advertisement


Ben Duncan's 'Supertuning CD' and 'Improved CD Electronics'

DoctorRad

pfm Member
I've 'digitally remastered' Ben Duncan's related 'Supertuning CD' and 'Improved CD Electronics' articles from the late 80s. These are concerned with re-engineering the output stages of TDA1541(A) 16-bit, 4x-oversampling players using Philips' standard chipset. In particular, it explains why simple parts substitution will only get you so far, and outlines a circuit which needs no DC blocking capacitor, instead offering a user-tweakable DC offset control.

I hope to re-draw the schematics for my own use, as I am musing over hot-rodding the output stage of the Rotel RCD-855 which should be arriving soon. I also hope to come up with a single set of schematics with equivalent parts correctly referenced from both texts. It's worth reading both articles as the style could hardly be different, and the content does not completely overlap. The Google document is open for comments, and if anyone else can make sense of the obvious typos, I'd be grateful.

My growing archive of Ben Duncan articles can be found here, including photos of the complete AMP-02 text.

Revisions to original document:
  1. 1st June 2020: Added 'Supertuning CD - A Refresher' - One year after ‘Supertuning CD', Ben Duncan returns to the subject of DIY improvements to the electronics of CD players. Includes notes on some (then) latest op-amp developments and a handy guide to swapping a TDA1541 for a TDA1541A.
  2. 2nd June 2020: Redrawn Hifi News main schematic.
 
Last edited:
Years ago Ben Duncan hot rodded some amps for me - it has to be said that he's the real deal, though the service wasn't cheap and he's quite a pushy salesman to boot!
 
Years ago Ben Duncan hot rodded some amps for me - it has to be said that he's the real deal
My personal viewpoint is that he takes and talks no bullshit - he does engineering properly in a way that makes sense to me but which I don't have the training or experience to emulate - only appreciate. His takedown of the standard Philips / Marantz TDA1541 output stage makes it obvious why these players are so heavily compromised.
 
Redrawn main schematic - removed NE5534 designations to avoid confusion - see below and article for opamp recommendations.

y4mTddi_8cTe64U14QI07WTRea3QUTjZqXTKEacvMIk0WCBwq4jlMELYXTAVBeDKUMeaiRPBXNaQnqOHGBotoufBWq2zFMVZWc5CI08Cezh7UsjfBwzfqd511Ll8RDfct5CtGoQgWqNH3EEuC8FgbIfBxfEXwD6UgputewySZJQ9riCqUOUHXrL5HfYGxufUbyWLv-BwwKDXkOzM45B_8Ak2w
 
Last edited:
Very interesting :), thanks for sharing. I'm now wondering if this will work on my Sony 337.

I'm sure that it would. +/-15V in the Sony and 2 off duel OPA604 Op-Amps.
 
Ben Duncan;s approach was good at the time; today it's not the best you can do for this dac, I'd say by a fair chalk.

I'd not lash this into a Sony 337 and think it better, for a start: The BD solution very context -sensitive to the original player it is based on; for a start, a bipolar opamp like the 5534 isn't the best choice for the I/V conversion, just about the best at the time.
 
...a bipolar opamp like the 5534 isn't the best choice for the I/V conversion, just about the best at the time

If you take a glance at the article itself, it goes into a lot of detail as to why the 5534 is the bare minimum you should use (i.e.: don't). He says that 'the performance of these chips is execrable in many areas, particularly the DC parameters', and goes on to describe its DC behaviour as 'disgraceful'.

I've used it as a placeholder in the schematic for various project-based reasons. At the time, the OP42 was the favoured op-amp, for the I/V conversion at least... that's what (the article says) it was designed for.
 
Last edited:
I'm now wondering if this will work on my Sony 337
Funny you should mention that, the 'epilogue' Revisited article specifically mentions that 'even up-market players like Sony‘s 337-ESD can benefit'... but as Martin suggests, don't go with 5534s. You can do a whole lot better, especially these days.
 
But the active crossover he designed is still in operation! Since 1983, with numerous upgrades.
When the central library reopens, I will be grabbing this and various other articles for the archive.

I hadn't realised until last week that in autumn 1990, Hifi News published a BD bitstream output stage project. I'd just arrived at university and had other fish to fry...
 
If you take a glance at the article itself, it goes into a lot of detail as to why the 5534 is the bare minimum you should use (i.e.: don't). I've used it as a placeholder in the schamatic for various project-based reasons. At the time, the OP42 was the favoured op-amp, for the I/V conversion at least... that's what (the article says) it was designed for.
The OP42 was indeed excellent for such uses (incidentally, what Naim used in their CD players for a decade or more)

One thing that the article doesnt give is why such a part works well -it is a FET-input opamp, and therefore handles the rf noise on the input more gracefully- less likely to suffer rectification effects, quite apart from the other figures of merit discussed in the articles.


Iirc Naim used the opa604s later eg in cd3 & cd3.5 /cd5; again, solid choice- and FET input
 
One thing that the article doesnt give is why such a part works well -it is a FET-input opamp, and therefore handles the rf noise on the input more gracefully- less likely to suffer rectification effects, quite apart from the other figures of merit discussed in the articles
The companion Improved CD Electronics article does mention that "Current-to-voltage conversion is best handled by a fast-settling op-amp with a (low trans-conductance) Bi-fet or degenerated bipolar input". This article, written later, mentions the following chips as being suitable for the I/V conversion stage: OP44 / 42, AD845 / 744, or Harris HA 2525 and HA 5137.
 
In my old Arcam Alpha 5 project I had OPA627 on I/V duties which were great.

I did try 2x OPA627AU on Brown Dog adaptors in the 337 but I wasn't convinced by the sound.
 
The 2604's are very decent. Nice analog flow to them. I've actually just swapped them back in after trying an 825/604 module. The 825 did give a bit more detail, but was too pushy for my tastes. I also tried it with local 0.1uF across the power pins, even more forceful.
 
If there is a way to directly couple the output from my 337 (or any Sony CDP) then please shout!
 
Well first measure the DC offset it has.

Does the input of your preamp have coupling capacitors?
 
If there is a way to directly couple the output from my 337 (or any Sony CDP) then please shout!
One of this circuit's explicit goals is to do away with an output coupling capacitor by using a precision DC current source to compensate for the DAC current offset, and also allowing for opamp DC offset adjustment if you want it.
 


advertisement


Back
Top