I've seen a lot of snide remarks or perhaps they are tongue in cheek comments about the AD797 opamp, what I am curious is, is this strictly a reaction to the cost of that device that limits its use?
No, it's mostly from animosity toward the designer, coupled with the recognition that it outperforms a lot of hyped-but-obsolete technologies. It's an excellent chip as long as the source impedance is reasonably low (see Samuel Groner's opamp survey).
Is there an accepted valid test for RFI susceptibility of opamps? I have seen comments that the JFET input opamps are better and just now that bipolars are better. There was a brief discussion on these forums about testing but no results. I read somewhere about a standard test but that is lost in the murk of time.
There should be no argument that bad execution or layout will cause issues with either type of device so I would be cautious about generalizing experiences.
There should be no argument that bad execution or layout will cause issues with either type of device so I would be cautious about generalizing experiences.
We used to wave a power drill over the test jig, preferably one with a dirty rotor.
Mr Soibelman,
you are aware that the AD797 has been among the top opamps for about a quarter century by now ? 1st ones cost private individuals an arm and a leg.
(cost wise, it would be identical to criticising the OPA627/637, or the LT1028 before the BB parts came out in '91)
Jacco,
That is what I thought but the comments in this thread are sometimes so obtuse that it is hard to know when it is just a personal animosity or off handed comment. Am I correct that Scott had a major hand in this chip? I don't know how anyone can be more helpful than Scott, but he seems to have to put up with some uncalled for comments on this thread.
I have noticed that some here drove off Jimmy Neutron, luckily I have other methods to contact him without needing to do it through this forum. He is missed by me here bring some sanity to these conversations. I do know by now which people add to the conversation or clean up the misinformation. Opinions are one thing, facts are another.
That is what I thought but the comments in this thread are sometimes so obtuse that it is hard to know when it is just a personal animosity or off handed comment. Am I correct that Scott had a major hand in this chip? I don't know how anyone can be more helpful than Scott, but he seems to have to put up with some uncalled for comments on this thread.
I have noticed that some here drove off Jimmy Neutron, luckily I have other methods to contact him without needing to do it through this forum. He is missed by me here bring some sanity to these conversations. I do know by now which people add to the conversation or clean up the misinformation. Opinions are one thing, facts are another.
" . . . which also happens to tolerate RF ingress just as well."
I am making the general case for degenerated bipolars here. Most IC opamps do not degenerate the IPS in order to maximize noise performance.
However, if you have RF ingress . . . use RFI mitigation techniques. The engineering is well documented and it works very well. I did a lot of EMI work on some auto electronics earlier in my career at 50 and 100V/m field strengths. The products had opamps in them and a lot of other circuitry. RF SMD ferrites, R's and C's plus careful layout and minimization of loop areas got me through the tests and the necessary certification.
Funny, I built an LM4562 based preamp (please see my X-altra Mini One on my website) when in Japan and then onto Taiwan and China, using it right up until recently. I never once had any RF ingress issues. Mind you, I did have a RC low pass filter on the front end just before the volume control. -3dB IIRC was 200 kHz - 20 dB/decade and the loop areas were well controlled.
My new 'Symphony' preamp used an AD797 front end fed either directly from the source (assume Rin then is <200 Ohms) or a jumper selectable 20 dB pad with Rin to the opamp of 900 Ohms for optimum noise performance. The subsequent stages use LME49710.
I do not think you can match modern day opamps with discrete designs in terms of overall performance and especially so when you factor in cost. But I'd say on cost no object designs, opamps still give the best result. Heresy? No, just solid engineering and a clear recognition that in the last 30 years of opamp development we've gone from the Model T to the Bugatti Veyron*
http://social.ford.co.uk/wp-content...l-the-cars-on-the-road-were-Model-T-Fords.jpg
http://futurecarslist.xyz/wp-conten...yron-super-sport-specs-release-photo-JWsm.jpg
*hope this helps you to visualize the progress more clearly John 😉
I am making the general case for degenerated bipolars here. Most IC opamps do not degenerate the IPS in order to maximize noise performance.
However, if you have RF ingress . . . use RFI mitigation techniques. The engineering is well documented and it works very well. I did a lot of EMI work on some auto electronics earlier in my career at 50 and 100V/m field strengths. The products had opamps in them and a lot of other circuitry. RF SMD ferrites, R's and C's plus careful layout and minimization of loop areas got me through the tests and the necessary certification.
Funny, I built an LM4562 based preamp (please see my X-altra Mini One on my website) when in Japan and then onto Taiwan and China, using it right up until recently. I never once had any RF ingress issues. Mind you, I did have a RC low pass filter on the front end just before the volume control. -3dB IIRC was 200 kHz - 20 dB/decade and the loop areas were well controlled.
My new 'Symphony' preamp used an AD797 front end fed either directly from the source (assume Rin then is <200 Ohms) or a jumper selectable 20 dB pad with Rin to the opamp of 900 Ohms for optimum noise performance. The subsequent stages use LME49710.
I do not think you can match modern day opamps with discrete designs in terms of overall performance and especially so when you factor in cost. But I'd say on cost no object designs, opamps still give the best result. Heresy? No, just solid engineering and a clear recognition that in the last 30 years of opamp development we've gone from the Model T to the Bugatti Veyron*
http://social.ford.co.uk/wp-content...l-the-cars-on-the-road-were-Model-T-Fords.jpg
http://futurecarslist.xyz/wp-conten...yron-super-sport-specs-release-photo-JWsm.jpg
*hope this helps you to visualize the progress more clearly John 😉
Last edited:
Scott designed that chip and (I think) patented the basic circuit concepts.
jneutron wasn't driven off, he has had other personal issues to deal with. Yeah, very smart guy, and even when we strongly disagreed, he always kept things on a logical and professional level.
jneutron wasn't driven off, he has had other personal issues to deal with. Yeah, very smart guy, and even when we strongly disagreed, he always kept things on a logical and professional level.
C'est le ton que fait la musique.
(Provided one does not check the poster for a sense of humor)
Jacco,
Yes there are many times when it is hard to tell when someone is serious or playing. Some here have personal relationships that the rest of us don't know about that colors the conversation, so it is hard to tell from the written word. Even with emoticons it can be questionable at time, but at least that helps to understand the meaning, sometimes. Short snippets of a conversation are just so often misconstrued as to what the real intention of the words mean. I try very hard not to use those types of sentences or even abbreviations that are not known to those who are not native speakers and can quickly get lost in the conversation.
Yes there are many times when it is hard to tell when someone is serious or playing. Some here have personal relationships that the rest of us don't know about that colors the conversation, so it is hard to tell from the written word. Even with emoticons it can be questionable at time, but at least that helps to understand the meaning, sometimes. Short snippets of a conversation are just so often misconstrued as to what the real intention of the words mean. I try very hard not to use those types of sentences or even abbreviations that are not known to those who are not native speakers and can quickly get lost in the conversation.
And how does the AD797 differ from a 741? They both have eight legs. 🙂
My experience is that when there is EMI that the noise floor appears to go up. This is from stadium experience listening by ear and confirming with an oscilloscope. The fix is to use metallic conduit, twisted pair cables, actual chokes, magic beads (haven't really had much luck with these or their varients) but the best result is to use is the AC power line filters that are two coils with slightly fancy magnetics. Very low power ones for microphone lines and bigger ones for loudspeaker level.
It is very common for hanging choir microphones to pick up radio signals. The bifilar chokes really do fix this.
Last edited:
Times moves on, new devices appear and naturally tend to replace older devices. However, I can see no reason for anyone to HAVE to change devices just because new ones are here, especially if one is to change devices with which he is very familiar and has a long standing relationship with. As just an example, I am in no particular hurry to give up OPA 37 op amps simply because they served me so well over the years. They have their shortcomings, to be sure, but they are dependable fo returning good quality sound.
In that sense, I understand John's "love affair" with his favorite op amp, but if memory serves, he did post a comment some time ago that he is looking at newer FET input op amps such as 2134 etc. My NAD CD player has also made me look at them muh more carefully than I did before as it clearly proves they can sound better than I previously thought.
In that sense, I understand John's "love affair" with his favorite op amp, but if memory serves, he did post a comment some time ago that he is looking at newer FET input op amps such as 2134 etc. My NAD CD player has also made me look at them muh more carefully than I did before as it clearly proves they can sound better than I previously thought.
Last edited:
PMA- Thanks for the link (http://www.ti.com/lit/an/sboa128a/sboa128a.pdf) . It seems to be a good test for conducted EMI and the presumption is that any radiated sensitivity would also show up as conducted.
The stuff to do the tests doesn't look too difficult to set up. Has anyone done these tests on the common audio opamps? I suspect that their absence in TI datasheets doesn't reflect well on the devices.
The stuff to do the tests doesn't look too difficult to set up. Has anyone done these tests on the common audio opamps? I suspect that their absence in TI datasheets doesn't reflect well on the devices.
Absolutely (IMO) nothing wrong with that. That said, I probably wouldn't use an OPA37 for new designs, even if I shared your same design constraints. Not that the OPA37 is a bad chip.
I realize that actives/passives cost is pretty minor in the grand scheme of DIY designs (chassis/PSU drive costs), but, and as absolutely no slight against Scott, the AD797 doesn't seem to strike the right $:benefit ratio for most places in an audio playback chain, where its superb linearity, GBW at high gains, and low noise is lost on the user*. IOW, to be used in the datasheet-recommended locations, where it's able to benefit the most. Fitting the tool to the application and all that jazz. 😀
* And this assumes the layout/supporting elements are in place to extract the benefit. As an analog neophyte, Analog's datasheets are wonderful to read/learn from.
** My cost:benefit is probably different than many here, acknowledged.
- Status
- Not open for further replies.