Discrete Flat/Transparent Doa - Opamp any diy about?
- Thread starter r2d2
- Start date
Help Support GroupDIY Audio Forum:
The Jensen (990?) was flat AFAIK... Why does it need to be a DOA? Modern ICs are quite good.
JR
JR
3nity
Well-known member
I found the ppa-992 to be extremely transparent.
and very affordable.
and very affordable.
zamproject
Well-known member
- Joined
- May 11, 2010
- Messages
- 1,510
Studer A101 ?
hakanai
Well-known member
my experience has been that a well implemented modern IC is the way to go if you want a "do no harm" kind of circuit. It's that "well implemented" part you have have to watch for! No matter how good the part spec is, we can make anything suck (new groupDIY motto?).
Ok anything else I have the urge to write gets into the not-so-productive topic of "transparent". So I'll stop myself here.
Don
Ok anything else I have the urge to write gets into the not-so-productive topic of "transparent". So I'll stop myself here.
Don
Henke
Well-known member
r2d2 said:
IME, it's hard to beat the THAT 1240-series line receivers / 1646 line drivers performance-wise
hakanai
Well-known member
Totally agree. If I wanted a few channels of basic, yet professional, mic pre's with not much in the way of color I would just go to the THAT data sheets and call it a day (1510->1646). Same for line receiver's. I am using the 1206 in the center section of a mixer I am laying out right now.Henke said:
Don
probably you are right ,
also mackie , soundcraft , allen & heat , many ssl units ...
are designed with ICs in the input and output stages ,
but are not considered " top/best " analog sound ,
and from that i read around ... , seem that many guys hate ICs and want only transformers , Doa , and tubes too .. ??? :-\
about the ICs you are put in the "prototype"
have a good bandwidth in addition to being very transparent ?
R.
also mackie , soundcraft , allen & heat , many ssl units ...
are designed with ICs in the input and output stages ,
but are not considered " top/best " analog sound ,
and from that i read around ... , seem that many guys hate ICs and want only transformers , Doa , and tubes too .. ??? :-\
about the ICs you are put in the "prototype"
have a good bandwidth in addition to being very transparent ?
R.
hakanai said:
Andy Peters
Well-known member
r2d2 said:
Not considered by whom?
And remember -- things are designed to a price point, and also remember that there's more to a circuit than just the choice of op-amp.
Depends on where and what you read. A lot of people have opinions, and those people have internet access and can post those opinions on forums. It doesn't mean that the opinions are correct, though.
-a
hakanai
Well-known member
and even the most anti-social toob hound will usually admit that if you can't make a great sounding album with a SSL, it's not the SSL's fault.
Thanks for post "Andy",
some time ago i have had between the hands some ssl Xrack modules
on the cards/pcb i found many many smd components , ICs too , and not "socket" type , but soldered ,
(very very less expensive than "normal" components)
very much more difficult for repair by a not expert tech guy,
is this "design to a price" ?
i think is "design for pockets"... ,
And many Neve too like 8816 are full of smd (only 2 transformers in the LR out) ,
and new 500 version of the 1073 pre-eq and other modules are made with smd
same components placed on mackie & co.... units
yess "Hakanai"
a great album
(well composed,well played with top musicians with best instruments ,well singed by great singers ,)
but recorded with "low cost" technology (microphones ,preamps,comp,eq ......)
cannot sound well as the same album recorded with "top" technology and a good engineer too (of course)
a great sound adds a lot of value to the music
ssl and 5532/5534 apart
mixing with a mci 500 desk is the same that mixing with a mackie ?
otherwise would be like saying
"a ferrari is same as a Volkswagen" ...
"a manley eq or comp is same as a behringer"
what you think about ?
peace
R.
some time ago i have had between the hands some ssl Xrack modules
on the cards/pcb i found many many smd components , ICs too , and not "socket" type , but soldered ,
(very very less expensive than "normal" components)
very much more difficult for repair by a not expert tech guy,
is this "design to a price" ?
i think is "design for pockets"... ,
And many Neve too like 8816 are full of smd (only 2 transformers in the LR out) ,
and new 500 version of the 1073 pre-eq and other modules are made with smd
same components placed on mackie & co.... units
Andy Peters said:
yess "Hakanai"
a great album
(well composed,well played with top musicians with best instruments ,well singed by great singers ,)
but recorded with "low cost" technology (microphones ,preamps,comp,eq ......)
cannot sound well as the same album recorded with "top" technology and a good engineer too (of course)
a great sound adds a lot of value to the music
ssl and 5532/5534 apart
mixing with a mci 500 desk is the same that mixing with a mackie ?
otherwise would be like saying
"a ferrari is same as a Volkswagen" ...
"a manley eq or comp is same as a behringer"
what you think about ?
peace
R.
hakanai said:
r2d2 said:
Nobody is saying that a Volkswagen and Ferrari are the same, but they are both made from steel and bolted/welded together. The Ferrari by design is a very different vehicle than the Volkswagen. I submit that the Volkswagen is harder to deign, because performance vs cost is superior to the Ferrari.
Another point worth mentioning, low volume hand built units, will always cost more to build than high volume mass production. Do not confuse this high cost with higher performance. Generally consumers will impose a discipline that the more expensive products should not suck too, or they will quickly fade away. This is a secondary effect and not intrinsic to the process.
Getting back to your Ferrari/VW comparison there are raw performance metrics, and then ergonomic or human engineering aspects (like leather seats) that do not measurably affect the car's performance, but does help justify the high price.
I repeat my original question, why do you want a DOA, when modern high performance opamps generally outperform them in almost every way? Note: The 553x opamp has not been modern since the 1970s. (While it does probably define the modern age of opamps faster than audio signals).
JR
Andy Peters
Well-known member
r2d2 said:
Oh, where to start. Let's see.
SOCKETS SUCK. The only reason to use a socket is because you expect to have to replace the component during the product's lifetime. We used to use sockets for microcontroller program memory storage EPROMS. Now that the microcontrollers use in-circuit-programmable flash, EPROMs have gone away, and with them, the last reason to use a socket. And praise $DIETY for that. Especially when the cost of a quality socket was generally more than the part you plug into it!
So say that you've got something on the bench and it doesn't work. You suspect that an IC has failed. Certainly, if it was socketed, you could pop it out and replace it with another. But, if that's the case, then you have to ask why the component failed, and then you'll be able to prevent it from failing again. These parts just don't die for no reason. So it's worth thinking long and hard before shotgunning parts.
SMD construction is superior to through-hole in pretty much every way. You can't do modern high-speed digital in DIP because the lead inductance alone kills your performance. Your board density can go up so board sizes shrink, and that's a cost savings. Certainly hand-assembly is possible with SMD but why? An assembly house can use automation to stuff boards faster and with much higher reliability than hand-assembled work. This notion that "hand assembled artisan electronics" are better than machine-assembled kit is rather silly.
Any competent tech can do SMD rework. It's not hard and the tools are readily available. Now, excuse me while I go and mount a BGA-packaged FPGA to this PCB ...
Well, it's not like Neve has a factory which makes custom parts!
-a
You have to understand that DOA's have not been invented for "color".r2d2 said:
In the jurassic times of audio opamps, design was influenced by the early works of the pioneers, and objective performance (BW, THD, noise, output capability...but also input current, offset...) was paramount. In particular API made sure their 2520 was capable of driving a low impedance (150r) in regard of their use of 1:2 (even 1:3) output transformers.
When the TDA1034 (a.k.a. 5534) came out in the early 70's, its performance was very similar to the 2520, and it became apparent that DOA's had no significant operational advantage over monolithics; that's why most designers gladly used them without moderation.
30 years later, when everything done in the 60's & 70's was so much better, it became systematic to bash everything new, so now tubes, discrete transistors, carbon-comp resistors and transformers are given a new lease of life.
In order to sustain a whole old-new industry, there must be a justification, based on unmeasurable parameters, such as color, transparency, sound-stage width, "ballz", and so on.
However, there are some instances where a DOA can be plainly justified, in particular if it's designed for a specific optimization.
Although most DOA's have noise performance that is just about the same as a 5534, there are DOA's with very low EIN voltage, that makes them particularly suitable for low impedance sources or summing amps. However, there are more and more IC opamps with equal or better noise performance than these DOA's.
The most common justification for DOA's is the fact they can run on +/-24V rails, thus conferring a maximum undistorted output of about +24dBu, or 4dB more than what can be achieved on +/- 15V rails. Indeed there are several IC's that run on +/-22V but they are rare and pushing on reliability. Higher voltage opamps are not new but are generally expensive and may be lacking in some aspects of their performance (noise). Only recently we have seen a new high-voltage IC opamp (ADA4700)of high performance, that runs on +/-50V rails, for a max output level of about +32dBu. Noise performance is not that good.
A DOA designed in accordance with the advances of technology made in the last 40 years, should be supremely transparent, just like are the über opamps (LME49710, ADA4898, OPA1611).
Andy Peters
Well-known member
abbey road d enfer said:
... and we need this why?
-a
Different people, different experience. I have used sockets for years; there was a time when IC's were not as reliable as they seem to be now. And the value of sockets for servicing does not have to be demonstrated.Andy Peters said:
That's a comment that comes quite often. I have always used the simplest sockets, which cost a fraction of the price of the IC. The few times I have used tulip sockets, I have experienced leads breaking and being stuck in the socket when swapping. I have never seen an IC falling off a non-tulip socket.
Indeed. That's part of the scientific approach. Using sockets has never led me astray.
These points are indeed valid on a manufacturing point of view. However, for a designer who does not have the means or resources of having prototypes assembled industrially, SMD's are a challenge. Working with binoculars and hot-air is not my idea of fun. In comparison, working through-holes is a vacation.
I do a number of different designs, some are complex some are not. I would choose through-holes for the simpler projects.
Agreed.
Because the mktg dept sez it looks better on the specs.Andy Peters said:
The marketing department says that because it resonates with the phoolish consumers. They think that more signal voltage means better signal above a presumably constant noise floor.
They take little notice of modern digital CODECs running from low singe digit DC power supplies, delivering SOTA dynamic range.
Faster horses, older whisky, more money.... etc... 8)
JR
They take little notice of modern digital CODECs running from low singe digit DC power supplies, delivering SOTA dynamic range.
Faster horses, older whisky, more money.... etc... 8)
JR
