AD 846 AN is ther a better sounding opamp out there

I've heard a lot of people say the AD8610 is a nicer sounding opamp... but, I've never done a direct comparison myself. You should also check out the OPA627 which is highly regarded. I believe Mark Levinson uses those as I-V converters in their DACs. There's a thread on here somewhere in which someone lists the results of their opamp comparisons. Or, if you check on audioasylum, I think there's a lot of opamp opinions on there.

I've never run into a digital opamp, and I sure hope I never do!


Paid Member
2002-03-29 5:19 am
Harry said:


Maybe that's why it is made by Analog Devices.....


Harry, they make DSPs as well and thems sure ain't analog.

Then there are ADCs and DACs which are only "half analog".

Maybe they should change their name to "Some Analog Devices"

FYI since nobody has mentioned it in this thread, AD846 is
discontinued. AD suggests AD8001 or AD811 as replacement.

Garwin, note that AD846 is a current-feedback amp, so it may
not be possible to replace it with a voltage-feedback amp
such as AD825. There is a lot of talking about the AD825,
probably because of the LCs modules. I am a bit suspicious about
LCs claims, since the comparison table is flawed. The data for
AD825 is not taken directly from the datasheet, but arbitrarily
selected from the ranges given, while the worst-case figures
are often used for the contenders. how it sounds is a different
matter, of course, I am just pointing out that LCs attempt to
couple their listeining experiences to data is questionable.
There are many other interesting choices out there, and there
are many opams with similar data to AD825, which is actually
a low-cost device not intended for audio (could still be good for audio, of course). I haven't heard AD825 myself, but a friend of mine is currently trying it an seems to prefer it to OPA627 when
used in front of a BUF634 in his preamp.


2001-06-12 10:15 am
Re: AD825

I've used these a lot, in all manner of applications.

They do sound unusually good for an op-amp, and at least one of the reasons for this is the unusually high open-loop bandwidth, something almost unique in voltage feedback op-amps.

The only VF device I currently prefer over the AD825 is the AD8610.

I would not Recomend the AD-825 be used ahead of the BUF-634. Both the AD-825 & the BUF-634 have High DC offset and this offset is in the same Direction resulting in an overall High DC offset of about 20Mv used in Typical Closed Loop gains of 20 dB. On the Otherhand the OPA-627 & the AD-8610 Result in an Overall DC offset of Less than 1Mv. The Bandwidth of the AD-825 is also Large and Evean with the BUF-634 operated in it's wideband mode will require a large Phase lead Cap around the Opamp of about 30 pF for stability. Moreover Sonicaly the AD-825 & BUF 634 do not blend well together. I would sugest the AD-8610 as this Opamp IMHO sounds Better than the OPA-627. The AD-8610 has the Smoothness of the OPA-627 but without the Distent Midrange of the OPA-627. Both the OPA-627 & the AD-8610 Sonicaly blend well with the BUF-634.


2001-06-12 10:15 am

Do you mean the unusally low open loop gain? There are many higher bandwidth op amps.

Well I guess one comes with the other, all things being equal. I did mean what I said though, most op-amps copensation schemes roll off OL BW at a very low corner frequency, the AD825 corner is 8-10k.

This is unusual, unless I've missed some devices - it's never part of a parametric search, making it time consuming to investigate.

IMHO wide open loop bandwidth, flat over the range of frequencies being amplified, is important for audio and almost all op-amps fail in this regard.

The conventional OL BW of 99% of VF opamps tends to introduce frequency depedant, non-linear phase responses in the amplifier as a whole, due to input stage transconductance non-linearities, worse with BJT inputs, so the AD797 mentioned above is not good in this regard, depite the impressive looking data sheet.

As always though, there's much more to it than this.

Yes this is one nice quality of the ad-825 that 10KHz. rolloff. Some other Opamps used for Video applications whitch is also the case for the AD-825 also have high Rolloff points and Low open loop gains, However most use BJT as input transistors and then input Bias curent becomes an issue with reasonable source resistences. I would like to sugest the use of an opamp Buffer Combo as an alternative to get constant amount of overall feedback over as wide of a range as you want See walt jungs Multiloop feedback Idea

>the AD797 mentioned above is not good in this regard, depite the impressive looking data sheet.<

My view is that the quirks of the AD797 have more to do with how the current-mirror is handled and comparatively less to do with the input stage treatment.

Topology-wise, the interesting aspect of the AD797 is the bootstrapping of the current-mirror. What this kind of topology typically accomplishes is a substantial but conditional increase in open-loop gain. Compared to what you would get without the bootstrapping, the OL gain at low frequencies keeps on increasing and increasing, but the roll-off slope in the upper frequencies largely does not change. Ergo, when you set a given closed-loop gain, you get a huge increase in the amount of loop feedback at the frequencies where the OL gain is available and a corresponding decreasing in distortion - but the benefits are definitely frequency-dependent.

So in the end, I would largely agree with your final assessment.

regards, jonathan carr