Marantz CD-67 measurements (newbie alert!)

[ReDFoX]

Hi everybody! @NTTY's recent CD players measurements sparkled an interest in me, so I tried to be trendy and measure my disc playing devices.

Attention: I'm a noob at measuring things (at least now) + English isn't my native language (as you can tell by location), so, sorry in advance)


Well, I forgot to take any fancy pictures of my unit (silly fox moment), so I dug up an ancient photo of my system (yes, I was using a VHS deck for audio recording). The player itself is sexy (although I prefer NAD industrial design) and I would lick every millimeter of that faceplate

Unfortunately, a few years ago, this CD-67 developed a ring crack (is this the correct term?) on RCA out, so I took it apart and resoldered terminals. Here are some photos during process:

Let's begin our quick dive! I'm well-aware of this thread, it was created just hours after I finished with my test CD...
All signals were generated with REW.
CD-67 was fed into line input of my E-MU1616 at 192kHz sampling rate (E-MU has ultrasonic noise, so I wouldn't show anything above 48k).
Analysis settings, unless said otherwise:

• 32k FFT
• 8 averages
• Dolph-Chebyshev 200 window
• 0% overlap

Starting with single tone THD measurement at 1kHz:

Yeah, that looks kinda nice, especially for that price and age. CD-67 easily clears 16bit threshold!
Next up is multitone (FFT length is 512K):

Again, no signs of any kinds of excessive distortion and mains leakage.
A start of a measurement sweep I burnt onto CD got corrupted somehow (I should've added a silence before it), so I used 100Hz+ portion. Still, it shows surprisingly flat FR! My dad and I were convinced that this 67 model has some kind of soft sound and I expected a slow roll-off filter with a dropoff at HF, but no, this thing's flat! That's crazy, since I literally heard this sucker play softer and richer than my Pioneer DVD...

But really, what about stop band length, antialiasing filter attenuation and out of band performance? 255 averages were used in this test

Ehh, attenuation is mediocre, to say politely, but I don't really care about that

Well, yeah, it's unimpressive, but expected. Not a concern, IMO...
Jumping onto IMD distortion tests!
CCIF shows, again, a deficiency in out of band atenuation

SMPTE shows quite a clear spectrum (256k FFT):

Jitter performance is also nice, well below audibility (256k FFT again):

Single tone THD sweep is showing some rise in distortion at HF, but that's mediocre filter's fault, I suppose

Here are manufacturer's specs:

I got 0.0006% or 0.003% THD, if we use THD+N. Well, I was prepared to see something horrible, but this thing delivers! This isn't some top of the line model, but it's audibly transparent (as far as I can tell) and costs ~100usd, here, in Russia.

 

[Blumlein 88]

Very nicely done. Cleanly laid out making it easy to understand.

I also way back when used a VHS HiFi unit to record music. The specs were pretty good. I made these 2hr long mix tapes for background music.

 

[ReDFoX]

Very nicely done. Cleanly laid out making it easy to understand.

I also way back when used a VHS HiFi unit to record music. The specs were pretty good. I made these 2hr long mix tapes for background music.

Thank you!
I was ready to save up and buy professional Panasonic model (no TBC) with manual volume pots, but at that time I learned about head switching and usage of generator to fill in gaps... That ruined a format for me, cause my child brain thought, that I'd get a true analog experience without throwing a ton of money on R2R deck. I lacked the knowledge at the time and was brainwashed by audiophile articles, praising analog origin as an ultimate quality of sound. Recently, I repaired that HD750 and maybe, I'll give it a chance but with null test this time

 

[AnalogSteph]

CD-67 was fed into line input of my E-MU1616 at 192kHz sampling rate (E-MU has ultrasonic noise, so I wouldn't show anything above 48k).

That's the plain, non-M 1616 (PCI or Cardbus), correct? ADCs PCM1804, DACs CS4392? There is probably modding potential in the thing, looks like PCM4202 may be largely compatible (you would just get about 1.58 dB lower digital levels as the newer model ADC accepts a somewhat higher input level). It wouldn't do much about the out-of-band noise above 50 kHz (if anything, there is even more up there), but in-band noise and distortion seem to be about 6 dB lower each. There is probably potential in the analog circuitry as well but I'm not sure whether anyone has reverse-engineered it in the 1616 or 1820 yet. Finally, yours may already have seen some recapping action as the electrolytics used at the time seem to have been quite cheap, but it doesn't hurt to check.

BTW, if you want to improve this player, mod the output stage to CD-57 level by removing RH23 and RH24 and equipping R619 and R620 instead. The HDAM section contains some nice (genuine) JFETs that may still come in handy for something else. As low as distortion already is and with the other low-level stuff floating around, you may have difficulty seeing the effect though. No idea how to tackle the so-so jitter performance, I bet it's some sort of layout/grounding related issue. I can already see that modern 100n SMD ceramics would allow for DVDD rail decoupling much close to the DAC (pins 15-16), and while trickier since you need to jump over an output pin, something in parallel to CD12/13 but closer should also be feasible.

 

[ReDFoX]

That's the plain, non-M 1616 (PCI or Cardbus), correct? ADCs PCM1804, DACs CS4392? There is probably modding potential in the thing, looks like PCM4202 may be largely compatible (you would just get about 1.58 dB lower digital levels as the newer model ADC accepts a somewhat higher input level). It wouldn't do much about the out-of-band noise above 50 kHz (if anything, there is even more up there), but in-band noise and distortion seem to be about 6 dB lower each. There is probably potential in the analog circuitry as well but I'm not sure whether anyone has reverse-engineered it in the 1616 or 1820 yet. Finally, yours may already have seen some recapping action as the electrolytics used at the time seem to have been quite cheap, but it doesn't hurt to check.

I'm well aware, that E-MU docks aren't as competent as they were 5-7 years ago. I picked one up just because it was packed with everything I can imagine: more-than-enough HP amp, hardware DSP (RME, unless it's Hammerfall, is well above my budget) and phono input (I just threw dock into another room and recorded on my main pc). Mine 1616 wasn't used much by the previous owner (and I shut down dock by disconnecting power every time), so caps are ok, at least for now.

BTW, if you want to improve this player, mod the output stage to CD-57 level by removing RH23 and RH24 and equipping R619 and R620 instead. The HDAM section contains some nice (genuine) JFETs that may still come in handy for something else. As low as distortion already is and with the other low-level stuff floating around, you may have difficulty seeing the effect though. No idea how to tackle the so-so jitter performance, I bet it's some sort of layout/grounding related issue. I can already see that modern 100n SMD ceramics would allow for DVDD rail decoupling much close to the DAC (pins 15-16), and while trickier since you need to jump over an output pin, something in parallel to CD12/13 but closer should also be feasible.

If it was my player, I might be looking into that, but I'm 99% sure, my dad won't let me fiddle with his unit . I searched for info about CD-67 when I was a child and literally everybody in russian forums was shitting on it, sayin' it's muddy, and you must tweak it to get at least sounding fine.

 

[NTTY]

These are relatively good results!
With the 1kHz tone, increase the FFT to at least double the sampling rate of the capture (eg: 256k for 92kHz).
Use Blackman Harris 7 windowing and 32 averages. It will give a better picture of power supply related noise
I would recommend to use 20Hz to 20kHz as the span (distortion panel of RTA) rather than 22kHz.
This player is likely to sound as transparent as many others.
Thanks for the review!

 

[ReDFoX]

These are relatively good results!
With the 1kHz tone, increase the FFT to at least double the sampling rate of the capture (eg: 256k for 92kHz).
Use Blackman Harris 7 windowing and 32 averages. It will give a better picture of power supply related noise

Thank you!
Honestly, I tried to mimic Amir's settings for most parts to keep measurements at least visually comparable with his. AP uses modified Dolph-Chebyshev window, so I used that.

I would recommend to use 20Hz to 20kHz as the span (distortion panel of RTA) rather than 22kHz.

That's a valid point, 20Hz-20kHz seems more appropriate for me too