Jump to content

Discrete op amp?


Guest sacd lover

Recommended Posts

Guest sacd lover

Can someone explain to me what constitutes a discrete op amp? Secondly, does this discrete op amp actually offer better performance than a typical op amp?

KG, Nate .... anyone?

Link to comment
Share on other sites

The burson and audio-gd are opamps made from discrete parts.

But so are dynalo, dynahi, dynafet, B22 and a whole bunch of other amplifiers

including parasound,krell, levinson and all other fully dc coupled solid state amplifiers.

I can always design something better with discrete parts than a single piece

of silicon. Can't get it in that form factor however.

Link to comment
Share on other sites

The burson and audio-gd are opamps made from discrete parts.

But so are dynalo, dynahi, dynafet, B22 and a whole bunch of other amplifiers

including parasound,krell, levinson and all other fully dc coupled solid state amplifiers.

I can always design something better with discrete parts than a single piece

of silicon. Can't get it in that form factor however.

Um while this is definitely correct, I think that the term more describes the burson and audio-gd products.

The term op-amp and discrete are quite an oxymoron, because opamp refers to the chip, and you can't have a discrete chip.

So when ppl use the term "discrete opamp", I guess there are really saying "discrete replacement for an opamp chip", i.e. the burson and audio-gd products.

Hope this helps.

Link to comment
Share on other sites

How does opamp refer to a chip? Opamp refers to a specific type of amplifier topology, the center of which is a differential pair, which has NFB fed back into it. Generally though, an opamp is attributed to having near infinite open loop gain, so I dont know that amps like the dynalo/hi fully meet the definition, though the argument could be made either way, and is ultimately unimportant. Your chip is known as an IC opamp.

Link to comment
Share on other sites

How does opamp refer to a chip? Opamp refers to a specific type of amplifier topology, the center of which is a differential pair, which has NFB fed back into it. Generally though, an opamp is attributed to having near infinite open loop gain, so I dont know that amps like the dynalo/hi fully meet the definition, though the argument could be made either way, and is ultimately unimportant. Your chip is known as an IC opamp.

Ok my bad I got it wrong then. :palm:

Sorry for the confusion OP.

Link to comment
Share on other sites

Can someone explain to me what constitutes a discrete op amp? Secondly, does this discrete op amp actually offer better performance than a typical op amp?

KG, Nate .... anyone?

I have been looking into this more also and have the same question. In particular, I have the North Star m192 mkII DAC, which is supposed to be "better" technically (e.g., better chips, better power supply). But it does not have a discrete output stage (I think I'm saying that right) unlike the mk1. I guess my question is, why would they use ICs for the output stage on the "better" version? As used above "better" is according to north star.

The burson and audio-gd are opamps made from discrete parts.

But so are dynalo, dynahi, dynafet, B22 and a whole bunch of other amplifiers

including parasound,krell, levinson and all other fully dc coupled solid state amplifiers.

I can always design something better with discrete parts than a single piece

of silicon. Can't get it in that form factor however.

Kevin (and anyone else), in addition to my question above, is there a specific area of improvement with a discrete design?

Um while this is definitely correct, I think that the term more describes the burson and audio-gd products.

Kevin, meet peanutbutter, peanutbutter meet Kevin Gilmore.

Link to comment
Share on other sites

I have been looking into this more also and have the same question. In particular, I have the North Star m192 mkII DAC, which is supposed to be "better" technically (e.g., better chips, better power supply). But it does not have a discrete output stage (I think I'm saying that right) unlike the mk1. I guess my question is, why would they use ICs for the output stage on the "better" version? As used above "better" is according to north star.

Kevin (and anyone else), in addition to my question above, is there a specific area of improvement with a discrete design?

I am not Kevin, so I will have to beg your forgiveness (pleasedontsueme), but I feel like taking a poke at this. I am just guessing at North Star's reasoning for using IC opamps for the output stage on the Mk2 v. Mk1, so could certainly be wrong.

The North Star m192 mk1 used a voltage out DAC, CS4396 IIRC. Discrete output stages for v-out DACs are very easy to design, as the output requires nothing more than a buffer and LPF, and there are plenty who even feel that that is redundant, and take the output straight off the DAC. You can make a great buffer with just a pair of JFETs, or a simple discrete opamp, or whatever floats your boat. Works great.

The North Star m192 mk2 uses a current out DAC, PCM1792 IIRC. Discrete output stages for i-out dacs are much more difficult to properly design, though there are a few approaches you can take. Common base/source amplifiers are a good way to do this, though I am sure there are those who use simply a discrete opamp. Some also use a passive I/V and discrete gain stage follower, though this really isnt a good approach (IMO). It is much, much easier to just use opamps for the I/V conversion, and they work well enough for most.

With a discrete design, you get to choose everything. If you dont want to use NFB, and there are some good reasons not to here, you dont have to. Its more about choices I guess. You can also bias discrete parts much hotter than the transistors inside an IC opamp.

Link to comment
Share on other sites

I have been looking into this more also and have the same question. In particular, I have the North Star m192 mkII DAC, which is supposed to be "better" technically (e.g., better chips, better power supply). But it does not have a discrete output stage (I think I'm saying that right) unlike the mk1. I guess my question is, why would they use ICs for the output stage on the "better" version? As used above "better" is according to north star.

Its a cop-out answer, but IMHO it all comes down to implementation. A well implemented IC OPAMP might not be 'perfect', but can be very, very good. They offer an incredible amount of functionality in a very small package - have a look at the schematic on the first page of the OPA637 data sheet as an example. Implementing something equivalent using discrete components can be difficult - but if done well can be better thanks to better part matching, tolerances and designing something EXACTLY for the specific purpose.

[EDIT] cetoole just said it all a lot better and in-depth than I ever could :cool:

However, I really think this trend towards using discrete OPAMPs like the Burson, Audio-gd etc. as drop in replacements for IC OPAMPs is a bit silly. They are often installed with long air-wiring, potentially use a lot more power than would normally be drawn by an equivalent IC...... I figure if you specifically want a discrete design, buy something built from the ground-up to be discrete.

Link to comment
Share on other sites

Are you saying that slapping in some discrete op amps in my ish won't be like slapping a Type-R sticker on my v4 Camry??? Because that noise is like NOS^2.

If you slap a Burson in it, it will be like putting a "bend over" sticker on, cuz at their price, you just did :eek:

Colin gave a very good answer, BTW.

Link to comment
Share on other sites

Are you saying that slapping in some discrete op amps in my ish won't be like slapping a Type-R sticker on my v4 Camry??? Because that noise is like NOS^2.

Uh, sure...... why not...... :)

*

Just to add to my reasoning above. Whenever I see talk of discrete OPAMPs I think of AMB's designs. Sure, an MMM with discrete OPAMPs might be a good upgrade over the default AD8610. But the extra cost has pushed you up into B22 territory. Buy/build something well implemented in the first place, and use the right tool for the job, then you won't have to worry about what the technology behind it is or pay for expensive (and possibly pure snake-oil) upgrades.

Link to comment
Share on other sites

I have been looking into this more also and have the same question. In particular, I have the North Star m192 mkII DAC, which is supposed to be "better" technically (e.g., better chips, better power supply). But it does not have a discrete output stage (I think I'm saying that right) unlike the mk1. I guess my question is, why would they use ICs for the output stage on the "better" version? As used above "better" is according to north star.

There's actually a very simple answer to this, in the case of the m192 MKII. In between the MKI and MKII Northstar released a new CDP the Sapphire. The MKII uses the exact same dac and output section as the Sapphire. Seems pretty likely to me that it was mostly a cost savings move.

Link to comment
Share on other sites

I kind of disagree with DR gilmore's statement that the dyna(whatevers) and some of the others mentioned are op amps. I dont think they have enough open loop gain to satisfy the condition of opampishness.

To Dr G's credit, im sure that if his only goal was to make 60dB or more of open loop gain he could have designed another gain stage into the amps only to cancel it out with more feedback later on. there are no free rides: are the more complex harmonics worth it? I'd say not.

on that note, not everything in an IC package is an opamp either.

Kevin (and anyone else), in addition to my question above, is there a specific area of improvement with a discrete design?

With a discrete design you can do a couple things you CANT with an IC.

At the end of the day, it boils down to the designer having control over what goes into his (her...) product VS selecting the best part that may be made for something else.

1: You can unconditionally guarantee that EVERY transistor in the gain stage(s) runs class-A no mater what. Many IC opamps run the internal transistors class-A but who really knows how hard?

2: In the design stages you can "roll" transistors in a discrete design, where you just swap the whole IC.

3: Since all of the parts are separate, they can dissipate much more heat individually. You can safely "burn" off a couple watts in the gain/driver stages of a discrete design if thats what it takes to make the amp work well... you only get a hair over a watt TOTAL dissipation with a dip-8 package.

4: If you are designing an amplifier as opposed to an op amp, you can build to suit a specific gain. you also have access to the internal transistors, so smaller local feedback loops (around 1 stage) can be used rather than 1 big one which is typical with an op amp.

Link to comment
Share on other sites

One day you'll be able to fund your retirement by selling your amps. You know, just like all the "Joe Grado personal" HP2s that appeared recently. :P

Not funny. When i finally do retire i'm going to want to move to a smaller place in

a warmer climate. No way i'm going to move over 3000 sq feet of vintage and

rare electronics, mill , lathe, tons of tools...

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.