discrete opamp class A

branet · 2011-03-28 9:30 am

I'd like to replace the chips in my cd player.
Now there are three ne5534 I would like too replaced them with discrete opamp working in class A.
Something like burson discrit opamps.
If anyone has some suggestion?

abraxalito · 2011-03-28 10:06 am

Yeah, my suggestion is check out Samuel Groner's extensive measurements on opamps first. Burson discrete doesn't fare very well at all and he says it looks identical to Audio-gd OPA-Earth which is much cheaper.

SG-Acoustics · Samuel Groner · IC OpAmps

AndrewT · 2011-03-28 11:20 am

there are a number of discrete opamps in this Forum.
I think Peranders has a board for one as well.
Mox thread had one but it required two pcbs for a dual.
Maybe Juma has one.
Pass shows some.

Search, you should get quite a few hits.

branet · 2011-03-28 4:27 pm

There are many schemes but little with the PCB layout and that is also wath I need!
And class A it would be nice!

tiefbassuebertr · 2011-03-28 7:03 pm

branet said:
I'd like to replace the chips in my cd player.
Now there are three ne5534 I would like too replaced them with discrete opamp working in class A.
Something like burson discrit opamps.
If anyone has some suggestion?

branet said:
There are many schemes but little with the PCB layout and that is also wath I need!
And class A it would be nice!

poynton · 2011-03-28 7:33 pm

Try this from Ray (of the HUGE CD63/67 mod thread) :-

Discrete output stage

PCB is available, too.

Andy

.

AndrewT · 2011-03-28 7:58 pm

tiefbassuebertr said:
All stages of Op-Amp runs in general in pure class A, except such versions suited for power amplifier applications like TDA2030 etc.

where did you get this from?
Or did you invent that hypothesis yourself?

Mooly · 2011-03-29 8:04 am

branet said:
I'd like to replace the chips in my cd player.
Now there are three ne5534 I would like too replaced them with discrete opamp working in class A.
Something like burson discrit opamps.
If anyone has some suggestion?

Three 5534's is an odd (as in not even) number ? so they can not be just audio left and right channels. Are you sure they are all 5534 ?

Do you know what they all do before replacing them.

As to replacing 5534's in the audio path such as buffers/filters or I/V conversion I would recommend the OPA604 from my own experience providing they are suitable.

branet · 2011-03-29 10:52 am

Mooly said:
Three 5534's is an odd (as in not even) number ? so they can not be just audio left and right channels. Are you sure they are all 5534 ?

Do you know what they all do before replacing them.

As to replacing 5534's in the audio path such as buffers/filters or I/V conversion I would recommend the OPA604 from my own experience providing they are suitable.

You have the right chips are 5532.

Mooly · 2011-03-30 11:14 am

Try the OPA2604 for 5532's. Should be no problems as a direct swap.

Ken Newton · 2011-03-30 2:59 pm

As far as key criteria for choosing among the many (diy & pre-fab modules) discrete op-amp designs, I suggest the following three:

1. Required closed-loop gain of your application. In the interest of having no more negative feedback than is necessary, the great advantage of a diy discrete op-amp is that you can control the open-loop gain and linearity via local degenerative feedback techniques and thereby control the amount of closed-loop feedback which gets employed. The main problem I have with commercial op-amps, both monolithic and discrete module, is that they are designed for general application and so have very high open-loop gain - usually well over 100db worth.

2. The current required by the load. Driving headphones is a far different current demand than is driving the input of a power amp. High class-A current requirements will almost certainly drive you toward a buffered discrete op-amp solution over a non-buffered, or especially over a monolithic (silicon chip) op-amp.

3. Maximum output voltage swing. Again, driving something like headphones or a step-down transformer coupled circuit will require greater swing and might necessitate voltage supply rails which could only be handled by a discrete solution, but likely not.

The desire to control stage open-loop gain usually overrides the other concerns for audio applications and usually drives my to a discrete diy solution - typically of only three to five transistors complexity. As far as discrete pre-fab'd modules are concerned, I find the seldom mentioned Weiss units to look quite interesting, although, I've not used them myself. Price is around US$132, I think.

http://www.weiss-highend.ch/op-amp/documents/flyer_bipolar_r1.pdf

http://www.weiss-highend.ch/op-amp/documents/datasheet_bipolar_r1.pdf

dutchaudio · 2011-10-23 6:11 pm

Are they really worth their high price?

I was a little sceptical but ones inserted , they will not be replaced with those plastic opamps again, never
Placed thies into my DAC AD1955 24bits/192Khz
For I/V configuration 2x OPA2604 and one burson for the output was a OPA2604.

More then over 10 times i feld realy into a shock, by the first half an hour warmingup (they can become nice heated up)

Messuring the opamp is not an option LISTEN TO IT completely differend story, they are producing music not nice diagrams.

Not any opamp come near this burson (i had opa2604 but now they are in the trashcan)

They need burnin time to get all out of the opamp

It is just as nice focusing like vinyl so easily do.

Bass is very very tide, mids and hights are amazing.

Duets voices are now clearly absolutely seperate with the opa2604 it was one setting one direction one big mess.

No opamp for me made from platic

An externally hosted image should be here but it was not working when we last tested it.

Only one disadvantage i could not get the opamps with in the casing.
I dont care ìt is standing in my audiorack so you can`t see it.
When tubes are inside like mine, it is a bit tricky with young kids (i don`t have any)

rabbitz · 2011-10-24 7:19 am

The Burson doesn't work well with low voltages but give them over ±20V and they start to perform. Still, more trouble than what they are worth as they can be a bitch to locate in the case plus there are heaps of good op amps out there for CD player use.

dutchaudio · 2011-10-24 9:51 am

rabbitz said:
The Burson doesn't work well with low voltages but give them over ±20V and they start to perform. Still, more trouble than what they are worth as they can be a bitch to locate in the case plus there are heaps of good op amps out there for CD player use.

I realy have other findings with thies opamps, even at low voltage.
i Don`t care about messurements, my ears are the best judgement.

So maybe i make another psu for the bursons it will make a lot of difference?
A higher voltage psu, maybe ...

Strange you comment on your country made opamp but your honist hehehe

jcx · 2011-10-24 2:57 pm

but did you actually decide only using your ears?

to do that you have to listen blind - not knowing which is which, with level matching

fast switching without sonic clues/artifacts - really requires at least 2 complete circuits to compare

the Burson's could be technically bad enough to give audible differences - unlikely though

it is easy to bias Class AB op amp output stages to Class A for most DAC loads with a pull down R or ccs, all other internal stages do operate Class A in op amps, better op amps for DAC I/V require good layout, bypassing because one of the most important specs is high GBW

the AD1955 app circuit in the datasheet shows AD797

dutchaudio · 2011-10-24 3:06 pm

yes on ears only.
Only swaping electronic parts, for others and then compare it
Most expensive highend is also judge by component switching out for others
I upgrade audio from 1980 till now seen a lot of bad stuff and very good ones.

jcx · 2011-10-24 3:33 pm

if you knew which op amp you were listening to after each swap you were also using your expectations, you can't control your brain from using the knowledge - and won't be aware of the bias - it a universal human condition

you would at a minimum have to have someone else swap the components where you can't see, you shouldn't even be able to see them during the tests due to the "Clever Hans" effect, the only "communication" should be the start, end of each test in the series

usually a series of tests, where you don't know if the parts have been swapped or not, in any order need to scored, compared with the hidden list of which trial was which circuit after the trials

20 trials is a good goal to reduce random chance in your results

this blind testing is so inconvenient it is best to 1st use good test equipment to verify you have removed all measurable errors not specific to the parts choice - Audio DiffMaker and a good ADC can help

rabbitz · 2011-10-25 8:04 am

dutchaudio said:
So maybe i make another psu for the bursons it will make a lot of difference?
A higher voltage psu, maybe ...

Here's some quotes from a 2007 email from Jack Thomas at Burson when I was using them in a CDP with ±12V supply.

Quote #1
1. Can you please confirm if the 12V power supply is only supplying power to the opamps? If that is the case try to change them to 15V if you can.

2. Our opamp do require more power compares to the standard IC opamps. So, have you try to add a bigger/better quality power cap to the power supply?

To sum it up if you can get the power supply right for our opamp you will have superior result than making them into an external buffer unit.

Quote #2
Yeah, I agree that 12V power supply which share by many other componet may be stretching it. Our buffer supplys 20V +/- to each module

I'm not knocking the Burson but it's important that the power supply is right and doesn't share the power supply with other parts of the CDP. In addition, unless there is a lot of real estate in the CDP case, they can be difficult to install and be located in a suitable position without long leads where as a op amp swap doesn't have these issues. If they can drop straight in then it's fine for sure.

In fact the modules from the Burson Buffer are superior to the op amp replacements but not really suitable for inside CDP use. Great for using inside an active pre. The one I did used around ±20V rails IIRC.

discrete opamp class A

Member

Member

R.I.P.

Member

Member

Member

R.I.P.

Administrator

Member

Administrator

Member

Member

Member

Member

Member

Member

Member

Member