I just noticed today on the TI web site that the OPA1611 (single) and OPA1612 are finally reality. I've seen mention of these on various threads in this and other forums. Maybe it's too soon for anyone to have obtained and evaluated samples, but -- well -- has anyone yet? They are supposed to be great audio op-amps. The datasheet looks pretty impressive -- good 'objective' specs. I would hope they sound as good as they measure. That would be most excellent.
Last edited:
Suspiciously similar to OPA211/2211
If you compare the OPA1611 to OPA211 you notice that they are essentially the same chip, only that OPA1611 has lousier specs for DC offset and drift. A clever way to make money of the rejects. The same "sorting" is done for OPA132 / OPA134, pitching the chips with more offset and lousier CMRR (OPA134) as suitable for audio.
From the hype surrounding the OPA1611 I had hoped for an output stage with a higher standing current (bias), but comparing the OPA211 and OPA1611 datasheets it's definitely the same chip - 3,6 mA idle current for both.
Oh well, marketing at it's best again ...
Edit: BTW i think OPA211 is great sounding op-amp!
If you compare the OPA1611 to OPA211 you notice that they are essentially the same chip, only that OPA1611 has lousier specs for DC offset and drift. A clever way to make money of the rejects. The same "sorting" is done for OPA132 / OPA134, pitching the chips with more offset and lousier CMRR (OPA134) as suitable for audio.
From the hype surrounding the OPA1611 I had hoped for an output stage with a higher standing current (bias), but comparing the OPA211 and OPA1611 datasheets it's definitely the same chip - 3,6 mA idle current for both.
Oh well, marketing at it's best again ...
Edit: BTW i think OPA211 is great sounding op-amp!
Last edited:
The OS looks borrowed from the xDSL drivers, low distortion on min possible standing current.
Pelle,
Are you sure your comment is correct? From what I've heard/read, the "audio" parts (1611 / x134) are not trimmed for lowest DC offset (relative to their siblings) as that's not an important audio parameter. That doesn't make them "rejects". Is a Toyota Highlander a "reject" Lexus RX since they share many common components and general design?
I believe I read somewhere (AES report on this forum?) that the 1611' AC response has been tweaked for lower distortion compared to the '211. I've not listened to anything with the 211, though I hear good reports; I guess I'll have to go straight to the 1611...
If only they were in PDIP. Sigh...
Are you sure your comment is correct? From what I've heard/read, the "audio" parts (1611 / x134) are not trimmed for lowest DC offset (relative to their siblings) as that's not an important audio parameter. That doesn't make them "rejects". Is a Toyota Highlander a "reject" Lexus RX since they share many common components and general design?
I believe I read somewhere (AES report on this forum?) that the 1611' AC response has been tweaked for lower distortion compared to the '211. I've not listened to anything with the 211, though I hear good reports; I guess I'll have to go straight to the 1611...
If only they were in PDIP. Sigh...
It is very likely he's correct. OPA211 is a SiGe part, and as in every new technology it is likely the parameters dispersion is still pretty large, so there's lot of room for "rejects". Usually, manufacturers choose to split the production in classes. Apparently TI decided to take advantage of the BB reputation in the hifi world and re-baptize lower class parts with the "audio" tag.
I don't think there's anything wrong in doing this, as you said, DC performance is not really critical in audio and OPAA1611/12 are actually 25% cheaper. OPA211 is an excellent opamp for audio, at par to the AD ADA4898 and the National LME4562.
BrianL,
I guess "reject" is a too strong word indicating a completely defective component (my bad - not native in english), maybe I should have written not-quite-up-to-spec or something like that.
I also wish they made them in DIP packages. NS definitely got good packagings for their new amps; both DIP and also TO-99 (at a premium).
I guess "reject" is a too strong word indicating a completely defective component (my bad - not native in english), maybe I should have written not-quite-up-to-spec or something like that.
I also wish they made them in DIP packages. NS definitely got good packagings for their new amps; both DIP and also TO-99 (at a premium).
I talked to a TI/B-B applications engineer today. My statements were verified:
- OPA1611/1612 are optimized for better distortion than the OPA211 (already pretty good). So, it is a different design and chip.
- high levels of DC precision are generally not required by audio customers (audio equipment makers). So, parts are not spec'ed for or tested for the same levels of DC precision as a part, i.e., OPA211, designed, spec'ed and sold for DC precision. (my note: 'customers' generally exclude us hobby types as we're not the guys who buy the quantities that keep vendors in business! So even if 'we' think extreme DC precision is important, we're in the minority as far as the market is concerned. And if the big customers don't want/need something that adds expense, then a company is not going to provide it in the product.)
So, as a user, Pelle or Syn08, I guess you (and I) get to decide which is best for our application. I hear the '1611 is the best sounding ever, so I'm willing to give it a try. With all this talk of DC specs, it's interesting to note that the OPA1611 equals or exceeds the LM4562 in everything except bias current.
The bottom line is that different markets need different specs and companies optimize products to match the market. Even if the base designs are similar or one is based on the other (as in the Highlander vs. Lexus RX comparison), doesn't make one necessarily 'superior' and the other 'inferior'
- OPA1611/1612 are optimized for better distortion than the OPA211 (already pretty good). So, it is a different design and chip.
- high levels of DC precision are generally not required by audio customers (audio equipment makers). So, parts are not spec'ed for or tested for the same levels of DC precision as a part, i.e., OPA211, designed, spec'ed and sold for DC precision. (my note: 'customers' generally exclude us hobby types as we're not the guys who buy the quantities that keep vendors in business! So even if 'we' think extreme DC precision is important, we're in the minority as far as the market is concerned. And if the big customers don't want/need something that adds expense, then a company is not going to provide it in the product.)
So, as a user, Pelle or Syn08, I guess you (and I) get to decide which is best for our application. I hear the '1611 is the best sounding ever, so I'm willing to give it a try. With all this talk of DC specs, it's interesting to note that the OPA1611 equals or exceeds the LM4562 in everything except bias current.
The bottom line is that different markets need different specs and companies optimize products to match the market. Even if the base designs are similar or one is based on the other (as in the Highlander vs. Lexus RX comparison), doesn't make one necessarily 'superior' and the other 'inferior'
Last edited:
Unlikely. Even the distortion spec is identical for OPA211 and OPA1611: -136dB @ 1KHz and 3V out.
But then OPA1611 is cheaper and, as you said, the DC performance is not critical in audio. Otherwise, I'm 99.9999% sure there is no difference between the objective performances of these two chips. Can't wait though to read about how much better OPA1611 sounds
I'm still waiting for samples they promised last October...
A first, they promised DIP for prototyping.
Then they said that there are no DIP doubles anymore, singles only...
Then they answered that there will be no DIPs at all, but they will start shipping SMD version soon, and I am in their list...
Now, almost an year later, I hear they have samples available, but no message from them...
May be it is a Fortune telling that I should not use them at all?
A first, they promised DIP for prototyping.
Then they said that there are no DIP doubles anymore, singles only...
Then they answered that there will be no DIPs at all, but they will start shipping SMD version soon, and I am in their list...
Now, almost an year later, I hear they have samples available, but no message from them...
May be it is a Fortune telling that I should not use them at all?
syn08: yes likely; in fact, true. Compare gain/phase curves (fig 11 OPA211; fig 5 OPA1611); they're not the same. Carefully compare comparable distortion plots at high frequencies. Yes, probably no noticable change at 1kHz, but definitely in the high frequencies. All of these show the effects of changing open loop response.
Reminds me of one of my favorite phrases from an equipment review in The Absolute Sound years ago:
"subtle, yet dramatic differences".
;-)
I, too, await samples. From the datasheet, it looks as if any hopes of PDIPs have vanished, again I'd guess that the large customers don't do PDIPs anymore. I love PDIPs -- you can see them; you can read the marking on the package...
Reminds me of one of my favorite phrases from an equipment review in The Absolute Sound years ago:
"subtle, yet dramatic differences".
;-)
I, too, await samples. From the datasheet, it looks as if any hopes of PDIPs have vanished, again I'd guess that the large customers don't do PDIPs anymore. I love PDIPs -- you can see them; you can read the marking on the package...
Last edited:
For all practical purposes, they are the same. Same unity gain frequency (30MHz) and the same phase rolloff. Don't tell me about the higher OL bandwidth, this opamp is single pole compensated (see the schematic) and hence has a GBW of 30MHz. OPA1611's are likely low class OL bandwidth OPA211's, hence the slightly higher OL bandwidth.
About distortion curves, I don't see any real optimisation. At best, a more careful triage in testing. All values deltas are well under anything that a process or design can control.
You know, it's so very hard to combat cynicism with facts...
What you're essentially telling me is the apps guy who talked to me is lying. Well, I've consulted with him on many topics over a number of years and have never found this to be the case. His advice has always been spot-on (as my Scottish co-worker would say).
There are distinct differences in the open loop gain/phase out at high frequencies. That's obvious from the plots. There's also a difference in the HF portion of the distortion plots. If you want to attribute this to lot-to-lot difference, then time will tell what the "mean' is. If you want to attribute the differences to cherry-picking data from a bunch of different devices, then I suppose you're free to that opinion.
I'm done here. The parts are available; the data out there for people to read. I intent to try these myself in a current project or two and hope to find myself another small step closer to audio nirvana. (once I get some SOIC-to-PDIP adapters...)
What you're essentially telling me is the apps guy who talked to me is lying. Well, I've consulted with him on many topics over a number of years and have never found this to be the case. His advice has always been spot-on (as my Scottish co-worker would say).
There are distinct differences in the open loop gain/phase out at high frequencies. That's obvious from the plots. There's also a difference in the HF portion of the distortion plots. If you want to attribute this to lot-to-lot difference, then time will tell what the "mean' is. If you want to attribute the differences to cherry-picking data from a bunch of different devices, then I suppose you're free to that opinion.
I'm done here. The parts are available; the data out there for people to read. I intent to try these myself in a current project or two and hope to find myself another small step closer to audio nirvana. (once I get some SOIC-to-PDIP adapters...)
He's not lying to you. He only tells you what he knows, and that is essentially what he was told about this part, and that is essentially what the company wants you to hear. Very few application engineers have much of an idea about silicon processing, testing, sorting, packaging, burn in, etc... Their knowledge stops at the gates of the SOT-8 package, which doesn't make him less knowledgeable when it comes to advising customers.
I have worked for quite some years in a semiconductor plant and still have a pretty good insight on how many parts are actually "manufactured". The number of original designs is usually much smaller than the number of pages in the catalog. The rest is sorting, re-packaging, etc... TI seem to excel in such practices (which are nowhere to be ashamed of)... on top of my head, another glaring example are OPA1632 and THS4131.
Good luck, OPA1611 is definitely a very good opamp.
Correction: In my post above, read "OPA1611's are likely low class OL gain OPA211's, hence the slightly higher OL bandwidth"
Last edited:
My intention in starting this thread was to find out if any experimentation and discussion was already ongoing with a promising new audio part that is of interest to me. The hope was to get reactions from other users to assist/guide my own work. It was not to get into an argument with syn08 (Ovidiu ?) on why he knows more/less than I, the apps. engineer or whomever.
I hope if anyone has given the parts a try and would like to report on their results, that they would feel welcome to speak up or PM me. I notice that folks on HeadFi are starting comparisons.
I hope if anyone has given the parts a try and would like to report on their results, that they would feel welcome to speak up or PM me. I notice that folks on HeadFi are starting comparisons.
Last edited:
At the cost of masks and wafers, it would be highly unlikely that the economics of two separate products would be justified for such trivial differences. OTOH I find it difficult to believe an op-amp would get released in the first place that had any yield loss to OLG unless it was known in advance.
Fair enough Brian, but OTOH if someone knowledgable makes it clear that the 'new' opamp is really an existing one in a new coat, that could save you a lot of useless work and you could go after something really new. I would call that very usefull advice.
jd
I think it was known in advance. OPA211, OPA2211 and OPA827 were delayed for quite some time, reason being the maturity of the new BiCom-3 SiGe process.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.