3F13DDzoodlewurdle(a)216.196.109.145>, >> Lostgallifreyan <no-one(a)nowhere.net> scribeth thus >>>George Herold <gherold(a)teachspin.com> wrote in >>>news:11271e1e-5ded-4bee-9852- 9152ec537b42(a)j13g2000yqj.googlegroups.com: >>> >>>>> further. :) I doubt that any input I'll put into this system will >>>>> have mu >>>> ch >>>>> signal above audio band. >>>> >>>> Oh, your opamps and resistors will have noise that goes out to MHz and >>>> beyond. And will only be fall off when you reach the gain BW limit of >>>> the opamp. >>>> >>> >>>Yes, but I won't be adding to it in any obvious or fixable way, and the >>>systems I'm adapting a DC coupler to will already have filtering for >>>their signal converters unless they weren't designed well. >>> >>>> The TI website is now working for me. The 'new' audio opamp is the >>>> OPA1641, 1642 and 1644. The 'audio' specs on these look pretty good. >>>> (You should at least give them a glance.) The only down side is the >>>> DC performance. The offset voltage is 0.5mV or 1 mV. >>>> >>>> George H. >>>> >>> >>>Thanks. I'll look at them, and if I can afford to, buy some at some >>>point to get used to them. I'm staying with OPA2277 for now though, as >>>DC is what I want, I just wanted to be sure it didn't let me down for >>>AC, and my slew rate and sample rate calculations suggest it won't. I'm >>>not building high gain mic preamps with them.. >> >> Is it possible to say more precisely what you are doing with this >> application at all JOOI?... > >No. Without disclosing a diagram I have no right to pass on, there isn't. In >the posts here I described it in a lot of detail too, so I won't repeat >myself. OK Fair comment!... -- Tony Sayer
From: Lostgallifreyan on 15 Jul 2010 07:11 tony sayer <tony(a)bancom.co.uk> wrote in news:8U387CdYTtPMFwFR(a)bancom.co.uk: >>> Is it possible to say more precisely what you are doing with this >>> application at all JOOI?... >> >>No. Without disclosing a diagram I have no right to pass on, there >>isn't. In the posts here I described it in a lot of detail too, so I >>won't repeat myself. > > OK Fair comment!... Cool. Thought you might be going to argue. :) I can tell you this much... I'm making a device that lets me turn a sound card into a logging tool. Assuming the ADC's have reasonable DC performance, I can use it to map out changes that are unique, too fast for a multimeter, too slow and unrepeatable for an oscilloscope. By using something like Sound Forge, which has really nice keyboard shortcuts for zooming, locating, selecting, etc, it makes fun easy work of sifting through truly enormous amounts of data, so watching for all sorts of drifting in meter circuits can be done. Further, you can do experiments logging to one channel while speaking a commentary into the other, so on playback you can know what you did that caused the responses you see. This could be a very useful diagnostic tool, and very cheap. All the earlier specifics are basically about trying to do this while not screwing up the sound card performance, so the adapter can be left in circuit. Out of general interest, Sparkfun Electronics do a nice multichannel logger called the Logomatic V2, which (with Kwan's firmware) can log two serial inputs and 8 (10?) analog inputs at up to 1KHz at 12 bits, but if, for a similar price, you can get two or more channels sampling at 48 KHz at 16 bits, it has to be worth trying... A lot of people have done this sort of thing for laser show control by modifying the outputs, but I haven't seen much to suggest anyone's doing it to inputs. Should be enough here now, with the other posts, to allow anyone to figure out what to do. Or come up with a better way, in which case, please post it.
From: Phil Hobbs on 15 Jul 2010 15:11 Lostgallifreyan wrote: > tony sayer <tony(a)bancom.co.uk> wrote in > news:8U387CdYTtPMFwFR(a)bancom.co.uk: > >>>> Is it possible to say more precisely what you are doing with this >>>> application at all JOOI?... >>> No. Without disclosing a diagram I have no right to pass on, there >>> isn't. In the posts here I described it in a lot of detail too, so I >>> won't repeat myself. >> OK Fair comment!... > > Cool. Thought you might be going to argue. :) I can tell you this much... I'm > making a device that lets me turn a sound card into a logging tool. Assuming > the ADC's have reasonable DC performance, I can use it to map out changes > that are unique, too fast for a multimeter, too slow and unrepeatable for an > oscilloscope. By using something like Sound Forge, which has really nice > keyboard shortcuts for zooming, locating, selecting, etc, it makes fun easy > work of sifting through truly enormous amounts of data, so watching for all > sorts of drifting in meter circuits can be done. Further, you can do > experiments logging to one channel while speaking a commentary into the > other, so on playback you can know what you did that caused the responses you > see. This could be a very useful diagnostic tool, and very cheap. > > All the earlier specifics are basically about trying to do this while not > screwing up the sound card performance, so the adapter can be left in > circuit. > > Out of general interest, Sparkfun Electronics do a nice multichannel logger > called the Logomatic V2, which (with Kwan's firmware) can log two serial > inputs and 8 (10?) analog inputs at up to 1KHz at 12 bits, but if, for a > similar price, you can get two or more channels sampling at 48 KHz at 16 > bits, it has to be worth trying... A lot of people have done this sort of > thing for laser show control by modifying the outputs, but I haven't seen > much to suggest anyone's doing it to inputs. > > Should be enough here now, with the other posts, to allow anyone to figure > out what to do. Or come up with a better way, in which case, please post it. What I did was to buy a second-hand HP 35665A for $350. A boat anchor, but a goodie. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal ElectroOptical Innovations 55 Orchard Rd Briarcliff Manor NY 10510 845-480-2058 hobbs at electrooptical dot net http://electrooptical.net
From: Lostgallifreyan on 15 Jul 2010 16:14
Phil Hobbs <pcdhSpamMeSenseless(a)electrooptical.net> wrote in news:4C3F5D7D.3020304(a)electrooptical.net: > What I did was to buy a second-hand HP 35665A for $350. A boat anchor, > but a goodie. > A bonny wee beastie. :) I like HP Agilent stuff, I have a nice 1740A scope that people compare favourable with a Tektronix 265B scope but having seen both, I think the HP wins tenfold. But here, I bet I could pay the equivalent of $350 just getting that HP 35665A shipped to me. Joking.. but it does look heavy. I'll settle for the adapted soundcard because the bang per buck is so good, and I can do it with no added weight of gear, or extra space found for it. And I can use all kinds of software to handle the data easily. |