AKAI GX-635-D Capstan Motor dreht nicht

+A -A
#1 erstellt: 18. Mai 2010, 04:50
Hi, bei meiner AKAI GX-635-D Bandmaschine dreht der
CapstanMotor nicht. Bei Schalten von Play oder RevPlay
zieht der Andruckrollen-Hubmagnet an, aber CapstanMotor
dreht nicht.
Eigentlich wäre es (Servicemanual vorhanden) einfach
die Servosteuerung zu messen/reparieren, wenn man denn
an die Platine herankäme. Nach Lösen von 3 Blechschrauben
läßt sich die Motorsteuerplatine nur wenig wegklappen,
geschweige denn herausziehen, dutzende (zu kurze) Wirewrap
Drähte verhindern es.
So kommt man kaum an die Lötseite heran.
Für Messungen muß die Platine auch angeschlossen bleiben.
Lose im Chassis rumhängend auch Gefahr von Kurzschlüssen.
Gern würde ich verdächtige Halbleiter (2SC458), Elkos und
Motorkondensatoren (RC-Glieder) erneuern,ist aber praktisch unmöglich bzw elende Fummelei.Je mehr man an den Volldrähten
biegt/bewegt, desto eher brechen sie.
Sollte man den Trafoblock demontieren, der vor der Servoplatine (Lötseite) liegt ? (Auch sehr aufwändig).
Oder alle Wrapdrähte zunächst dokumentieren, abwickeln, Platine prüfen, dann alle Drähte wieder anlöten ?
Ebenfalls elend aufwändig.

Frage zur Schaltung: Ein NPN Transistor TR13 steuert den
Gleichrichter D9 so an, daß die 80VAC für den Capstanmotor
zugeführt bzw durch den GR geleitet werden.
Die Transistor-GR Kombination scheint ein DC-gesteuerter
AC-Schalter zu sein.
Wie funktioniert dies genau ?
Kann man die AC-Anschlüsse des GR D9 testweise kurzschließen, um zu prüfen, ob der Motor überhaupt noch dreht ?
(Ohne TR13 zu beschädigen?)

Habe beide Teilspulen des Capstanmotors gemessen,
eine mit ca 160 Ohm, die andere mit ca 100 Ohm.
Motor wurde von der Platine abgeklemmt.
Diese Ungleichheit könnte auf defekte Wicklungen deuten.
Müssen beide Wicklungen den gleichen Widerstand haben ?
Sind die nominalen Werte bekannt ?

Gruß und Dank
#2 erstellt: 18. Mai 2010, 05:35

Kann man die AC-Anschlüsse des GR D9 testweise kurzschließen, um zu prüfen, ob der Motor überhaupt noch dreht ?
(Ohne TR13 zu beschädigen?)

NO, but you may short-circuit the DC-side of D9 (= Collector-Emitter of TR13)
TR13 is just a sort of a variable resistor across D9. If TR13 doesn't conduct, no AC will be going to the motor and if it conducts fully, all AC will go to the motor.
By varying the ON-time of TR13 across D9 you control what AC is supplied to the motor and hence how fast it's rotating (Pulse-Width Modulation)
This is explained in the service-manual on pages 20~22


#3 erstellt: 18. Mai 2010, 15:05
I read that pages in SM dealing with capstan servo operation,
on collector TR12 appears the DC voltage (measured 4V DC) with ripple (pwm pulses) like shown in figure (J) on page 21.
This signal fades across R30 and is weak at base TR13.
Collector and Emitter TR13 are zero volts.
R30 is ok , 100 Ohms.
TR13 is ok, measured.
I shorted CE of TR13 , no reaction, no castan motor move.
GX-635 is in PLAY mode.
Seems the motor gets no AC , maybe C19/20 are defect.
AC 80V is present on wirewrap-points 72 and 67.
I have (i guess) the CEE/UK model. So CR devices are separate
R and C. Also C19 is installed and the 50/60Hz switch misses.

OK, forcing motor to run by shorting CE of TR13 does not work.
I will again short CE TR13 and look for reason why motor gets no AC.

When does the Capstanmotor start running ?
- with poweron AC Mainswitch
- with press of PLAY or REV

#4 erstellt: 18. Mai 2010, 15:17
"If TR13 doesn't conduct, no AC will be going to the motor and if it conducts fully, all AC will go to the motor."

This means that both + and - ends of D9 are clamped to ground
and shorted via TR13. I did this, no motor move.
Then AC will go through D9.
D9 ist checked: ok.

So it must be riskless to desolder D9 and make a short between both AC ends on circuit board ?
To "hardforce" AC to the motor and exclude D9/TR13 as possible error source.
#5 erstellt: 18. Mai 2010, 16:13
Rectifier D9 was defect, replaced: Motor runs.
Now PLAY , REV , FF , REW , PAUSE is OK.
Also the start,stop,reverse delay times when switching
PLAY to REV or REW to FF. No tapeloops with 18cm plastic reels.
But (the big but): PLAY/REV speed is too high and does not
react on speed switch 19 - 9,5cm/s.
I will check speed switch, pitch potentiometer and
speed prelevel trimpots VR1 VR2.
Also again servo signals TR5 - TR13.
I think speed change must have effect on servo signals.
@TRICOMP: When this 635 runs, the 630 is free...
#6 erstellt: 18. Mai 2010, 16:43

Nice work
On my GX-630D that was stolen, TR13 (TR1 on GX-630D) became defective within the first year. I replaced it with a MJE3055 and it never failed again.

I'm planning a trip this summer....



[Beitrag von tri-comp am 18. Mai 2010, 17:45 bearbeitet]
#7 erstellt: 18. Mai 2010, 17:37
MJE3055: Nice tip.
I desoldered TR13 and test standalone:
seems to be OK, but from C to E i have ca 700mV in one
polarity direction (MultimeterTransistorTest) , viceversa
isolation. So as if there were a built-in CE diode.
But datasheet does not say anything about CE diode.
For safety i will replace TR13 (2SC2336) 180V 1,5A 25W 95MHz
by MJE3055 or similar type.
Testwise shorting CE of TR13 when running in PLAYmode has
no effect on motorspeed, so as if TR13 ist totally conducting.
#8 erstellt: 18. Mai 2010, 18:15
TR13 is defect, the diode effect between CE was a damage.
Replaced for original 2SC2336 with an 2SC3446.
Speedswitching now works ok.
MJE3055 seems to have too few voltage, only 80V.
Original 2SC2336 has 180 V.
#9 erstellt: 18. Mai 2010, 18:45
Yes, I know. I didn't check that at the time of repair. I was working at an US-Airforce base at the moment and had to take what was available.
Never the less the MJE3055 worked for more that 10 years until I lost the recorder.
Your choice seems to be a better one.
Congrat's with repairs

#10 erstellt: 18. Mai 2010, 20:08
2SC3446 is much stronger as original 2SC2336 , but 2SC3446
is marked as switching transistor.
For TR13 is used as variable resistor, maybe we must have
a definite "Kennlinie" (Uce curves).
So i am not shure if that replacement has negative effect on
Wow and Flutter speed modulation.
For safety it is a must to replace original part or equivalent type.
#11 erstellt: 06. Jun 2010, 19:59
I have exactly the same problem today with my GX-635D. It ran in a normal way, but at a certain point the speed went uncontrolled up and up. At that point I stopped the tape. Since then, I never got it running again.
As described in this thread, the capstan motor stopped running, although it should run all the time.

Do you think that my taperecorder suffer the same problem? I'm not a specialist although I have some basic knowledge and skills on electronics. Do you think I can manage to repair my recorder? The components themself are not expensive and relative easely to obtain.

Please, I appreciate your advice!

Regards, Ivan (Belgium)
#12 erstellt: 06. Jun 2010, 22:50
Refer to servicemanual:
Replace CapstanMotorDriver TR13 and Rectifier D9 .
Desoldering/partchanging is a litte bit difficult
on Motordriver/Logicboard (the upper one), because of
short wires and few mounting space.
Pull AC Mains cord.
Loosen 3 screws that fix board to chassis, tie wires off
holders and pull board out as far as wirelengths allow.
D9 is located at one rear edge of board, TR13 is MOUNTED SEPARATE on chassis near main transformer between Reelmotors.So only change of TR13 needs no demontage of motordriverboard.First change TR13, then test.
If works, do nothing, if not, replace D9.
Do not forget insulation plate between new TR13 and chassis.
Apply new thermal conduction grease both sides of insulation plate. Metal case of TR13 MUST NOT BE CONDUCTIVE TO CHASSIS !

In my case capsmot stopped because of burnout D9, mot got no power.
After replace D9 capsmot ran full speed too fast,
because TR13 was (partially) shorted.(Same as your problem).
After replace TR13 capstan speed was nominal and reacts
on pitch (varispeed knob/between the reels).
This is a sign that capstan servo loop is "closed" and all OK.
So if your capsmot runs too fast, i think TR13 is shorted.
As "tricomp" said: Shorting one AC-end of D9 causes capsmot
run max speed, disconnected AC-end lets capsmot stop.
TR13 gives "a half connection" (works like variable resistor)
depending on tachogenerator mounted on capsmot shaft
who generates pulses.
If one shorts + and - pins on a rectifier, also both AC pins
are connected (with a voltagedrop of 1.4 Volts).
So if both + and - pin of D9 are shorted to ground (which does TR13), capsmot runs max speed.

Replace also all transistors 2SC458 on audioboard (the lower one) against 2SC945 .Defect 2SC458 (Hitachi) cause
NF-signal interruption or distortion.

Apply contact spray TESLANOL T6 into FWD-REV switches
on audioboard driven by a big electromagnet.
Do not use KONTAKT 60 which contains acid.
#13 erstellt: 07. Jun 2010, 21:18
Thank you for your reaction!
I read your explanation and checked on my recorder and the Service Manual.
I found the components TR13 and D9. I understand now why working on TR13 is a lot easier.
Availability of the components:
- TR13 is no problem, it is still available on the market.
- D9 however requires an equivalent. I think KBP02M/3N248 from General Semiconductors may do the job (it has the same dimensions too as the original part). Do you agree on this equivalent?
I will first try to fix the motor problem since this is the real problem at the moment. The audio part will come later and is probably worth working at (is it recommended to change certain capacitors too?).

Thank you too for the other suggestions, I will certainly consider these after fixing the problem.

I will update this thread as soon as I make some progress (time is limited due to professional occupations).

See you!
#14 erstellt: 08. Jun 2010, 01:23
Original rectifier D9 has 1,5 Amperes and the capsmot
runs on less than 100V AC. So take the "electrical biggest"
spare type that fits mechanical and pinout. But first check desoldered D9 if its 4 diodes are ok. There are 2 more rects on motor board, replace them also.

Often the phase-shift motorcapacitor (the one also nearly
all AC 2-phase mots have) has lost capacity.
Check or replace.

The CR-units across motor relais sometimes blow off with
much fog. Replacing them can avoid this.

Replacing all electrolyt caps can not harm, but is an "endless" work. If all runs and audio performance (do a
amplitude vs frequency measure) is fine, its not necessary.
But with all new caps you are on the safe side.
Thats the difference between repair and restauration.
Perhaps "only" change elcaps on audio board or only those
in signal path.

Look for unbroken paint seals on head justage screws.
If you have a reference playbacklevel tape and some other
equipment you can adjust heads mechanical and trimpots electrical to achieve minimal source-tape sound difference.
I had to to this on my GX-635 , all 6 heads were deadjusted
by an neanderthaler.
Adjustments mechanical: tape height (tracking), tilt,
azimuth , both play and rec , both run directions.
Electrical:playback level, meter level, reclevel, bias.
Also nominal tape speed.

Do not lubricate caps- and reelmot bearings except with
special sinter-bearing-special-oil.
All other bearings need some high quality acid free oil
or grease.Do not forget the hidden skate-bearings (under head plate) of capstan gummiroll pull sledge .

Apply TESLANOL T6 to all trimpots,other pots and switches.
Dont forget the single ("tongue") contact near reelsize switch. Determines the start/stop reel timing when switching
FF to REW and vice versa depending on selected reelsize.

Change pri-AC-voltage from 220V to 240V , change one wire
position , use 240V pri transformer coil. See servicemanual.

#15 erstellt: 14. Jun 2010, 18:46
OK. I'm back with a little feedback: replacement of TR13 was very easy to do, although the result was nill .
I ordered a set of bridge rectifiers for replacement of D9 and the other 3. I expect to get these this week.
I need some more patience...

I'll be back!
#16 erstellt: 16. Jun 2010, 18:39
I received the bridge rectifiers yesterday and replaced D9 first. Measering the old component revealed that 2 diodes were down, they conducted in neither way.
Replacement resulted in a succes. The capstan runs again AND in the correct pace on both speeds.
I'll leave the machine open until I'm convinced of the good mechanical operation.
At the same time I also replaced D1, D3 and D50 that resides on the same board. Because I suppose these components may soon suffer the same issue (wear by time?). I read on another forum that D50 went down too...

Anyway, I will continue the other restauration works in order to end up with a perfect working machine with the least possible chance of breaking down again.
In the end, it remains an old machine ... but I love it!

Thanks again for all the good advice!
#17 erstellt: 19. Jun 2010, 10:01

Capstan runs still ok, I suppose this issue is solved.
2 questions:
1/ Today I lost for a while the right channel in reverse mode only. I examined the path of the signal and by accident I solved the problem by touching slightly the capacitor C31.
Could this mean that C31 has a problem? Or is this a faulty contact on C31?
However, I don't understand how the right channel operates ok in forward play because C31 lies in the signal path of both forward and revers play.

2/ "Often the phase-shift motorcapacitor (the one also nearly
all AC 2-phase mots have) has lost capacity.
Check or replace."
Can you explain more in detail where these capacitors are on the board (C-number) and how I determine if they are still OK?
#18 erstellt: 19. Jun 2010, 10:27
1a/ this maybe contact problem of relaydriven FWD-REV-MultiContactSwitch.
Apply TESLANOL T6 and move switch by hand often.
Better: Dismount mechanic above FWD-REV-switches and
desolder them.Take them out the 635.
Fill a metal polish fluid (i take SIDOL Metallputz)
into the switches and move them often. Give some time
to solve oxid layers , half hour.
Put switches in warm waterbath and wash out ALL polish fluid.
Dry with hairdryer and seal with TESLANOL T6 or KONTAKT 61.

1b/ or a 2SC458 Transistor is defect.Replace all 2SC458 vs
2SC945 . In contradiction to circuit diagram there were
built in 2SC458 at locations where already 2SC945 were drawn.
So check IN THE MACHINE every transistor, if it is an 2SC458 replace vs 2SC945.

2/ in the circuit diagram the motorcaps are located/connected near the 2 motor coils.
The cap shifts phase of one motor coil to force a clock
or counterclockwise turning.
If the capstanmotor runs quiet without hum the cap maybe
OK , but check is safe.

#19 erstellt: 19. Okt 2010, 15:35
OK. I left this issue to rest for a while, but finally I found some time to look more closely at the problem with the right channel no operating in reverse mode.
1a/ I simply applies the Teslanol T2, but did not dismount the switch. For some reason I had the feeling that this was not the issue.
1b/ I then replaced all the transistors, as you specified, in the signalpath of the right channel. But this didn't change a thing.

I than started to follow the signal path from the heads on. The first thing I saw was that the connector of the REV-head on the signalboard was gone. The wires were directly soldered onto the board. The soldering was very messy, so I removed the four wires and all remaining soldering material. I soldered the wired again and YES this solved the problem. Of course: if no proper signal is coming from the heads, no sound will come from the output!
Why didn't I look at this first!?

Next steps:
- I will replace all remaining old transistors.
- I will look at the capacitors, but the machine is not humming at all.
- I will take the machine to my brother who has all the gear for tuning the machine. 2 years ago we revived a Revox B77, installed brand new heads (made in Belgium, by the way) and tuned this machine for LPR35, which is the type of tape I'm using all the time.
Although the Akai sounds OK, it seems to sound a little dull compared to the Revox.

Update coming up!
#20 erstellt: 20. Okt 2010, 00:34
That anyone has done a crash repair was not expectable.
Those things are much worse than the worst "normal" problem/error.
Look also for broken colour seals at head adjust screws,
most dilettants "work" here.
For replace bad 2SC458 vs 2SC945 the BC639 also works fine.
Same Pinorder, direct replacement without turning legs.
#21 erstellt: 11. Jun 2012, 16:07
Final feedback:
It's been a while ...
My tape deck GX635D is now running fine for 4 months AND the problem on the right channel is solved.
After many headaches, I decided to replace the playback head on the reverse side. This solved the problem. The original head seemed to have intermittant issues ...
The "new" head is of course a used head of another deck. I managed to buy one from a broken deck. This took some time (6 months!).

Many thanks for all the help!
Das könnte Dich auch interessieren:
Akai GX-266 D Capstan dreht nicht
DaPickniker am 20.10.2010  –  Letzte Antwort am 25.10.2010  –  10 Beiträge
Reanimation AKAI GX-635-D
bukongahelas am 23.05.2010  –  Letzte Antwort am 23.05.2010  –  2 Beiträge
AKAI GX 635 D - Aufnahmeproblem !
Anna_log am 18.04.2007  –  Letzte Antwort am 29.04.2007  –  11 Beiträge
AKAI GX-635 Zählwerkriemen
bukongahelas am 31.12.2010  –  Letzte Antwort am 09.01.2011  –  14 Beiträge
w.georgi am 15.11.2008  –  Letzte Antwort am 18.11.2008  –  2 Beiträge
Tonbandgerät AKAI GX-635 D ganz leise?
spycatcher am 17.03.2007  –  Letzte Antwort am 22.03.2007  –  8 Beiträge
Akai GX 635 D - Aufnahme defekt
cyrano_de_b. am 25.12.2007  –  Letzte Antwort am 27.12.2007  –  6 Beiträge
Bandmaschine Akai GX-635 Bezugspegel
bukongahelas am 11.01.2014  –  Letzte Antwort am 12.01.2014  –  9 Beiträge
Akai 635 D Audioplatine
thesaurus am 11.12.2011  –  Letzte Antwort am 16.12.2011  –  6 Beiträge
Akai GX 653D blowing fuse capstan motor stops
Woadl am 04.07.2013  –  Letzte Antwort am 19.09.2013  –  24 Beiträge


Produkte in diesem Thread Widget schließen

Aktuelle Aktion

Partner Widget schließen

  • beyerdynamic Logo
  • DALI Logo
  • SAMSUNG Logo
  • TCL Logo

Forumsstatistik Widget schließen

  • Registrierte Mitglieder872.160 ( Heute: 37 )
  • Neuestes MitgliedAchim57
  • Gesamtzahl an Themen1.453.821
  • Gesamtzahl an Beiträgen19.662.951

Hersteller in diesem Thread Widget schließen