TDA1541A DAC chip corner

The spec sheet can be found here: TDA1541A LINK

This is the data mining page of mine:   DAC chip data mining

BLOCK DIAGRAM of TDA1541A/S chip

WHY TDA1541A DAC ?

THE BIG QUESTION: Mr. Fikus, are you dogmatic about using old shit and why you deny the technical progress of our wonderful industry? Are you some kind of a lunatic, a gaslight advocate? Are you a hi-fi Amish? Are you on drugs? What is wrong with you that you do not like 24 bit/192 KHz and HDCD and SACD and VLSI chips?

Here is what I can say: There are two kinds of resolution: digital and analogue. We can talk in digital "language" and conclude, that a digital information can have a certain - say 16 bit - word length. And the chip that reads it must be able to operate with 16 bit length of words. But can it (DAC) put out the resulting sine wave with similar accuracy ??? That is the KEY QUESTION.

Bits are the smallest measurement units which we can apply to the process. It is sort of like in digital cameras. Would you buy a 100 Euro Casio pocket camera which has a CCD with 10 million pixels or rather a 8 million Canon D30 XLR? The number of pixels on the casio is better, but it is all lost in firmware, compression, lens quality, accuracy of power supply, etc. In music reproduction, our CD has 16 bits. No matter what DAC we use - it is still 16 bits. That's means that we can reproduce a 2 V pp signal with accuracy of 2/65500 Volts.
So why the cheap DVD player with real 24 bits is not immediately better? Because increasing the digital resolution does not have a chance to come through the process. The fact that digitally speaking the chip can "understand the word -length of 24 bits" does not guarantee, that after reading it will output music of equally great nuance resolution. Remember - reading is not understanding. Just like with this text. Because the other elements which play role - the timing of impulses (clock) and the power supply stability and the analog stage - all have possibility to reproduce signal with accuracy of say 0,01 % which is 1/10000 which is far from 1/65500 and VERY FAR from 24 bit number, which is 2 to the 24th power, which is 65500 times 2 to 8-ght power which is A LOT. Actually it is 17 millions.
The non mathematical explanation is possible if we take the example of say the printer. Some printer manufacturer can claim the resolution of 20 000 dpi. This may be a valid claim for the ink injecting head resolution and it's software, but the "analog" result will be that the drop of ink will smear over an area 10 x bigger than its injector nozzle, and B) the paper structure has fibres 10 x bigger than the 20 000 dpi requires. So the picture limitation is in the paper ink interaction, not the head injector resolution. Not to mention accuracy of head movement on the rails, temperature of injectors, air pressure and humidity etc.

So somebody promising a resolution of 24 bits is effectively claming that he produced electrical device which controls the output signal to a wide range of loads with accuracy of one/17 millionth part of a volt.
THIS CLAIM IS NOT VERY SERIOUS. We take it with a grain of salt.
So dear readers, there is no correlation between increased number of bits above 16 and sound quality. At least it is not automatically guaranteed. I suggest we rather concentrate on the best usage of the bits which we have, than worry of these bits we don't have.

The taste in DAC chiops is something very subjective, but I have much more opportunities to compare DACS because I remove other variables like output stages, capacitors, opamps and stuff like that. I listen to NAKED CHIPS via the same lampizator every time, so I know how DACs compare.
TDA1541 is very strong contender to being the best, some rivals may actually come close, equalize or have an edge in ONE area but no DAC can really seriously beat the TDA. Yes, it is THAT good. Especially in NOS mode and with very good capacitors around it.

For those interested in the TDA1541 debate - S1, S2, A - non-A, fake or real - here is the must read link:
http://www.diyaudio.com/forums/showthread/t-32587.html



Philips CD304mkII (non-A chip) I LOVE the non-A, it is very vigorous.





Two A1-"S" in the Grundig 9009 (but only parallel, not differential)




Grundig 9000 with standard Philips PCB and the A chip.




This is the mother of all DACs - the fourteen bit TDA1540 as found in the Loewe CD 9000 and Marantz 73, Philips 103 and 104.




The S1 in the Revox B226S (decoupling caps - the red ones - already good !)



The Naim has very similar config to the Revox, but the PCB is of military style. An absolute overkill of engineering in all aspects.





BOTTOM SIDE OF Marantz 60 with new caps

squeezed in.
The marantz cd60 is almost identical to Marantz 40, 50, Grundig 9000, Philips 630. It is very simple and cheap in fact, but after electrolytic cap upgrade, decoupling cap upgrade and lampization - it can be VERY GOOD.
PCB is standard Philips one.



SMD caps (the small ones) and the "proper" MKP or MKT capacitor that replaces them.
- the very necessary job in cheaper CD players from second league.
Most players use here 100 nF, the better players use 220nF. I swear that 220 sounds better than 100. I suggest to use 470nF by 63 V from WIMA. Like on the photo somewhere way below - Grundig 8400 MK2. Nowadays I learned that the SMD caps can stay, we ADD the new caps across them. So a nice job would be adding 220n to the existing 100nF to make 320 nF) Without removal job - the likelihood of damaging the PCB is very small.
Another good option is to add the tantalums - something like 1uF by 10 V. But remember that the signal is polarized negatively (-5VDC) so put the cap plus to the ground. Definitely keep the SMD in place then. Tantalums have more dry but more detailed sound. Lots of thin air and micro clues.




The Marantz CD 40 DAC area before cap replacement / addition. . All the brown SMD caps can actually stay.
These extreme left one and extreme right one are in parallel (bypass) with large electrolytic mounted on the other side. These are power supply reservoir caps. Replace SMD with MKT (220nF/63V or better 470 nF) and the electrolytic too - with os-cons or tantalums, at least 47 uF/16 V) .  Observe polarity of electrolytics. Of course you can use Black Gates too.



In Philips' flagship DAC 960 - the non-S chip is surrounded by premium 220n caps and gigantic power supply.

This is identical looking PCB of Marantz 94 MK1 - just like the Philips DAC above - both made at the same time in the same Marantz Japan factory.  Non "S" variant of the DAC but the caps around are 220nF not 100 like everywhere else. This is key element of their sound signature.
The red dots here show the wire jumpers which we must lift for lampization, and yellow dot is a ground point.

Superb Danish engineering in the B&O CD5500. Premium parts and kosher solutions. 100 nF caps - white ones - around the DAC



Grundig 8400 mMKII - after removal of decoupling caps.


Grundig 8400 - capacitor part substitution.



Grundig 8400 - job completed.




Installation of 220nF decoupling caps in Philips player (and later - on the identical pcb - in Marantz 40, marantz 60, Grundig 9000 and over a dozen others, which share same Philips PCB) On the picture the job is not finished.

The biggest size of these caps that I found on any CD player is 220nF but one CD - Arcam delta DAC has 4 out of 14 caps sized 470 nF.(the orange tantals)



This is Arcam 70.2 DAC section with the Tantals visible. - the orange head caps in a row of white cubes around the TDA dac.

Why decoupling is important, what really happens there ?

The process of musical signal creation inside the DAC is explained on this drawing of the TDA1540 guts. All elements are inside the IC but the capacitors CAN'T be integrated that's why they are outside. The switches take some current from each cap and build up the music signal.



As you can see, the caps are DIRECTLY responsible for music, for the sound , color and flavour.



MArantz - who was owned by Philips who invented the TDA chip - used 100 nF in cheap players but 220 nF in their best ones.

A solution by Lukasz Fikus will be: 220 nF all 5 caps in each row, then 470 nF like Arcam, and last one 2,2 uF a'la Rogic.
I will follow up with some scope measurements.
Rogic DAC is HERE

One clever guy Thomas who tries to build a commercial DAC spent many years optimizing the TDA circuit and arrived at the dual differential balanced config with some GIGANTIC decoupling caps - In my estimate may be close to 1 uF but I don't know for sure - just looking at the pic: the blue cubes are huge.


The story of this  project (probably best TDA 1541A DAC ever designed) is here:
http://www.vt4c.com/bb/viewtopic.php?p=13568&sid=a105dc5d441c7e6c616e714c3b4ee190

Later on I found this PCB set to be used in SATCH DAC from Hong Kong  DIYHIFISUPPLY.com
I describe it HERE
Funny, the huge blue caps turned out to be 100nF by 250 V as the data sheet of philips advocated 100 nF.
Here is what Simon from DIYHIFISUPPLY had to say about the cap size :

In the end, the satch took 3 months to play FANTASTIC but this required very heavy mods including lampization of the satch.

ENTER "GREENDAC"

We can buy the GREENDAC for circa 15 USD and we need to populate only less than half of parts, but this time no mercy only the best parts need to apply.

 



This is Revox B126 with non-s but otherwise IDENTICAL board as B226-S. Best sound was achieved with a half of BB-OPA2604 as a converter and amplifier. Both outputs from leg one of opamp via PIO to RCA's. No lampization in this one, just bypass.


PHOTO: Absolutely superb player - Sony 227 ESD with two parallel TDA's.  Not "S" like Grundig, but still very good. Not differential like in Marantz 94/2 - sorry. But the caps, the grounding, the opamp arrangements -, the regulators - all is FIRST CLASS. Similar to Philips DAC960 and Philips CD880, Which were made by Marantz Japan.



This is a bigger brother of 227 - the Sony CDP-337ESD. Two "A" chips in parallel, sometimes - non-A, depending on the batch.


This is again a Sony, but this time it is a 555ESD. One chip, probably S, I must check it because it is covered.




One funny observation: after removing of all 17 capacitors from SMD board, the player still plays good sound, Not as good as with new 220 nF caps, but quite good indeed.


For the fans of TDA1541A: the absolute champion of TDA'ization is Cambridge Audio CD3 player from UK.
It uses 4 DACs in parallel.


It must be seen and heard to be believed.
See it here
CAMBRIDGE CD3



This picture shows the DAC board of Marantz CD-94-2 which has two single crown chips in differential balanced mode (no balanced output is available prior to lampization).


The best CD for beginners to start the adventure in my opinion is this one (equivalent of Marantz 40)  Grundig CD8100 - see it below.


Just imagine taking this lampized 20 Euro player to your buddy who has Wadia 860X and when you wipe the floor with the Wadia.

By the way, it has the same mechanism as the NAIM CDS1. Hard to find - 25th variant.



This is the inside of Musical Fidelity DAC  with TDA1541A. All parts are premium and layout is perfect, there is nothing to tweak really, only add tubes.




This is a CREEK CD60 player, photo courtesy of Aldo from Torino. It has the S1 and the 4/14 mechanism. What a JOY TO LOOK AT I guess it sounds superb on its own, even without tubes.
The parts and layout is one of the best , just like NAIM !!!
BUY THIS ONE !!!

And this -  tadadada drums rolling - is the "British Cottage Industrys best PCB design award" CD player - Sugden

Lampization of the TDA chip

6N2P

The short list of  CD players which have TDA1541A chip.

SOME HISTORY of TDA1541A DAC from Philips

"Desperation is the Mother of Invention." Isn't that how the proverb goes? Certainly it applied ten years ago in the case of the Philips engineers working on the development of the Compact Disc system. Given a specification that had included a 14-bit data word length, they had duly developed a 14-bit DAC chip, the TDA1540, only then to be informed that the CD standard decided upon after Sony joined forces with the Dutch company would involve 16-bit data words. (Thank goodness!)

Philips having already committed the 14-bit design to silicon, they would not have a 16-bit DAC ready in time for the medium's launch in the Fall of 1982. They were thus faced with the problem of squeezing four times the resolution from their existing 14-bit DAC. The result was an ingenious digital filter that combined 4x-oversampling and noise-shaping—the latter is effectively a digital feedback loop, the error produced when the filtered digital data are truncated to 14 bits being fed back to the beginning—to give a digital system with full 16-bit resolution.

Philips's true 16-bit DAC chip, the TDA1541, followed in 1985, but the seeds of ingenuity had obviously been sown: if the combination of oversampling and noise-shaping can increase the resolution of a DAC using too few bits, then why not go all the way and implement a system that used a simple 1-bit DAC and make up for the shortfall in resolution by taking the oversampling, noise-shaping process to the limit?

The result was a D/A system, internally referred to by Philips as "DAC3" (the two earlier systems were DAC1 and DAC2, of course), which was introduced in the summer of 1989. I discussed the design of the system in detail in June 1989 (Vol.12 No.6, p.57), but briefly, the SAA7321 DAC3 chip massively oversamples the input data at a 256x rate, interpolating the new sample values to produce a 17-bit datastream sampled at 11.02MHz (footnote 1). The data words are then fed to a 1-bit DAC, with the 16-bit error fed back in a noise-shaping loop. Mathematically, this should—and does—result in the pulse stream output by the DAC having the full 96dB+ dynamic range of a conventional 16-bit system. But unlike multibit systems, this "Bitstream" DAC is inherently linear and monotonic over its entire range—the reasons why were given in Peter Mitchell's "Industry Update" in January 1990 (Vol.13 No.1, p.36)—requiring no laser-trimming of on-chip resistor values or in-production adjustment of linearity, both of which add to a CD player's manufacturing cost, hence price.

I was told by Philips last year that the Bitstream DAC was therefore intended to be used in low-cost and portable players, the company saying that they would remain with their TD1541-based chip set for high-performance players. Very rapidly, however, once designers had tried the Bitstream approach, it became apparent that, correctly implemented, it could surpass traditional D/A conversion in the preservation of low-level detail. (Even Philips has now introduced a Bitstream player, the LHH500.)

ABOUT THE CD RED-BOOK

ABOUT THE NON-OVER-SAMPLING modification