PS-8750 : speed issue solved !

turntables, tonearms, tuners, open-reel & K7 recorders...

PS-8750 : speed issue solved !

Postby Axel » Sun Jul 28, 2013 1:03 pm

As lucky as I was with this mint PS-8750, it developed random speed issues after a day or two.
Sometimes it would need to "warm up" during the night to stay on X'Tal speed locking the next day... but it wouldn't always work. Other times it would be spot on at power-on, then it would digress, generally faster, after a while - unstable, unusable, imprevisible.

I remember having read about the caps that feed (100µF/10V ; C131) and filter (10µF/350V ; C132) the neon/stroboscope - mainly the filtering one ; both are on the X'Tal lock PCB :
(ctrl+click to open in a new window for full-size)
Image


Pop the lid up, find the servo PCB is on the front (beige color ; the two blue flat pots are the primary speed controls) and the X'Tal PCB center (green color) - that's where the culprit is :
(ctrl+click to open in a new window for full-size)
Image


Unscrew the X'Tal PCB, locate the soldering points, remember where the negative is (facing the Mitsu ICs) :
(ctrl+click to open in a new window for full-size)
Image


Unsolder the old Nippon Chemi-con cap and discover the 'lyte was gone entirely - an empty cap that had left marks on the PCB but nothing beyond that - VERY lucky me !
(ctrl+click to open in a new window for full-size)
Image


The problem.
(ctrl+click to open in a new window for full-size)
Image


The new one in place.
(ctrl+click to open in a new window for full-size)
Image
________________________________________________________________________________


Put everything back together, power-on, start platter, hope, but the speed is variably fast again :(
Loud noise in the street, I go check on the balcony, maybe eight, ten seconds tops, I come back : the speed is locked !
It has been so, unwavering, for twenty minutes now. Twenty-five. Thirty. Thirty-five. More than an hour now - solved !

I'll neverhteless recap the rest of the Servo and X'Tal boards just in case, maybe even the power-supply and those unpolarized for motor startup, then deal with the auto-return function (just the belt ?) but I've got me a fully-functional and mint PS-8750 !


All I need now is a just-as-mint dustcover and that is going to be difficult - can't win all the time.
User avatar
Axel
Founders
 
Posts: 497
Joined: Tue Mar 02, 2010 10:39 am
Location: Paris

PS-8750 : speed issue NOT solved !

Postby Axel » Sat Aug 03, 2013 9:10 pm

Alas... NOT solved.

The repair above worked for a bit more than an hour, then X'Tal locked speed went back too fast.

A few days after, I put the 8750 back upside down, resoldered a few points of the X'Tal PCB, including those of the oscillator itself plus others at random, and this worked to perfection for almost four hours.
I thought I finally had it fixed then I switched it off for maybe a minute, then on again... and X'Tal speed was gone, just like that ! It hasn't come back so far and checking solder points again today didn't change anything so this 8750 be off to a tech very rapidly.

The regular servo mode works perfect (but there is °some° drift over long playing time) so the problem clearly rests with the X'Tal PCB only and is quite probably related to the double-sided / pass-through soldering of some of the points.
The rest works like a charm and even the X'Tal, when not obsessed with "faster I must go", gives perfect A / la / 440Hz pitch (which is important to me).
User avatar
Axel
Founders
 
Posts: 497
Joined: Tue Mar 02, 2010 10:39 am
Location: Paris

Re: PS-8750 : speed issue solved !

Postby Sam Z » Sun Aug 04, 2013 4:19 pm

I think a full recap is the first thing to do because although I've never had an 8750, my experience with my old Denons and Luxmans is that the caps are critical.

Can you correlate the problem with any other factors or is it totally unpredictable?
Sam Z
 
Posts: 17
Joined: Tue Jun 14, 2011 5:20 am
Location: Australia

Re: PS-8750 : speed issue solved !

Postby Axel » Sat Mar 15, 2014 5:20 am

At last, thanks to TVK Knobber Pat66's wit and persistence, the PS-8750 speed issue is declared done with and universally SOLVED.

This will be useful for many 8750 owners so I will post the long and short of it here in the coming days.
User avatar
Axel
Founders
 
Posts: 497
Joined: Tue Mar 02, 2010 10:39 am
Location: Paris

Re: PS-8750 : speed issue solved !

Postby Sam Z » Sat Mar 15, 2014 5:45 am

Nice. I'd be curious to see what the problem was. These old decks can be tricky.
Sam Z
 
Posts: 17
Joined: Tue Jun 14, 2011 5:20 am
Location: Australia

Re: PS-8750 : speed issue solved !

Postby Stoffie » Sat Mar 15, 2014 5:54 pm

Glad to hear that and indeed a testament to Pat66's skills! Too bad he is so far away because finding a good tech is hard!

Anyway, enjoy the table, another priceless classic saved.
User avatar
Stoffie
Knobber
 
Posts: 99
Joined: Tue Jun 29, 2010 9:10 pm
Location: Belgium

PS-8750 speed issue solved at last !

Postby Axel » Sun Mar 16, 2014 8:03 pm

Like many 1970s direct-drive turntables with any of the speed locking circuit variants, the PS-8750 can go crazy.
It often does, randomly or not, and it only does so when its star feature is used, for perfect A / la / 440Hz pitch, namely : the X'Tal Lock.

Pity not to use its second best feature (the first being the carbon PUA-1600S tonearm and its SH-160 carbon headshell :) because when the 8750 works properly, it is a truly amazing turntable that shamelessly, radically beats most of the contemporary competition. And beyond, easy.

Sony, however, like Luxman, even if less compulsively so, is known to have produced many undocumented circuit variants (the TA-3200F or the many different filters of the APM-22ES come to mind) and unnecessary versions depending on planned export markets (how many TAE-20F/FB versions can you live with ?)
Sony even used between late 1977 and 1979 sub-par PCB manufacturing/soldering contractors (and/or quality control routines) for some but not all of its audio production, rendering the repair of otherwise truly remarkable components never ending nightmares. Sometimes it was due to Sony itself (circuit design, layout, marketing demands paired with too much sake), sometimes not.

But sometimes it's only the parts themselves that have aged beyond what they were planned for in the first place.
Who'd have thought that about two thousand fervent lads would still use a PS-8750 or an ST-A7B fourty years after their production ?
Nobody ! All this was at the same time made to last a lifetime and at the same time… not. New stuff was produced every year, if not every season, and what made the profitable part of the hi-fi business did indeed buy the newer and junk the older every year, or every other year - same difference.

The PS-8750's problems come not from a lousy circuit design, not from manufacturing hiccups, and not from the salmon mousse either but, simply, from ageing parts. The minor work I did on mine didn't change a thing, or not beyond an hour or a day at any rate, so these "easier" solutions proved completely useless : dead strobe caps, MSM7756 replacement, resoldering here and there etc. Pat66 on the other hand only made the repair outlined below and his 8750 locked steady instantly and has remained so, unwaveringly, ever since. Not that having a non-flickering strobe light isn't good but that clearly wasn't the culprit.

So, on with the show, thanks to TVK Knobber Pat66, who owns every ES, ESII, pre-Esprit and ESPRIT component Sony ever made and knows them inside out, having restored and repaired all of them.
His diagnosis was made possible studying his japanese PS-8750, my own multivoltage EU model and that of another Knobber, all three suffering from the same symptoms, plus a fourth which made the exception confirming the rule : it works and has never even been opened !


Being the angel he his, Pat66 also pointed me to a mislabelled but absolutely mint yet cheap 8750 dustcover on eBay which mine was missing despite its equally "new" state. Pat', you're the man !
(Well, no, bad memory : 'twas Knobber Romain who pointed it out so thanks again Romain - you're da man too !)
______________________________________



Said symptoms are largely the same on all PS-8750 discussed here and there : erratic speed fluctuations or, mostly, too high a speed with the X'Tal Lock circuit ON, differing levels of madness whether using the 45rpm or 33 1/3rpm speeds, different behaviours whether the turntable is hot, cold, if the sun is shining or if your income tax bill just came in.

A contrario, the non-Locked mode works like a charm, always, so this points rather clearly toward the green PCB dedicated to X'Tal Lock mode for the problem's source. Plenty can go wrong there as the ICs used are old and "outdated" items, the caps may have dried and this two-sided board can also prefigure the madness of the ST-A7B - shorted or cut PCB traces, humidity, heat, dry soldering, hygrometry, argh and argh again.


Measurements & Diagnosis
The PS-8750's circuit uses a 960Hz pilot frequency, reduced and split differently for either speeds (33 1/3 and 45rpm) when the X'Tal Lock mode is ON.
This frequency division(s) is where one should investigate and to confirm the mess is there, and there only, measure pin 8 of IC106.

33 1/3 rpm mode :
The square-wave should read 35,55Hz (28,1µs period) ; this corresponds to a x27 division of the 960Hz pilot found on pin 11 of IC101.
When the speed problem is up, the same pin 8 of IC106 gives a reading of about 73,8Hz (13,5µs period) which makes for a division ratio of about x13 instead of x27 ; this explains the "too fast I must I must I must go" manners.

45 rpm mode :
The same measures should give readings of 96Hz (too fast) or 48Hz (normal), making division ratios of x10 (too fast) or x20 (normal) respectively - it's the very same x2 error !


Repair #1
As IC104 and IC106 are both present only to perform such frequency divisions, they are potentially both the source(s) of the speed problem(s). These ICs are Mitsubishi made, marked M53200P, but are in fact copies of Texas Instrument's SN7400N ICs as the PCB marking clearly states - but you have to take the Mitsus out to see that :)

Pat66 suggests to stick to original SN7400N replacements and to avoid the later versions (74LS00, 74HC00, 74HTC00…) as if they are equivalent in function, their electrical characteristics are not - no need to solve one bloody problem only to have another rise up !
Once done, that's it : perfect A / la / 440Hz pitch and no reefer madness no more.


Repair #2
Some 8750s not only go too fast but suffer from strange speed fluctuations, as if the platter hesitated between START and STOP and couldn't decide. Pat66 still has to investigate this as neiter his or mine suffer from that (supplementary) illness.

In my (very humble) opinion, this may be related to the electromagnetic brake which comes in at the end of a side or when using the "reject" function. If the auto-stop belt is loose or gone, or the mechanical cam disengaged, the tonearm's auto-return motor will engage but not be able to go through its entire cycle, therefore not ending said cycle ; the 8750 will sit in limbo wondering if it has to stop or keep going, the electromagnetic brake being triggered on and off relentlessly. Maybe.
Or also related to the dead /damaged /aged IC104/106 which wander electrically and divide the 960Hz pilot by whatever they feel like dividing on account of the income tax bill, the cat's mood and the neighbour upstairs singing Barry Manilow's best in Karaoke mode with the loudness on.
______________________________________



That's it.
Now go scavenge old whatevers to retrieve as many precious SN7400Ns as you can !

Image


The PS-8750 is worth every effort as noted here for instance : http://www.audioasylum.com/audio/vinyl/messages/93/930780.html.

It is no wonder the Japanese magazine MJ chose it as one of the landmark turntables in its 1970-1980 anthology published in 2001.
The other items were the Technics SP-10, Linn LP-12, Micro DDX-1000 and RX-5000/RY-5500, Aurex SR-M99, Lux PD555 and Trio L-07D - same league, top of the pops.

As a sidenote, the PUA-1600S works properly only when paired with fairly easy-to-comply-with MC cartridges (MMs don't exist in my world :).
A Sony XL-33 /33L /44 /44L /55 or the Denon DL-110 /160 if you still can find them will do fine just like the Ortofon Cadenza and newer Quintet, all in the 10...20 compliance range ; the Ortofon MC20 is the upper limit at 25 x10-6/Dyne.
But don't go put a huge static bug in need of 5g tracking force on it and certainly not a DL-103 with its compliance of... 5 !
User avatar
Axel
Founders
 
Posts: 497
Joined: Tue Mar 02, 2010 10:39 am
Location: Paris

Re: PS-8750 : speed issue solved !

Postby Sam Z » Wed Mar 26, 2014 1:06 pm

Excellent, excellent post Axel. Sony did good with this one.
Sam Z
 
Posts: 17
Joined: Tue Jun 14, 2011 5:20 am
Location: Australia

PS-8750 speed issue solved - videos !

Postby Axel » Sun Apr 20, 2014 8:23 pm

Now for the visual proof the above works.
Posted two videos done with an iPhone 4S and the X'Tal Lock ON.

1. http://www.thevintageknob.org/tvk_talk/axelimages/PS8750-repaired1.mp4
Some wavering, frightening at first but mostly befuddling because... I cannot hear it !
Being a musician first and foremost, I am very sensitive to pitch and this stroboscope wavering goes completely undetected, whatever the music (boombang, concerto, opera, piano, whatever) or speed (33 1/3 or 45rpm).

2. http://www.thevintageknob.org/tvk_talk/axelimages/PS8750-repaired2.mp4
Different point of view. Still wavering. However, close scrutiny and a third video (not posted) taken nearly tangentially from the strobe area show that only the diecasting of the platter itself is to be blamed : the strobe dots do not have the same depth depending on their location ! They can be seen, very slightly, "moving" forward, then backward, then forward etc.
Also, the clear line separating the 33 1/3 dots from the 45 dots above and the line above the latter vary in thickness quite visibly. Since this and the dots' thickness changes is happening in sync, 'twasn't the salmon mousse either but the platter's outer geometry diecasting accuracy.


Small differences but when filmed so close and with a centering frame added it looks as if there must 1% pitch differences with each wave. There isn't. None. Nada. Nichts !
Well, there must be because the 8750 isn't an L-07D or a TT-801 but it remains completely beyond human hearing - and I got mine checked last year and it still is rather extended and accurate within ± 2dB between 60Hz and 12kHz. (pretty amazing given the amount of time I've spent with cans on my ears and blasting guitar amps)

I wish the strobe dots could come out sharper but that has very little importance : changing IC104 and IC106 brings a PS-8750 back from the dead and with reliable pitch accuracy.
Done - time for music now.
User avatar
Axel
Founders
 
Posts: 497
Joined: Tue Mar 02, 2010 10:39 am
Location: Paris

Re: PS-8750 : speed issue solved !

Postby Axel » Fri Apr 24, 2015 2:05 pm

OK - a year after the above post.

The speed issue wasn't *really* solved then : after a few weeks of perfect speed, one morning, out of the blue, it went too fast again.
I got bored of trying, set it aside and brought the PS-X800 back in.

Then Pat66 suggested that the center crux of this frustrating Scheisse was the X'Tal Lock switch itself and its uselessly complex activating system based on a pushing spring. On one of his, wiggling the switch (it's metal so one has to wiggle with authority :) triggered either perfect speed or mayhem. A full cleaning and re-setting solved the problem, same on his TTS-8000.
So... I wiggled, cleaned, wiggled, cleaned for hours with various WD40 equivalents... no go : still too f****** fast.
I got really bored hoping and cleaning and wiggling and hoping and pushed it aside to forget about it entirely - enough.

BUT
Today I got it out of retirement to get a big book stored behind it... and plugged it in - just for the fun of deception.
Switch ON, play with X'Tal Lock ON and.......... jkl#Fgkm/Uwd,k]Cmw”K,ù¡ NO GO: still too fast ! So I wiggle the bloody switch, again, and again wiggle and then some but I don't unplug it out of despair and do whatever it is I have to do elsewhere for a while.

Come back to it after about 45 minutes, having indeed forgotten about it entirely. But I hit "play". AND IT WORKS : perfect speed at last.
I will now leave it spinning for 24 hours and see what happens.


UPDATE, the day after
Well, here it is : IT WORKS !
No need to wiggle, De-Oxit or pray : the key to that is... heat.

When "warm", it works as in 1975 : perfect.
When "cold", it doesn't.

So : change the strobe cap, change the SN7400N ICs, check everything else as you normally would. And keep it powered ON.
User avatar
Axel
Founders
 
Posts: 497
Joined: Tue Mar 02, 2010 10:39 am
Location: Paris

Next

Return to Sony analogue sources

Who is online

Users browsing this forum: No registered users and 1 guest

cron