Review and Measurements of Technics SL-PG100 vintage CD player

[pma]

I am still using a Sony ES CD player from the same era, it still works, still sounds great and the build quality urinate all over most modern gear.

Great! And I agree, when I see photos of those miracle $50 boxes here ...... they are just - cheap, and mass production and circuit technology progress gives them good parameters, if their design is competent. But it is a pressure from customers for cheapest everything that is responsible for low built quality and low reliability.

 

[pma]

The digitally derived pure test signals are not intended to be dithered.

You may be confusing emphasis with dithering.

Following my comments on undithered tones not good for distortion measurements, I am posting measurements on Technics SL-PG100 at 1kHz/0dBFS with dithered and undithered CD test tones. With better players, the difference in digital rubbish is even bigger.

With dithered sine, noise bottom should be clean, we can see some DAC created digital rubbish.

When using undithered sine only, it is impossible to tell what is quantization noise and what is DAC digital rubbish.

 

[restorer-john]

I cannot agree with you on using of undithered tracks for distortion measurements. You get zillion of distortion-like and digital hash like spectral lines then. Same mistake was done even by AES recommendations and test files decades ago. You may use -90dB undithered sine to test lowest bites, that's all. Badly dithered recordings were also the key to horrible sound of some of earliest digital music. Plots of undithered and dithered 44.1/16 sine follows.

Following my comments on undithered tones not good for distortion measurements, I am posting measurements on Technics SL-PG100 at 1kHz/0dBFS with dithered and undithered CD test tones. With better players, the difference in digital rubbish is even bigger.

With respect, the digitally generated frequencies chosen on the test disc listed above (notice they are not even numbers) have been chosen so the samples fall on theoretical 16bit whole values and consequently do not need dithering. I could be wrong, but it has served me well over the years for my THD testing on players. The Sony YEDs has more THD on its 1000Hz track, than the Denon with 1001Hz at the higher harmonics.

When using undithered sine only, it is impossible to tell what is quantization noise and what is DAC digital rubbish.

However, even with your plots, the final numbers are virtually identical in this case are they not? I'll pull up some numbers tomorrow using ARTA and the same fft size, window etc and see what I come up with. I could send SPDIF straight into the M-Audio 24/192 card to analzye the actual discs if you want? This is an interesting exercise- the best CD player I have here is under 0.0015% at 0dbFS and I have a Marantz CD12/DA12LE which is also the same THD spec. Previously, the lowest I could resolve was 0.002% with reliability, so I called them good.

Cheers.

 

[pma]

So guys if you needed 44.1/16 sine 1kHz/0dBFS test files for your CD players, here you are with a download link for both dithered and undithered. We can prepare a complete set of files, in future, if members are interested.

http://pmacura.cz/1k_sines_cd_format.zip

 

[pma]

However, even with your plots, the final numbers are virtually identical in this case are they not? .

I do not care so much that final numbers are similar. That's the problem of evaluation simplified to numbers only. The case and biggest issue of undithered signals and music is that they create signal correlated quantization noise which is strongly level and frequency dependent and in some cases is extremely audible. 16bits is simply not enough to be truncated only. As you know, proper dithering may make even 8-bit files well listenable. This was explored and described in detail in the nineties and many links are easily searchable.

 

[pma]

We can see that even the distortion "numbers" are not same for dithered and undithered test signals, because of strong correlation of quantization errors to level and Fsampling/Fsignal (error lines may overlap with harmonic frequencies), as shown for the -90dBFS 1kHz below:

44.1kHz/16bit -90dBFS 1kHz signal dithered

44.1kHz/16bit -90dBFS 1kHz signal undithered

 

[pma]

The review just updated, two measurements are added, DIM (square 3.18kHz + sine 15kHz) and Multitone, in the 1st post of the thread. Personally, I find the DIM sine+square test more revealing than the Multitone test.

 

[Berwhale]

Unfortunately, people tend to rotate discs in the cases and concentric light scratches become the problem, not radial ones. Once the linear burst error (2.4mm) is exceeded, linear interpolation kicks in which is fine for music, not for test tones. Label side Al layer damage is rarely concentric.

Did anyone think of putting a notch in the center hole of the CD with a corresponding key in the case so that the disk couldn't rotate in storage? I guess you'd have to put symmetrical notches to balance the disk whilst it rotates in the player...

 

[Jack richer]

I am reading a CD disc. I have several "pro" CD tests discs and a number of those I prepared myself. The reason is that even at the "pro" discs the test signals are often not properly dithered and this is a real issue with 16-bits. I always check the discs in CoolEdit Pro before I use them. So the "hash" seen here really comes from the player. It is audible only on very low level signals (-60dB, -80dB) after you amplify the output above standard listening level.

One remark to the 50Hz seen at -95dBV in every plot. This also comes from the player. When the output is muted (to GND), the 50Hz line disappears and also the noise floor is clean then.

Also please note that my plots are calibrated in dBV and NOT in dBr or dBFS. This is for the reason to read easily real amplitude of the spectral lines. So always please read dBV for the base frequency or just read "RMS" value which is shown in every plot and which is the true rms signal value.

Very nice write-up. Thanks for doing this.

 

[Blink Too Fast]

I enjoyed reading this. Thanks for the write-up.

I converted my late 80s / early 90s Technics SL-PC705 and SL-PC364 to S/PDIF out, and love using them.

Link to Technics S/PDIF conversion

 

[Ulrik de Renouard]

I enjoyed reading this. Thanks for the write-up.

I converted my late 80s / early 90s Technics SL-PC705 and SL-PC364 to S/PDIF out, and love using them.

Link to Technics S/PDIF conversion

View attachment 330778

Found a 100a in a bin and because of the front was hanging loose I could see it's cd drive and was very pleased to see the Philips CDM4. After recapping and some repair to the front it works fine. Lampizator have written a lot about the Philips drives and I don't use/want other ones.
The 100a is an upgrade to the 100 because of the first version had an inferior cd drive and got bad rewiews.
Luckily this player uses the same pcb as the bigger ones and it is easy to add a toslink connector. You need to make a hole in the backplate but thats not difficult.
I connect it to my Bifrost dac and the combo works great.

 

[Adaboy4z]

Old thread but thought I say my Technics SL-P555(1989) is still going strong.

 

[Blink Too Fast]

Old thread but thought I say my Technics SL-P555(1989) is still going strong.

It sent me a private message and asked me to ask you to give it an SPDIF out.

 

[morillon]

The review just updated, two measurements are added, DIM (square 3.18kHz + sine 15kHz) and Multitone, in the 1st post of the thread. Personally, I find the DIM sine+square test more revealing than the Multitone test.

View attachment 39700View attachment 39696

any details on this “dim”? in its realization and also its mode of reading that you make of it?
thanks
;-)

 

[Sokel]

any details on this “dim”? in its realization and also its mode of reading that you make of it?
thanks
;-)

It's included in Multitone for a long time now after @pma 's suggestion (two signals,DIM30 and DIM100 )



looks ok in this one but in other newer DACs it does not looks nice.

 

[pma]

any details on this “dim”? in its realization and also its mode of reading that you make of it?
thanks
;-)

DIM = dynamic intermodulation distortion test. 3.18kHz square + 15kHz sine, 4:1 amplitude. You may create it yourself or use the test signal in Arta (multitones). You should see only 3.18kHz + its 3rd, 5th harmonics and 15kHz line. Anything else are intermodulations or mirror images.

 

[morillon]

sorry...
in fact I had come across him on M-T without having realized his interest (nor his principle)... (I was recently looking more towards the NID side)
the stress factor is obvious with this DIM indeed
Do you use it especially in what circumstances and with precautions? (square 3k15)

 

[pma]

It is an old TIM test (Otala, Curl, Leinonen)

AES E-Library » A Method for Measuring Transient Intermodulation Distortion

A description of the square-sine intermodulation measuring method is given. The practical measuring procedure is described, and measurement results of commercial audio power amplifiers and integrated operational amplifiers are presented, in many cases showing appreciable transient...

www.aes.org

maybe obsolete now, but sometimes it discloses non-linearity easier than other signals. It should look like this in 44.1/16 CD format. All the spectrum lines except for those are a mess.

You may compare this with the result from SL-PG100.

But, there is a pretty good result from DM+ driven from iPad: