Magnavox CDB-460 Measurements and Review (CD Player)

[tpd]

This post provides detailed measurements of a Magnavox CDB-460 CD player.

Released in 1986, the CDB-460 was the littlest brother of the Magnavox CDB-650, the US version of the Philips CD650. This line of CD players is based on the TDA1541 and SAA7220P/A chip combination and CDM-2/10 transport. There was quite a bit of contemporary discussion and modding around these players. You can read Walt Jung's thoughts in the June 1987 issue of Audio and his POOGE-4 articles (among others) in The Audio Amateur magazine.

This particular CDB-460 is one of two I have restored and is a steady presence in my main listening system. This one was in made in Taiwan, not Belgium, in 1987. The CDB-460 also has a coax digital out and can be used as a transport. I did not measure digital output in this review.

It was non-functional when I acquired it. It's been cleaned up and re-capped, but retains the original LM833N filter op-amps. Other than substituting poly caps for some small-value electrolytics, it has not been modified.

Measurements (analog output from CD)

I made measurements using the test CD files version 3.7 kindly shared by @NTTY. I am trying to follow Flo's protocol for comparability.

RCA output is:

Right: 2.002Vrms
Left: 2.041Vrms

NOTE: The following measurements were not captured correctly. Please see updated measurements in this post: https://www.audiosciencereview.com/...ments-and-review-cd-player.58110/post-2136510

1kHz sine @0dBFS (dithered)

I also tested 997Hz @ 0dBFS to see if there's a significant difference. Only the right channel is shown here.

1kHz @ -6dBFS:

Bandwidth measurements:

Multitone test:

Jitter test (16bits/44.1kHz) with the source overlaid:

THD (excluding noise) vs Frequency at @-12dBFS:

And there we have it. Assuming I've my measurements correctly, these seem like decent results for a 37-year old CD player. I certainly enjoy listening to it.

This is an old, budget CD player and has its quirks (as do I). There is no remote for this model. Front panel controls can be confusing -- seek controls are immediately below the play/stop controls, but track forward-back controls are under the main display. But the tray opens and closes solidly, TOC read happens quickly on commercial CDs, and navigating between tracks is quick.

CD-R playback is finicky and depends on the media and burner. I'm finding that my older burned CDs work fine, but recently burned CDs may or may not play. I suspect it may be due in part to the very poor quality of currently available blank media.

I hope to later post measurements of my second CDB-460, which lives in my headphone system, for comparison. I also have a CDB-650 that needs some restoration. When it's feeling up to reading a CD-R, I will do some "before" measurements, then follow up with "after" measurements post-refurbishing.

Cheers

 

[Watire]

Great post. I would hold on to it. The frequency response is everything but aggressive in the high end.
Is changing all the capacitors was all that was needed to get it playing again ?

 

[restorer-john]

What's going on with the high end? What ADC/settings are you using?

You should be within 0.5dB at 20kHz.

 

[NTTY]

Nice review!

Did you calibrate the Focusrite for the bandwidth measurement?
Follow that video in case not:
Create REW Calibration file

Calibration is not mandatory and not perfect because it will include DAC compensation of the FocusRite (while we only need to compensate for the ADC), but sometimes it’s better than none.

Do the calibration for 44.1kHz sampling rate input only, because REW forces that sampling rate for the bandwidth measurement. For the other measurements, it’s not required.

Alternatively, use the white periodic noise and capture 300+ averages with the Motu set at 192kHz sampling rate input. Then zoom 20Hz to 20kHz and you’ll get the correct bandwidth.

Otherwise, I’d recommend increasing the input gain of the Focusrite so that for a 0dBFS tone played by the CD, you see like -1dBFS at the input. Else it makes some calculations wrong in the main dashboard (eg: ENOB, N+D, N).
But you will need to lower the input for the intersample-over test tones as they would saturate the input. There’s no best option here. Maybe @JohnPM can find a way to address this need of having few dB headroom at the input for us to measure, yet ensuring that the dashboard takes this required headroom into account?

 

[tpd]

While I have calibrated the Focusrite, it's possible the calibration was not applied while capturing some of the measurements. I will take another look. This is my first set of measurements so it's okay to assume user error if something looks off.

 

[tpd]

Great post. I would hold on to it. The frequency response is everything but aggressive in the high end.
Is changing all the capacitors was all that was needed to get it playing again ?

There are some power supply capacitors known to go off on, particularly on the driver board. The symptom is inability to detect the TOC. Unless there's a good reason to do otherwise, I often take the shotgun approach of replacing all electrolytics in any gear 30+ years old. They are all well past their rated lifetimes and post-replacement measurements frequently show elevated ESR.

Unlike Sony-based CD mechanisms, the Philips CDM drives seem to be pretty robust and often just need cleaning and a new belt. If they don't read, it's usually because the power supply is wonky. At least that's my experience across 5 Philips/Magnavox players so far.

 

[Apesbrain]

Nice. I remember this unit. I had an "Emminent Technology" CDB-650 which I recall sounding very good.

But who knows. This was back in the day when I religiously demagnetized and black-lit CDs.

 

[tpd]

I re-ran calibration and did indeed come up with some slightly different results. Most different for the bandwidth measurement. This is curious enough that I'm going to push up measuring my second CDB-460 to see whether there's consistency between them.

And here's the 999Hz @ 0dBFS measurement revised.

 

[AnalogSteph]

I re-ran calibration and did indeed come up with some slightly different results. Most different for the bandwidth measurement. This is curious enough that I'm going to push up measuring my second CDB-460 to see whether there's consistency between them.

That doesn't look realistic (like, at all). Which interface at which sample rate are you using? I would recommend using one far beyond 44.1k in order to minimize the role of its digital filters. Not sure whether REW can handle this, I've had to (record and) resample externally when using RMAA.

 

[tpd]

@AnalogSteph to be clear, you're talking about the bandwidth measurement specifically, or all of the measurements?

I'll have to confer further with @NTTY about my process. I must be overlooking something. Thanks for the feedback, however. I would like to get this sorted.

 

[Ruhled]

The budget siblings i.e. the non-oversampling cdb-470 has a flatter response than this and only drops maybe 4-6db by 20khz iirc...and we know it's broken from the factory. Something is indeed probably off here. The 470 is easily my worst measuring CD player but it isn't 20db down at 20khz.


*** Edit sorry I misread the graph! 6db down seems par for the course on these.

 

[AnalogSteph]

@AnalogSteph to be clear, you're talking about the bandwidth measurement specifically, or all of the measurements?

The bandwidth. No CD player has a filter response that looks like this with a jump up at the end, that's undoubtedly a calibration artifact.

 

[NTTY]

I don’t know why you’re getting this result.

Remove the cal file completely. Use the white noise periodic (last file of the test CD) and capture 300+ averages with the Focusrite set at 192kHz sampling rate input. Save the measurement and send me the file, I’ll have a look.
.

 

[tpd]

Yeah I have something jacked up here. @NTTY I will follow up this evening. Thanks.

 

[Timmy]

It's been cleaned up

It shows. Nice pictures, Al!

 

[NTTY]

The budget siblings i.e. the non-oversampling cdb-470 has a flatter response than this and only drops maybe 4-6db by 20khz iirc...and we know it's broken from the factory. Something is indeed probably off here. The 470 is easily my worst measuring CD player but it isn't 20db down at 20khz.


*** Edit sorry I misread the graph! 6db down seems par for the course on these.

Yep, non oversampling CD player would show -3.92dB at 20kHz because of the sin(x)/x roll-off, if not compensated.

 

[tpd]

To follow up: I'm unable to capture sweeps from CD media reliably with my setup from any CD player. I'm seeing the same rolled off response from every player I've tried to test (other Magnavox players, a Pioneer, and a Yamaha). So there's clearly something jacked up in how I'm doing this measurement. I'll report back in the future with better measurements once I make sure my testing devices are in fact set up correctly. Thanks for your patience and advice.

 

[mcdonalk]

I normally prefer to eschew nostalgia, so excuse my uncharacteristic shameless reminiscing about how I came to possess a Magnavox CDB-650 CD player, which I still have. You may want to skip this; there's no science, and it is not about the CDB-460.

Back in 1987, I had just read the rhapsodic review for the CDB-650 in the March issue of Audio magazine (I think I've attached an image of the cover below), to which I subscribed. At that time, I didn't care for our NAD CD player, and my wife really hated it. (She appreciated good sound as much as I did; we always did our audio shopping together. In fact, we built all our Heathkit audio components together, but that's another story.) So, I decided I should get me one of those CDB-650's.

This was pre-internet, so I had to locate dealers for this CD player via ads in the back of audio type magazines. I quickly discovered, after many fruitless long distance landline calls, that the CDB-650 was THE HOT CD player. Everyone from anywhere in the US I contacted was sold out. Crutchfield was sold out. I contacted Magnavox for information on local dealers, and the only local dealer was a furniture chain store. They were sold out.

At that time, I worked at a large GTE facility where central office telephony switches were designed. There was a small 1-room company store adjacent to the cafeteria where Sylvania light bulbs and Magnavox clock radios and the such were sold. Not really thinking anything, I stopped in and perused their catalog. There, in the catalog, was a listing for the CDB-650! There was some kind of affiliation among GTE, Sylvania, Magnavox, and Phillips at that time. The clerk ordered it for me, and a few days later, it arrived, and I paid for it via paycheck witholding over a period of weeks.

The CDB-650 had a feature that I never saw anywhere else back then: Favorite Track Selection (FTS). Using a small foldout numeric keypad, one could select the tracks for a CD to be played, and the unit would remember this programming. This allowed me, for example, to automatically skip the annoying narration tracks that preceded each movement of Vaughan Williams' Antarctica (#7) Symphony, every time I played it.

I still have the CDB-650, and it still works. But these days, I rip all CD's to NAS and stream them via jRiver Media Center, so I have not used the CDB-650 in at least 10 years.

 

[NTTY]

To follow up: I'm unable to capture sweeps from CD media reliably with my setup from any CD player. I'm seeing the same rolled off response from every player I've tried to test (other Magnavox players, a Pioneer, and a Yamaha). So there's clearly something jacked up in how I'm doing this measurement. I'll report back in the future with better measurements once I make sure my testing devices are in fact set up correctly. Thanks for your patience and advice.

Remove any .cal file from REW.
Use the periodic white noise file, with input at 24bits/192kHz. Run more than 400 averages measurements and save for left and right channels.
Then, in overlay view, add 1/48 smoothing to both traces. You’ll get something like this:

This will be very much independent from the FR of the Focusrite which is flat from 20Hz to 20kHz at that sampling rate.

 

[tpd]

Thank you @NTTY I'll give this a try ASAP. Much appreciate your help on getting my measurements sorted.