My CMT-PX7

[sfbp]

CMT-PX7, mythical machine

This bookshelf machine has the perfect specs. It's Type-R, MDLP (CXD 2662) like its littler brothers PX5 and PX3, and can do the PCLK interface.

 I could never see any on sale until one day, almost by accident I found not one but two on Yahoo jp listed under HMC-PX7. Now you see why i never found it.

https://buyee.jp/item/yahoo/auction/x1078822831#group=nogroup&photo=1

The CMT-PX3 was only ever called that in Japan. Elsewhere it became known as the DHC-MD595. Not sure about the CMT-PX5,

https://www.minidisc.wiki/equipment/sony/bookshelf/cmt-px5

 which has 1 MD and 3 CDs.

Unlike the PX3, the PX7 does not support a tape unit via the I2C cable controlled interface, at least as far as I know. But it has 3CDs and 3MDs which is a pretty stunning combination of equipment. Of course it can do Optical and Analogue in, and there's a radio though it's the Japan frequencies (wish I could figure out the way to substitute the tuner block, but not that important, owning this is much more about recording long programs easily off the Internet).

So I ordered one (picked at random from the 2) for $35 buy-it-now. Panic stations - the shipping was initially going to be $500, ouch. So realised that if I got Buyee (my proxy bidder) to ditch the speakers, things would be a lot better. The package duly arrived about 3 weeks after I first bid, shipping cost around $120 (US). Buyee knows I won't pay $500 for a $50 item, so they try hard.

First problem was interface. It came with no connecting cable. So with suggestions from our Kevin here, patched together something that would work.

Of course, turned it on, lights came on, but not much happened. No response to button presses other than occasional light changes from orange to green, seemingly at random (I now know they weren't).

Forced the console on the master unit open but it clearly isn't happy. More on this later.

Open up the satellite unit and discover that there is "belt jam" everywhere. The belts had completely disintegrated in a sticky mush that was jamming everything. Isopropanol works wonders but still awfully messy, all over everything especially my fingers.

Ordered a bunch of belts (circular cross section) from some place in Germany.

https://www.ebay.com/itm/324114190779

Sadly the "belt-jam" type belts are what Charlie (linx audio) offers, and not in the right sizes anyway.

And every time I tried to get the MD stocker working, one or both of the following noises:

a. a ticking sound for about 20 seconds followed by silence (when machine first turned on)
b. a loud cracking sound that sounded really, really bad (when I pushed eject/load button)

Problem now arises - no Service Manual. In the course of opening up the unit and attempting to fix it, I think I completely ruined the mechanism of the stocker for the MD. I managed to get the CD stocker working because, by a wonderful accident, CHC-CL5MD

https://minidisc.org/part_Sony_CHC-CL5MD.html

uses the same CD stocker. (sorry I have no idea why this image is upside down)

This helped enormously, and when I had given up on my primary objective (fixing the MD section) I got the CD working. I immediately noticed that with a couple of Sony speakers I already had, the sound was absolutely wonderful on CD playback. (I just saw promo material for a motherboard with its own sound, claiming the superiority of Japanese capacitors, so in this machine that sort of thing may be part of what the original $1500 price tag was all about).

Get the console opening nicely after I clean off ALL the belt jam (it uses a belt to open and close) and also a bit of bicycle oil plus manually tweaking until there's little or no resistance as the motor turns. New belt, of course.

https://forums.sonyinsider.com/uploads/monthly_2023_09/20230911_153440.mp4.2cc0dc383db12f86ec249d8fb2d608f4.mp4

(sorry about the movie, still puzzling how to get it to load in most browsers, so here is a download link for you)

Anyway I got the MD open (by the wrong method), cleaned it up, discovered there is a broken piece of plastic that transmits (mechanically) the motor motion into something which opens the MD door at the front of the machine.

 I glued that back together (by this time I am getting slightly better at assembling the MD stocker) whereupon it broke again. So I got a friend to make (3D print) a replacement. Same thing, only worse because the plastic is less rigid and therefore has more trouble initiating the door opening. I finally discovered the correct way to open things up and reassemble, although I fear too late for this one which is well and truly gibbled. Back to the original (reglued twice now) piece. Still jamming.

Here's a picture of the guts of the MD stocker in pieces. The two small pulleys (minus their belts) are the two motors which drive the opening and vertical positioning.

Quite some time has gone by, and I am off on my travels and finally decide it's time to pull the trigger on the second machine which is still there on yahoo.jp. Maybe it will be better (so it is) and specifically it has an interconnect cable which is worth the cost of the item ($5.00) several times over. Same problem, another $120 to ship it, minus speakers, but maybe that's ok.

This one duly arrives. It's quite rusty especially around the feet (pictures) that the units (both main and satellite) stand on, and nothing works, but this time no belt jam to be seen. Just broken belts.

The second machine's MD stocker has one disk in the MDM-7A, and two in the (!!)5(!!) slots for disks. This looks quite encouraging. Part of my problem is I don't know which 3 slots are actually used, lacking the service manual. So after clean up and belt replacement I get it to the point where (it looks like) it's working. Minor problem replacing the belt on the MDM-7A itself (the one on the first machine), I bust the plastic claw. But did that once before on another machine. Half an hour later with some glue that's back to normal. I've got the other MDM-7A, worst case, but this one seems to be ok.

Finally get one disk in, but the ejection is still iffy. Break the (same/similar) plastic piece which is very sad. What did I miss? Something subtle about the order of assembly, I bet. Oh yes, somewhere along the line I discovered this 19-pin interconnect cable. Machine goes from almost working to perfect.

After a lot of fiddling and regluing (2x more) I finally get the stocker working. I almost had a heart attack when music played for the first time. Occasionally the disk won't eject but if I simply wait a few seconds, turn the machine off and on again, everything rights itself. No more crunching noises.

This machine is neat. You can do continuous recording (though of course you miss a little bit of time when it switches from one MD to the next), or you can tell it to sync-record.

 In the latter case it always makes one MD per CD, slightly annoying as even in LP modes it switches disks when the CD changes. But you can take charge of all the modes and settings from M-Crew, so the display which is a little hard to read sometimes no longer matters. Interestingly, there are TWO PS/2 sockets (one at front and one at back), one labelled Keyboard/PC-Link and one labelled just PC-Link. That is unique, in my experience. They were very serious about loading this model down with features.

An oddity with track metadata: I can play an MD which I transcribed from CD, and on the PX7 front panel the title information shows up. But you can't see the titles in M-Crew. The converse is not true. Once M-Crew has the data (eg. if I type it in by hand) then the metadata always shows up on the dashboard of the machine (as well as M-Crew) as recorded on the MD. Still wrestling with that. One possibly related annoyance, M-Crew is always switching out of Group Mode no matter how often you set it.

Back in the day this might have worked with Gracenote, but anyway that's all gone now - and even then I cannot access the CDDB button suggesting it was maybe never possible.

The sound is amazing; the capabilities are amazing. I'm just waiting for something to break. I've even got spares for most parts, and my friend has made another plastic piece (or 3) of a harder material, just in case.

Stephen

PS look what I just found, idly poking around

https://buyee.jp/item/yahoo/auction/c1106919218?conversionType=search_suggest

 

 

[BearBoy]

You've been on a real mission with this one! I really admire your persistence, and ingenuity, in getting it working. Sounds like your efforts have been rewarded with the audio quality!

 

Stylish looking unit as well.

 

How did your friend 3D print the replacement part for you? Did he scan the original broken part somehow or was it a case of measuring everything and drawing it out in a CAD program?

 

18 hours ago, sfbp said:

This machine is neat. You can do continuous recording (though of course you miss a little bit of time when it switches from one MD to the next), or you can tell it to sync-record.

 In the latter case it always makes one MD per CD, slightly annoying as even in LP modes it switches disks when the CD changes.

 

That seems like a bit of a missed opportunity on Sony's part. One of the big selling points of MDLP was that you could fit multiple albums onto a single MD and yet the machine does not really exploit that.

[sfbp]

I think he worked from the picture I sent him. Nope, I gave him the busted bits.

 

The problem with changing disks might be operator error........

[sfbp]

Looked a little bit more at this.

 

Trying to put 3 CDs onto multiple MDs at a single button press is a bit of a boondoggle. Every program step ends up being coded. Sooner or later you run into "too many program steps" (probably some internal structures in either firmware/chip or in M-Crew never were intended to do this much, rather like the 1700 character limit on metadata on legacy MDs).

 

However you can easily transfer one disk. And then another. And then the remains of a third (since LP2 it will only fit about 2-and-a-bit disks). And of course to calculate all this in the firmware - it ain't gonna happen.

 

I will try and copy 3CDs to LP4, but that has never worked that well. The only good way I know to do LP4 is at playback speed, using optical input. Then it always turns out very well.