Philips LHH2000's secret output (versus CD100 & others with TDA1540)

[flavio81]

Of interest to people with TDA1540 CD players!!

Hi,

As some of you already know, the Philips LHH2000 professional CD player is a highly valued beast; some reviewer said the sound was identical to a SACD player (?) and there is one currently at ebay for like $14000 dollars (!).

I am in possession of a humble Marantz CD63 (1982, 1st gen), which has the same circuitry than the Philips CD-100, and Magnavox FD-1000. Those players have a good reputation, but not cult status as the LHH200.

The thing is, they all use the same chips: SAA7030 (4x oversampler), 2x TDA1540 (14-bit dac)... so in theory they should sound the same. But of course there should be a difference between the LHH2000 and others.

So i dug out the schematics for both. The differences are in BLUE on the attached image.

I'm not an electronics engineer but as far as i can see, the main difference is a transformer XLR output; before the two opamps that make the XLR output there are no big differences except for (1) using two decoupling caps in non-polarized configuration [in LHH2000] instead of one, (2) more decoupling caps or higher-valued decoupling caps, (3) using one chip per opamp instead of 2-opamps-on-a-chip.

The TDA1540 on both schematics was wired practically the same. Again, the 1540 got higher value decouling caps,

I don't know where exactly the I/V conversion is done in both schematics, i'm assuming it's done at the transformer on the LHH2000 and done at the 3579 and 3580 resistors on the CD100.

So where is the reportedly big difference in sound?

 

[flavio81]

Note: The image is big, expand it so you can see the component numbers better.

Also, a question for the electronics experts: What kind of transformer can be used in that XLR output stage (number of windings, number of turns, in/out impedance etc...)?

 

[Steerpike]

Interesting!
I have a Grundig, same circuitry, but it is reported to sound better than the Philips, apparently becase of better circuit board layout, and better power supplies.
The LHH may have similar issues. It may have much to do with the quality of the capacitors, etc, and not so much their values.
The NE5534 as superficially the same as the NE5532-dual, but you can buy 'selected' 5534's that are picked for superior perfomance.

Output transformers like that are usually 1:1 or low ratios, designed for impedances of 600R. They may have been made by Studer - who collaborated strongly with Philips on matters CD. Generally transformers don't improve the sound, and the very costly ones (which professional types ARE) are intended to degrade the sound as little as possible. In Peru, I cannot help out, but RS-Components of the UK sell 600-600 audio transformers.

Another notable difference is the muting and deemphasis on the LHH are switched by FETS, on the CD100 by relays.

 

[le´flu]

Hi Steerpike,
which Grundig do you have?

Thanks and greetings Ulf

 

[Steerpike]

@Ulf

which Grundig do you have?

The CD7500 in silver.

http://web.eject.co.za/s8nspawn/hifi/ak/#cd7500

 

[poynton]

Of interest to people with TDA1540 CD players!!



I don't know where exactly the I/V conversion is done in both schematics, i'm assuming it's done at the transformer on the LHH2000 and done at the 3579 and 3580 resistors on the CD100.

So where is the reportedly big difference in sound?

The i/v is performed by the first opamp after the TDA1540 chip in both cases.

Andy

 

[flavio81]

The i/v is performed by the first opamp after the TDA1540 chip in both cases.

Oh!! Hmm... so then the difference in audio is just better (& more) capacitors? Interesting. Note also that the resistors on the LHH2000 are precision resistors. For example "1.78K" instead of "1.8K" in the CD100.

Interesting!
Generally transformers don't improve the sound, and the very costly ones (which professional types ARE) are intended to degrade the sound as little as possible.

Exactly, that's why i was wondering if there was some coloration induced by the transformer that could alter the sound in a nice way.

Another notable difference is the muting and deemphasis on the LHH are switched by FETS, on the CD100 by relays.

Yes, but i don't think that would make such a big difference in the sound.

 

[poynton]

Interesting!


Another notable difference is the muting and deemphasis on the LHH are switched by FETS, on the CD100 by relays.

Yes, but i don't think that would make such a big difference in the sound.

In most modern CD players, switching is usually done by transistors ( generally bipolar rather than fet )

One of the more popular upgrades is to remove these transistors as they are acknowledged to have a detrimental effect on the sound - leakage, capacitance etc..

Relay switching is usually the better option and indeed is another tweek - replace the muting transistors with relays.

I always remove the relays when I have to scrap an old CDP.



Andy

 

[Zoran]

IV is performed with 1st op amp,
transformer at the end (and last dual op) is just for single-end to ballanced line out...
in the middle is reconstruction filter,
which can be override, because I dubt the transformer have high bandwith up to 44.1K
so You already have a filter?
cheers

 

[flavio81]

In most modern CD players, switching is usually done by transistors ( generally bipolar rather than fet )

One of the more popular upgrades is to remove these transistors as they are acknowledged to have a detrimental effect on the sound - leakage, capacitance etc..

Relay switching is usually the better option and indeed is another tweek - replace the muting transistors with relays.

I always remove the relays when I have to scrap an old CDP.

Exactly. But the CD-100 already uses relays, that's why i said i don't think using MOSFETs [like the LHH2000 does] would make a difference in sound.

 

[Richard]

Go with valves

Hi Flavio,

I have built a valve output stage for my Philips CD 304. I initially used a resistor for i/v but then changed to a transformer. I’m using a UTC KB-143 at the moment wired for 1:2 step-up and a small 20 ohm i/v resistor on the primary, this feeds into a 6N2P in SRPP, similar in theory to Audio Note’s current DAC output circuits. The 1540 DACs are in NOS mode and the player is clocked. I have tried various opamps but prefer the transformer valve combination. In fact it is the best sound I’m getting from CD and prefer it to my Marantz Cd94 MkII and a PCM 1702 DAC. I’ve also done many extensive structural mods to the player. I plan to try other valve circuits.

I do have an output board from a Studer player with transformers, but am so busy with work I haven’t been able to compare it. There a several good solid state options one can try, but be that as it may I’m still convinced that a properly designed valve output is in a different league. My friend is running a TDA1541A DAC with a choke loaded 6C45P and Duelund capacitor output and he reckons it’s on par with his Technics Sp10 Denon DL 103 DIY Schroder arm combination, but he may have been drunk ;-). As for the vinyl comparison I haven’t compared but I have heard the above mentioned DAC with RCA LC1A speakers and it was truly stunning. Have look at some of the mods in the “Philips Cd 104 tweaks” thread.

Cheers,
Richard

 

[flavio81]

Thanks Richard, that's exactly what i wanted: Opinions!!

My friend is running a TDA1541A DAC with a choke loaded 6C45P and Duelund capacitor output and he reckons it’s on par with his Technics Sp10 Denon DL 103 DIY Schroder arm combination, but he may have been drunk

Well, the DL 103 is a very controversial pickup... the stylus is spherical so it will give lots, LOTS of even-order harmonics and noticeaable sibilance/distortion on high level passages... hardly an "accurate-sounding" pickup by any standard!!

Seems your friend likes the sound of groove-mistracking distortion.

Ok, now back on topic. Please tell us how it goes when you try the Studer output on your CD player.

 

[Richard]

Hi Flavio,

I started with a Philips Cd 202 which is very similar to your Cd 63 and then upgraded to a 304. I recall a member on this site by the name of TDA1540, say that the second generation players that used the 1540’s were better due to the layout and goundplane etc. After doing similar mods on both players and comparing I soon realised they were right, so concentrated my efforts on the Cd304/104. I also played around with the power supply of the valve output and am using a valve rectified power supply.

I think most TDA1540 fans lust after the LHH2000, but instead of trying to get our players to sound like it, why not strive to surpass it?

Have you had a read of the following:
http://www.audionotekits.com/agrovedac.html

They discuss the use of transformers as filters, it’s working well for me. As for the Studer output I’m afraid it’s going to be a long time before I hook it up and compare, but when I do I'll post the results.

Cheers,

Richard

 

[flavio81]

I think most TDA1540 fans lust after the LHH2000, but instead of trying to get our players to sound like it, why not strive to surpass it?

Have you had a read of the following:
http://www.audionotekits.com/agrovedac.html

They discuss the use of transformers as filters, it’s working well for me.

Hi Richard,

The circuit describes I/V conversion done with a transformer, and tells some reasons why no-oversampling should sound better. But it does not speak of any filtering on the output stage. As far as i understood, the transformer is there for I/V conversion not filtering.

In any case, all the NOS (non oversampling) DIY designs i've found so far apply just very mild filtering at the output. No brickwall filtering.

Someday i will try the NOS modification for this player. It's reported to sound much better. Of course this does not mean "more accurate" but "better sounding."

 

[Richard]

Hi Flavio,

The transformer can also be used to filtering high frequency content. The UTC transformer I’m using rolls off just after 20kHz. A resistor on the transformers primary also helps adjust the I/V resistor. TDA1540 in NOS is very nice, but my findings suggest there may be some loss in micro detail.

Here are some links to the sound of the Revox Studer TDA154x players. I’m not sure if they are using the transformer output. Lukasz of Lampizator fame has found that taking the output from the 1st op amp to sound better, nothing is said about the output board with transformers. I’m waiting for him to add the valve output for this player and see his findings.

http://www.lampizator.eu/lampizator/REFERENCES/Revox B225/RevoxB225.html
http://www.tnt-audio.com/sorgenti/revoxb226_e.html

Cheers,

Richard

 

[lcsaszar]

I have seen on the Lampizator site some internal photos of the Revox, and it seems it uses the ceramic version DAC chip, the TDA1540D, not the plastic TDA1540P. I suppose this also makes some performance difference, just like at opamps.

 

[poynton]

I have a B226S and a B126.

Neither uses a transformer in the output.

The professional Studer version uses a transformer to convert the output to balanced.



Andy

 

[flavio81]

I have seen on the Lampizator site some internal photos of the Revox, and it seems it uses the ceramic version DAC chip, the TDA1540D, not the plastic TDA1540P. I suppose this also makes some performance difference, just like at opamps.

I just disassembled my Marantz CD-63 (1982, predates the CD100) yesterday, to replace some power capacitors...

... it uses the ceramic versions of the TDA1540 too;
... all the resistors are metal film...
... almost all the non-polar capacitors are NOT ceramic but better quality ones (the ones around the DAC were BIG!!!)
... almost all the ICs have a Marantz code instead of the normal Philips code (ex: "M45xx" instead of "SAA7030")

I also replaced the 22uF capacitor that's in series before the final output, for a 22uF non-polar electrolytic, just in case. BTW the old philips blue axial capacitor still measured around 22uF. I did not notice a significant sound improvement.

This machine is really BUILT LIKE A TANK!!... the frame is 100% bulky, heavy metal, the CDM mechanism is metal too, the outer shell is actual aluminium, all the circuit boards are shielded with big grid-like metal shields... a very heavy player.

 

[flavio81]

Hi Flavio,

The transformer can also be used to filtering high frequency content. The UTC transformer I’m using rolls off just after 20kHz. A resistor on the transformers primary also helps adjust the I/V resistor. TDA1540 in NOS is very nice, but my findings suggest there may be some loss in micro detail.

Thanks for the info. So you like the NOS mod for the TDA1540 or you would rather not use it?

 

[guido]

... almost all the ICs have a Marantz code instead of the normal Philips code (ex: "M45xx" instead of "SAA7030")

I don't think Mxxxx is marantz. My philips cd650 contained an M4804 instead of a SAA7210. Player made by philips belgium. IC factory naming convention is also three characters, like all the other philips stuff (LHG in this case).

Guido