Best OpAmp for Audio Alchemy DITB

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
stephensank said:
Many guitar people swear by the JRC NJM4558 as the ultimate. Might be the worst sounding chip ever for music playback, but I have no difficulty believing it's great for electric guitar applications. So, those two guys recommending the 5532 may certainly have their points, but not in any credible way can their opinions be related to music reproduction, which I gather is the focus of this particular forum.

Actually most of us guitar guys think the 4558 is a really lousy chip. Most of us ditch them for the 5532 or 2604. The 4558 sounds really good in a handful of distortion circuits but in general, it's often despised. Some guys also love the TL072 which is another lousy chip.
It's the industry that claims the 072 and 4558 are such great chips, not us gearheads building and modding our amps.

Now my amp like many others is equipped with a CD player in and is stereo. I still swear by the NE5532OPA even in audio applications. I suspect the Hi Fi guys loving these other chips love them because they add density to the overall body of the tone, characteristics of this nature. As a musician that demands the most intricate detail so I can hear every perfection or flaw to my playing, these fattening chips just don't cut it. The 5532 has a very balance tone and while I admit it's sort of thin, that's due to it's clarity and detail, it delivers the most accuracy.
A lot of guitar guys prefer the 2604 or 2228 because of the fatness they add to the overall tone but plenty like the 5532. The other chips may add a fullness but the 5532 allows me to hear each instrument more distinctly or the vocals more distinctly.
It could be described as a more sterile sounding chip, that's for sure.

If we as guitarists didn't care about resolution, we'd be building tube amps and pedals where clarity and frequency definition go out the window and it's all about natural distortion and tube overdrive which vastly degrades the sonic quality of the signal but results in a very thick, pleasing tone that is very forgiving to a players weaknesses and in fact makes what they're doing sound even more impressive than it actually is.

It all comes down to WHY you think this chip or that chip sounds so good. In the DIY Hi Fi world, I think it's about having both detail and a fat richness. On the musicians side, it's generally about choosing one more than the other or even over the other.

.
 
Hello all !!
I have just completed my DITB mods with the folowing:

1. Added 220uF Nichicon Muse FG in the incoming PS section.
2. Transferred the 0.1uF PP to the underside & added further 220uF Nichicon Muse FG to the incoming PS
3. Added 470uF Nichicon Muse FG to Dac analog PS with existing 0.1uF underside each 470uF.
4. Added 470uF Nichicon Muse FG to Opamp PS with existing 0.47uF underside each 470uF.
5. Replaced the OP-275 with OPA-2134PA.

Now - how does it sound ??

First impressions are that the color of tone changed for the better, much richer, nice low end and wonderful upper end without a hint of sibilance. But, I kind of feel there's a loss of airiness which was there earlier in my DITB. Also, imaging may have also taken a "slight beating" but am still not sure on this.

Can someone tell me what series PP caps are/may be necessary to bring about the best from this BB OPA-2134?

Thanks to all of you for helping me reach my goals!

Cheers!
arup
 
" ... what ... may be necessary to bring about the best from this BB OPA2134? ..." ... or 4558 or 5532 or 5535 or 4560 or any op-amp, FET input or not.

It looks like you have already taken care of this as you already refered to several 0.1 and 0.01 uF caps in your mod ... just make sure they are plastic caps:

See: http://focus.ti.com/lit/ds/symlink/opa2134.pdf ... top of page 8 is a simple test circuit diagram. Note the presence of those 10nF (0.01 uF) caps on the power pins. This is becoming SOP for better quality audio op-amp designs and applications, adding plastic caps to the power rails (PP = polypropylene, polystyrene or MKT type). These should be "close coupled" to the op-amp power pins ... I install 'em so they look like they are making love = so close you can solder both connections with one pass, like the cap through holes are within 1/10" of the power pin holes ... These plastic caps are in parallel with the "standard" electrolytics usually placed on the circuit boards as additional power supply decoupling / filtering.

This applies to just about any single or multichannel op-amp or chip amp. The various quality op-amp makers, TI/Burr Brown, Analog, etc., are all adding plastic caps like this to their test circuits 'cause that way they get better specs. The "rule of thumb", if you need power filtration on your board, do so and add these plastic caps as well, if think you don't need 'em, add 'em anyway. Figure: for every 1uF of electro, add 0.001uF to 0.010uF of plastic ...

Bob Pease of Electronic Design Magazine has written several columns about exactly this and there as a bunch of Internet references as to which type of cap works best ... sometimes making reference to the fact that this makes all the "poor quality" electrolytic caps look good. :smash:
 
" ... 470uF of lytic ... plastic should be 0.47 to 4.7uF [?] ... I have 0.47uF. Should I add another 0.47 to the existing one? ..."

As the electrolytics get bigger, the plastic caps don't have to "keep up" ... and one 0.47 uF (470 nF) for each power rail should be sufficient ... unless you have more than one op-amp chip on your board, of course, then you can easily, reliabily "split the difference" ... 0.22 uF for each rail, each op-amp = total of 4 plastic caps on two op-amps, etc. :D

That "rule of thumb" is pretty flexable. As long as the plastic caps are greater than about 0.047 uF, each, the higher frequencies will be "sharpened up" and the overall parallel capacitance need not be "compounded" ... It is a case where enough is enough and too much won't hurt, but gets to be ah, well, too much. ... I use 0.022 uF (22 nF) everywhere, I buy these by the handfull, about US$0.25 each. I try to get the yellow MKT types or the redish/orange poly types with ratings above 100 volts. Two X 0.022 uF is enough for any single op-amp, but 0.47 uF is not really "overkill". :smash:

(I wish I had some of those liinks to Bob Pease's articles ... you can see the effect on a 'scope. You can see a difference with just 0.001 uF [ 1 nF]. The last project I did this to was an Oppo '981 DVD player, adding two 0.022 uF MKT types to the +/- 9 VDC analog DAC power bus = and it made some difference = better highs, a better sound overall IMOP.)
 
arupg said:
Hello CT
Thanks for your feedback. Which 2132 did you use - "PA" or "P" ??
Also, did you try OPA627/637 or any AD types. Please report your remarks as it will help me.

Thanks in advance!!

--------------------
AG


I have used both in two different SACD players (sony C555es and Denon 2900) but did not compare them directly.

I have used 627 in my headphone amps. They are fantastic. But I have not done a direct comparison with 2132.
 
2132 vs 2134 and 627

hi all,

I have tried 2132 nd 2134 some time ago and find them to be quite good, but can't tell if I can differentiate between them. However, there is very big different with 627/637, they are just a whole class better. I have 2846 and THS 4032 to try, but they are SOIC and i am not sure if I can solder them properly into DIP.
How would any one think about 4580 which is i Behringer? Sud I change them to 2132/4?

I still like my 15 years old LT1057...

Best rgds

William Lee
 
Hi Guys!

I know, this is a very old thread. But I recently purchased a DITB (not arrived yet) and I had a look at the inside (picture from Peter Daniel) and the schematic (see below), and I wonder: Are there no output coupling caps after the OP275 output stage?? If so, how does this unit handle dc offset on the output terminals?

This is important to me because the amplifier I'm currently using also has NO input coupling caps! (This is because every source I know has either output caps or a dc servo.)

Anyone?

Thanks a lot!
martin
 

Attachments

  • DacInTheBox - Inside.jpg
    DacInTheBox - Inside.jpg
    94.3 KB · Views: 362
  • DacInTheBox - Schematic.pdf
    186.8 KB · Views: 164
two things I would do.
1.) measure the DC resistance at the output of the DITB.
2.) measure the output offset over a range of operating conditions.



1) Output R sits practically at 100 ohms.
2) negligible offset - not worth the cap - hence have left it flat open to DC.

Happy that way - helps with the Low Freq esp. after moving to 12V reg for the analog power.
 
two things I would do.
1.) measure the DC resistance at the output of the DITB.
2.) measure the output offset over a range of operating conditions.

Thanks Andrew!
You are everywhere, aren't you? ;-)

I will do that of course as soon as I get the DAC. But I was just wondering if there is a theoretical explainaition about why this DAC in particular doesn't need neither output caps nor a dc servo to handle dc offset... Seems a bit odd to me...
 
1) Output R sits practically at 100 ohms.
2) negligible offset - not worth the cap - hence have left it flat open to DC.

Happy that way - helps with the Low Freq esp. after moving to 12V reg for the analog power.

I read about a different approach to improving the analogue-out of the DITB which involved changing forn the DAC chip(s) V (voltage) -out to its I (current) -out with a resistor load.

This is supposed to allow two benefits;

It bypasses the on-board opamp circuitry of the DAC chip and allows one to choose what voltage is fed to the op-amps, and hence the DITB's overall output.

One could drop the DAC's line-out from it's (IIRC) 2.4v to, say, 1.5v which would then make the standard 5v rails adequate for the purpose, which it isn't for 2.4v.
 
@ Aleksunder

Indeed - your method is the alternative one. In the DITB, the AD1860s dacs do the I/V themselves internally. Pulling out the "I" off the dac chips would work best if one was to use a top-notch transformer to couple to the output jacks. Simple, minimalist yet great sound is what I can surely predict BUT all depends on the transformer quality!

===========================================================================


I've been very much delighted with the level of uplift this tiny DITB has provided me by using a completely re-done PSU that is at +/-15V DC 2.5 Amps. The PSU is extensively regulated right from using HEXFREDs rectifier diodes, top-notch electrolytics along with Multicap double bypass.

A similar sort of attention was also given to the DITB board & the end results are just awesome!! Can't believe its still that $200 dac from the mid '90s.

DITB owners' would be highly recommended to go either way, i.e. using the voltage out off the dac chips OR taking out the current off the dac chips and then to outputs. However, you still need a "stealth" PSU for best results.
 
Discussing the uplift in sonics the DITB has provided with another stalwart who inspired me to do this, we agree that R2R ladder type dacs did offer a superior performance wherein mods were accepted with grace. . . and similarly performance kept shooting up & up !

Pity modern R2R ladder dacs aren't there anymore either from AD or BB/TI. :sad:
 
@ Aleksunder

Indeed - your method is the alternative one. In the DITB, the AD1860s dacs do the I/V themselves internally. Pulling out the "I" off the dac chips would work best if one was to use a top-notch transformer to couple to the output jacks. Simple, minimalist yet great sound is what I can surely predict BUT all depends on the transformer quality!

===========================================================================


I've been very much delighted with the level of uplift this tiny DITB has provided me by using a completely re-done PSU that is at +/-15V DC 2.5 Amps. The PSU is extensively regulated right from using HEXFREDs rectifier diodes, top-notch electrolytics along with Multicap double bypass.

A similar sort of attention was also given to the DITB board & the end results are just awesome!! Can't believe its still that $200 dac from the mid '90s.

DITB owners' would be highly recommended to go either way, i.e. using the voltage out off the dac chips OR taking out the current off the dac chips and then to outputs. However, you still need a "stealth" PSU for best results.

I have to confess I've never got around to any of the mods, despite owning two of them (one with a single dual-channel DAC, the other (slightly later) has twin single-channel ones).

I suspect how well the DITB works in a given system depends on the input they're feeding (decoupled or not, impedence etc') - I've heard comments on it (in unmodded form) ranging from "one of the worst outboard DACs I've ever listened to" to, well, very good!

Mine is between a PC (fettled for audio) and a Rotel RA820BX4 intergrated, and I love the music the combo makes, maybe the Rotel's line-in suits the DITB (I know it's a "proper" c.47 Kohms).

Having said that, my intention has always been to try the conversion to the DAC's I-out, roughly halving the overall output to c. 1.2v, which I think could be done without making it irreversable.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.