new ultralow noise, high bw opamp from ADI
this looks pretty promising, although you are going to be soldering it with a magnifying glass, the AD8099 from Analog:
http://www.analog.com/UploadedFiles/Data_Sheets/40280151071581AD8099_0.pdf
this looks pretty promising, although you are going to be soldering it with a magnifying glass, the AD8099 from Analog:
http://www.analog.com/UploadedFiles/Data_Sheets/40280151071581AD8099_0.pdf
AD8099 not in audio band???
What do you mean by: "AD8099 not in audio band" ?
At the spec-page:
http://www.analog.com/Analog_Root/productPage/productHome/0,2121,AD8099,00.html
I read the following:
"Applications:
* Pre-amplifiers
* Receivers
* Instrumentation
* Filters
* IF and Baseband Amplifiers
* A-to-D Drivers
* DAC Buffers
* Optical electronics"
For me AD8099 seems to beat all other opamps by several length - Extremely low noise and extremely high slew-rate.
Has anyone tried it?
And what about the AD8067?
http://www.analog.com/Analog_Root/productPage/productHome/0,2121,AD8067,00.html
- Ask
What do you mean by: "AD8099 not in audio band" ?
At the spec-page:
http://www.analog.com/Analog_Root/productPage/productHome/0,2121,AD8099,00.html
I read the following:
"Applications:
* Pre-amplifiers
* Receivers
* Instrumentation
* Filters
* IF and Baseband Amplifiers
* A-to-D Drivers
* DAC Buffers
* Optical electronics"
For me AD8099 seems to beat all other opamps by several length - Extremely low noise and extremely high slew-rate.
Has anyone tried it?
And what about the AD8067?
http://www.analog.com/Analog_Root/productPage/productHome/0,2121,AD8067,00.html
- Ask
Nice chip. The advantage of these high speed amps is that the performances below 100kHz or so are very, very good for audio. Low noise, low THD up to well above the audio band, although I noticed a difference in THD performance between the two versions (SOIC and CSD).
They also can drive large loads, they can directly drive a headphone.
Disadvantage is the low supply voltage of 15V absolute max, which in practise means +/-6V, giving max output of 3.5VRMS or so. And, because of the high speed, they may be hell to stabilise - the PCB design will be quite critical I think.
Jan Didden
They also can drive large loads, they can directly drive a headphone.
Disadvantage is the low supply voltage of 15V absolute max, which in practise means +/-6V, giving max output of 3.5VRMS or so. And, because of the high speed, they may be hell to stabilise - the PCB design will be quite critical I think.
Jan Didden
Re: AD8099 not in audio band???
Preamplifier or DAC buffer does not automaticly means audio preamp or audio DAC buffer..
The noise level from the AD8099 is very high in the audio band.
askbojesen said:
Preamplifier or DAC buffer does not automaticly means audio preamp or audio DAC buffer..
The noise level from the AD8099 is very high in the audio band.
Re: AD8099 not in audio band???
So the opamp won't be particulary noisy for a 1 V application but is it good for a mic or MC preamp? .... but in other aspects the performance is impressing.
I mean that you must read and understand all important parameters and make an opinion from that. 0.95 nV/Hz says nothing if the application is high level and the signal source is "normal". You must see how the noise looks like in different bands.askbojesen said:
So the opamp won't be particulary noisy for a 1 V application but is it good for a mic or MC preamp? .... but in other aspects the performance is impressing.
I was going to try out the AD8067.
The nV root Hz curve seems to be pretty similar to the AD843 I was originally using, if not slightly better. The 8099 curve seems to be ten times better than the 843s.
The AD8067 has a lower slew rate than the 8099, but it's still two or three times quicker than the AD843 and the bandwidth is also lower than the 8099, but still better than the AD843, so stability should be less of a problem compared to the 99. It also has a wider power supply range than the 99.
Maybe I'm missing some detail or you're just used to quieter opamps. 😀
Any recommendations for replacing the AD843?
The nV root Hz curve seems to be pretty similar to the AD843 I was originally using, if not slightly better. The 8099 curve seems to be ten times better than the 843s.
The AD8067 has a lower slew rate than the 8099, but it's still two or three times quicker than the AD843 and the bandwidth is also lower than the 8099, but still better than the AD843, so stability should be less of a problem compared to the 99. It also has a wider power supply range than the 99.
Maybe I'm missing some detail or you're just used to quieter opamps. 😀
Any recommendations for replacing the AD843?
I just saw a thread mentioning the AD797 for low noise.
This wasn't included in my ADI searches because it's slew rate is more than ten times worse than the AD843 and I disliked the AD8610, which is three times quicker than the AD797. I also preferred the AD843 over the OPA637, which has only half the slew rate of the 843. So it seems slew rate is winning out over noise for me at the moment. Although I'm using this as a simple headphone amp, not a preamp in a RIAA stage.
This wasn't included in my ADI searches because it's slew rate is more than ten times worse than the AD843 and I disliked the AD8610, which is three times quicker than the AD797. I also preferred the AD843 over the OPA637, which has only half the slew rate of the 843. So it seems slew rate is winning out over noise for me at the moment. Although I'm using this as a simple headphone amp, not a preamp in a RIAA stage.
eeka chu said:
Hi,
Have you looked at the slew rate that you are likely to see at the signal input, worst case, and compared that to the opamp slew rate (AD843, 797, etc)? You probably will see several orders of magnitude overkill. Why are you so worried about this particular issue?
Jan Didden
janneman said:
You ain't kiddin' -- I am trying to implement a 100kHz low pass filter to replace a switched-cap filter -- there is some "clock-through" with the latter and I thought it would not be that difficult with one of these new high speed chips -- it's nice to have high GBW Product until you have to stabilize it -- the filter easily becomes an oscillator.
jackinnj said:
Yes, one important one is to avoid ANY capacitance at the inputs. 2pF here to a ground can wreak havoc. Not easy, I tried some of these chips in a super-duper regulator but abandoned it.
Jan Didden
It looks like a spectecular opamp, but indeed both current- and voltage-noise start to rise below 10KHz..... the PSRR is pretty good throughout the whole audio band...maybe input impedance is a bit low, and might have to be taken in account when designing filters...?
see selection of graphs..
see selection of graphs..
Attachments
janneman said:
Yes, I am going to have to go FDNR or state-variable and see how it works -- both Sallen Key and MFB have been very unstable.
the 4-pole 100 kHz Sallen Key low pass filter using the LM4562 attained about 83dB of attenuation -- and then rang like a bell -- it almost looked like a band-reject filter. I know that we are getting into "terra infirma" when you go over a few tens of kHz with opamps.
I have a reel of NE5532D's -- maybe I should try these!
A nice latecomer from AD.. 😀
See LMH 6624 [which in itself is a parent of CLC425]
Or BB opa846.
Ciao, George
See LMH 6624 [which in itself is a parent of CLC425]
Or BB opa846.
Ciao, George
janneman said:
Only just now!
I was listening to three opamps, the AD8620, the OPA637 and the AD843. I found my preference for them went down with the slew rate, before checking what the slew rates for each where. So I just assumed that slew rate is probably representative of something I like about the sound.
The AD843 sounded more open, clear and upfront.
After a quick google, I found some single figure V/us values for slew rate being suggested for a recording, but the person writing also started by mentioning that he wasn't looking at maximum slew rate. And having read audiophile website for more than two minutes, I'm now of the mindset where I dismiss pretty much everything unless I see it stated in at least three separate places by different people. 😀
Maybe I'm wrong, but I couldn't see anything so clearly different between the three amps I was listening to.
Doing the maths, if we assume 20kHz as the maximum signal frequency you can hear...
The time period period cycle is then 1/20kHz. I'll assume the wave is a perfect sinewave shape, so the rising edge will be 1/4 of that. If my amp is driving 3V into some headphones, that's 3/((1/20kHz)/4) for volts per second. / 1,000,000 to get it into microseconds and I get about 0.24 V/us
Perhaps the error in that logic is assuming a regular sinewave shape. Maybe there are some rising edges on the waves with steeper angles. I'm not so sure about the limits of human hearing when applied to slew rate.
Next time I'm in the electronics lab I might connect might see what the maximum is I can find.
eeka chu said:
And even then, it is not uncommon to see someone posit a nonsense statement only to have it repeated many times at varying places. This is for me the single most significant problem of getting info from the web. I don't think that there is any substitute for researching issues yourself.
One thing I look for is whether the person stating something also gives the rationale or arguments for his opinion. If he/she only says: "It is well known that..." I get suspicious...
Jan Didden
janneman said:
Quite, trying to learn electronics through audiophile sites is an indescribably tedious process at times. The only problem is that there's also some extremely talented people mixed in with it all that you won't find elsewhere.
Anyone feel free to shout out if you can think of another reason for why my preference on the three opamps went the way it did. Or if you can think of a way of determining the maximum slew rate of human ears (could assume 3V into my headphones is the 'loudest' my ears can hear, without bursting anyway, so the maximum slew rate will be the shortest rise time to that voltage, but does that follow that I can just use a sinewave example at 20kHz and get the maximum? I suppose more importantly, it doesn't even matter if I don't have anything higher on the recordings. If I had a fancy scope I could just set it to autohunt the steepest and tallest edges, but I don't 😀 maybe there's something suitable around the department somewhere).
just had a quick think about what kinds of instruments in a recording would have a huge slew rate, since it implies that the waveform won't go smoothly from base to peak if the frequency is kept constant at 20kHz and amplitude is some constant arbitary level.
Electric guitars came to mind. If they're heavily clipped, the slew rate could be extremely high. Now the question becomes, how high and can I hear it?
I've looked over the datahsheets for the OPA637, AD8620 and AD843 again and the slew rate is still the only thing glaring out at me.
There's also possibility for me being able to hear that difference because I listen to lots of rock and have been (possibly) overly interested in the sound of guitar distortion for half a decade or so. If you were listening to classical music, slew rates might be less of an issue given than classical instruments aren't designed to have rocketing slew rates into clipping; they'll be closer to the sinewave example, which means a lower slew rate for a given SPL compared to a nearly squarewave clipped guitar at the same SPL. Which seems to at least roughly agree with tradition given that lots of people who like valve amps like their smoother sound, and valves have low slew rates compared to opamps like the AD843.
Electric guitars came to mind. If they're heavily clipped, the slew rate could be extremely high. Now the question becomes, how high and can I hear it?
I've looked over the datahsheets for the OPA637, AD8620 and AD843 again and the slew rate is still the only thing glaring out at me.
There's also possibility for me being able to hear that difference because I listen to lots of rock and have been (possibly) overly interested in the sound of guitar distortion for half a decade or so. If you were listening to classical music, slew rates might be less of an issue given than classical instruments aren't designed to have rocketing slew rates into clipping; they'll be closer to the sinewave example, which means a lower slew rate for a given SPL compared to a nearly squarewave clipped guitar at the same SPL. Which seems to at least roughly agree with tradition given that lots of people who like valve amps like their smoother sound, and valves have low slew rates compared to opamps like the AD843.
jackinnj said:
I'd disagree with Itzhak, and I'm pretty sure I have more science on my side.
It's also born out by the trend of people who prefer solid state for modern music as opposed to valves and vice versa.
A guitar string and a violin string are doing something very similar. If the output of the guitar is also being fed into some form of reasonably high gain, high slew rate amplifier (e.g. a solid state distortion pedal), it's going to annihilate the violin in terms of slew rate.
If you were to pluck a string, hit a cymbal or drum, that should also produce something with a high initial slew. But I'd still go with modern music for consistently high rates.
- Status
- Not open for further replies.