Thanks, Ian. for this info. However before I can make my order for replacement caps, I find there is a 'logistics' problem - inasmuch as the Service Manual doesn't list ALL the caps on the two PCB's. 'Frinstance on the Pre Amp PCB (TH-5001A) the List jumps from C3 to C15 and C15 to C20 etc. So I cannot tell what, for example, C8 or C12 are. (That's just sticking a pin in the problem) Ergo this Manual is incomplete and I am therefore unable to make up my shopping list. Has anyone access to a full, comprehensive components list for these two PCB's on the AKAI GX-635D? Oh dear! Michael (UK)
I looked for and only found 2 free manual sources on the web. Hifiengine's manual was corrupted but Electrotanya's is linked here. You are correct that the Akai parts list is only for recommended Akai spares - presumably, other common parts were sourced by the service agent as required. I doubt you will find the required PCB level parts lists for these now.
As I'm not a pro. service technician, until the internet caught up with me, I made do with only a bare schematic or nothing at all. All I do is scan over the schematic, section by section, looking for the polarized capacitor symbols and writing a list according to the spec alongside e.g. 1.5/10 = 1.5uF / 10 volt. Check with the real parts for anything unusual (as in different appearance to the typical), find the reason so you aren't looking for redundant (eg tantalums) types and not forgetting the NP caps. Where space permits, film caps of ~1uF or less can be substituted for these, with some benefits. Obviously, you need a proper supplier catalogue and a precision, short rule for this.
Then edit the list down to one or two voltage ranges and you are down to fewer types, less confusion and better price breaks.
I looked for and only found 2 free manual sources on the web. Hifiengine's manual was corrupted but Electrotanya's is linked here. You are correct that the Akai parts list is only for recommended Akai spares - presumably, other common parts were sourced by the service agent as required. I doubt you will find the required PCB level parts lists for these now.
As I'm not a pro. service technician, until the internet caught up with me, I made do with only a bare schematic or nothing at all. All I do is scan over the schematic, section by section, looking for the polarized capacitor symbols and writing a list according to the spec alongside e.g. 1.5/10 = 1.5uF / 10 volt. Check with the real parts for anything unusual (as in different appearance to the typical), find the reason so you aren't looking for redundant (eg tantalums) types and not forgetting the NP caps. Where space permits, film caps of ~1uF or less can be substituted for these, with some benefits. Obviously, you need a proper supplier catalogue and a precision, short rule for this.
Then edit the list down to one or two voltage ranges and you are down to fewer types, less confusion and better price breaks.
Many thanks, Ian. I had already started scrutinizing the two Cct.Diags. for Caps. in order to fill in those gaps in the Lists. Thanks for the suggestion of the other site for a Manual - I looked at it and it is identical to my Hard Copy. The caps are missing there as well. So far I have got to C15 22/25 -[|- on the Pre Amp Cct. Diag. This is, of course, but one of two caps (L) and (R) but the info is only given against the one cap in the (L) channel. There is a Symbol I haven't come across before, against some of the Caps, employing a square bracket: -[|- I wonder what this designates? There's also the standard Cap symbol. At this point I fired up my computer and found your Most Helpful information. I shall continue to scrutinize both Cct. Diag's. logging my findings and then comparing the results to those incomplete Lists on Pages 58 and 59 in the Manual. If, with help, I am able to complete the Lists, it might be helpful to others to publish them here. I'm currently cogitating on the purchase of a PEAK ESR+70 which can test Caps IN CCT. This way I may find many of the Caps are OK - what do you think? Michael (UK)
Last edited:
If you plan to replace the caps, do it completely, save your money on testing the old ones only to find a range of figures from fair to diabolical. You will also find that ESR varies with cap. type, brand and grade. Modern types will be far lower ESR than 35 year old types, so the scales of "normal" readings won't be comparable. It takes a bit of experience to use an ESR meter confidently unless you have apples/apples comparisons to make. I have an in-circuit ESR meter too but it is still not suitable for all "in-circuits", as I found.
In a nutshell, even though there are cheaper meters just as effective, you would spend a lot to save a little. Another consideration is that you need to spend hours to-ing and fro-ing for opinions from others on whether a 20p component should or should not be replaced when it could or might not be marginal, depending on the application specifics and readings you get. Sheer madness! Replace them and be done with it - one board at a time.
The symbol -[- used by Akai is the symbol for all the polarized electrolytics on that drawing. The smaller terminal is + . If you have been looking at plain black symbols like = signs, they are other, unpolarized types like polyester film, styrol, mica types. You shouldn't need to touch them. Note, there are LL (Low leakage) electrolytic types too, as defined on the lower right of the drawing. I didn't see any non-polarized electrolytics, or Tantalums specified on the preamp/playback board, though.
Tantalums are usually solid coloured beads, sticking up like balloons but I wouldn't expect to see them on Japanese gear. Akai were very conventional and economical with components. The PCB layouts suggest axial caps here - that's cylindrical with leads protruding from either end, lying flat on the board. This is awkward because few types are manufactured like this any more and it means stretching out the leads on radial caps to span the holes instead.
Types of capacitor - Wikipedia, the free encyclopedia
I suggest you remove and inspect what you can't read or deduce from the schematic. Sure, it's not going to be an easy task but that's just one reason why service costs are high or uneconomic on old gear in particular.
In a nutshell, even though there are cheaper meters just as effective, you would spend a lot to save a little. Another consideration is that you need to spend hours to-ing and fro-ing for opinions from others on whether a 20p component should or should not be replaced when it could or might not be marginal, depending on the application specifics and readings you get. Sheer madness! Replace them and be done with it - one board at a time.
The symbol -[- used by Akai is the symbol for all the polarized electrolytics on that drawing. The smaller terminal is + . If you have been looking at plain black symbols like = signs, they are other, unpolarized types like polyester film, styrol, mica types. You shouldn't need to touch them. Note, there are LL (Low leakage) electrolytic types too, as defined on the lower right of the drawing. I didn't see any non-polarized electrolytics, or Tantalums specified on the preamp/playback board, though.
Tantalums are usually solid coloured beads, sticking up like balloons but I wouldn't expect to see them on Japanese gear. Akai were very conventional and economical with components. The PCB layouts suggest axial caps here - that's cylindrical with leads protruding from either end, lying flat on the board. This is awkward because few types are manufactured like this any more and it means stretching out the leads on radial caps to span the holes instead.
Types of capacitor - Wikipedia, the free encyclopedia
I suggest you remove and inspect what you can't read or deduce from the schematic. Sure, it's not going to be an easy task but that's just one reason why service costs are high or uneconomic on old gear in particular.
So far I think the Caps marked -[|- are polarized Electrolytics and the -[ is the NEG. pole? Michael
Many thanks, Ian. You are right - I need to replace all the Electrolytics and Tantalums - but what of the NP's, the Styrol and the Metallized Film types? Maybe they are not prone to ageing - or are they? As soon as I have all the Caps logged from the two Cct.Diags. I'll post some observations for discussion and go on from there. In the mean-time there are several on the PRE AMP PCB List with bracketted suffixes like:
C15 Elect./C (LL, Homing) 22mF(M) 25WV
When finding this Cap on the Cct. Diag. it shews:
C15 22/25 adjacent the -[|- symbol in the Cct.
Other suffixes on the List: (Homing Type) (LL Homing) (Vert. Type) -which sounds obviously a vertically mounted Cap. etc.
Nearly finished the Pre Amp Cc. Diag.
Michael (UK)
The Tantalums are in th SYS.COM. PCB. Here also the List is deficient in several missing Caps. The Cct.Diags for the SYS.COM PCB. are spread over seven Cct. Dwgs on pp. 11 - 17 shewing the various Modes - STOP, FWD-PLAY, REV.PLAY,FWD-REC,REV-REC, FF,RWD MODES. so I shall have fun working that out! Here goes, but the drawings are so small and the printing so blurred it is impossible to decipher the Cap numbers with any degree of surety . . . I wonder if there's a clearer Dwg out there somewhere? Michael
Last edited:
I have enlarged (TH-1015S) p.11 (p.12 counting front as p.1) by (R) click+rotate and have made some headway identifying the Caps. Several are simply too blurred to identify. I'm missing values for: C8,C27-29,C32,C42, C50 & 51 and C55 to the end. In enlarging on the PC Monitor I get lost moving the image by the 'hand' so I have tried printing (CTRL+P) after selecting enlarge to 150% but cannot select which part of the image to enlarge. Any ideas about this? Michael
You won't be able to print out a PDF file with the freedom to adjust parts of it, as you can with image formats such as jpg, png or gif. PDFs have to be converted to an image file first before you can manipulate its scale and feed resized parts of it to a printer, AFAIK.
I don't even think you will be able to rely on the schematics for sufficient details when looking for equivalent parts, either. You will have to physically inspect them, check the leg spacing and read the labels of some types to be certain of what substitute types you should be looking for.
At this point, I don't see a need to tackle the motor and relay control boards unless they are showing signs of misbehaviour. As I understood, you have an input preamplifier only working on one channel and that should be checked first with a signal tracer to be sure where the signal is being lost. I assume from your description that the problem is confined to the preamplifier board (schematic 2-2). The logical place to start is there or you will become thoroughly confused by taking a random approach to problem solving by starting somewhere irrelevant. Use a logical process and work each signal stage through to completion, a step at a time and begin with the clear schematic, gaining some experience and confidence as you go.
A signal tracer can be as simple as a continuous mono radio or music signal at low line level, such as from another preamplifier, MP3 player, radio etc. via a shielded lead such as an RCA lead. The shield is clipped to input ground and the signal pin or wire to an ordinary polyester film capacitor, around 0.1 uF, in series to a crocodile clip or similar to inject a signal where you think it should be found, according to the schematic.
Another preamplifier, or line output with volume control, is used to control the signal if it does not have a volume control, so that you can attenuate the volume and avoid overdriving the tape preamp, since you won't have the protection of its own level control if you inject the signal after that point. Obviously, the tape player has to be working and connected such than you can hear or see the meter movement of the output of the tape preamplifier on test. You shouldn't need to have tape running for this monitoring, though.
Recapping is secondary to this fault finding but if you cannot locate the problem, this shotgun approach won't hurt as it's due now but don't try to tackle it all at once or you will never solve any problem satisfactorily.
I don't even think you will be able to rely on the schematics for sufficient details when looking for equivalent parts, either. You will have to physically inspect them, check the leg spacing and read the labels of some types to be certain of what substitute types you should be looking for.
At this point, I don't see a need to tackle the motor and relay control boards unless they are showing signs of misbehaviour. As I understood, you have an input preamplifier only working on one channel and that should be checked first with a signal tracer to be sure where the signal is being lost. I assume from your description that the problem is confined to the preamplifier board (schematic 2-2). The logical place to start is there or you will become thoroughly confused by taking a random approach to problem solving by starting somewhere irrelevant. Use a logical process and work each signal stage through to completion, a step at a time and begin with the clear schematic, gaining some experience and confidence as you go.
A signal tracer can be as simple as a continuous mono radio or music signal at low line level, such as from another preamplifier, MP3 player, radio etc. via a shielded lead such as an RCA lead. The shield is clipped to input ground and the signal pin or wire to an ordinary polyester film capacitor, around 0.1 uF, in series to a crocodile clip or similar to inject a signal where you think it should be found, according to the schematic.
Another preamplifier, or line output with volume control, is used to control the signal if it does not have a volume control, so that you can attenuate the volume and avoid overdriving the tape preamp, since you won't have the protection of its own level control if you inject the signal after that point. Obviously, the tape player has to be working and connected such than you can hear or see the meter movement of the output of the tape preamplifier on test. You shouldn't need to have tape running for this monitoring, though.
Recapping is secondary to this fault finding but if you cannot locate the problem, this shotgun approach won't hurt as it's due now but don't try to tackle it all at once or you will never solve any problem satisfactorily.
Thanks to your detailed analysis of the Akai problem. I feel I can now go ahead with more confidence. Since the Sys.Con board has, so far as I'm aware, no problems. It would probably be best to leave well alone and to recap just the Electrolytics in the Pre Amp and insert the new KSC1845's. Then re-assemble up to the stage when I can apply power.
My test equipment is pretty rudimentary - the PEAK DCA55, a couple of home-made signal generators with saw,sine,square wave plus a couple of DMM - a 'Holdpeak HP-9804 and an old Thandar TM361. I built the two Sig Gens with variable outputs of freq. and levels to inject tones into Stereo equipment. (A Uher Report 4200 actually) That's it. No 'scope.
Michael
It seems you have plenty of gear for this task. Any generator will do if you can adjust output down low enough for the gain in your tape preamplifiers. As you will be familiar with your DIY generator(s), you should have few problems. I would use sinewave mode and if the generator is DC coupled to the output terminals, i.e. direct from the oscillator with no coupling cap - just an output level switch and attenuator controls.
In that case, add a cap. as previously described to isolate the 2 circuits, avoiding DC coupling between them which could be catastrophic.
Needless to say, you compare the response to that in the good channel to determine what you should be getting. "Earsight" and your tape recorder's VU meters are quite good enough for this type of level testing when you are locating where the problem is,
In that case, add a cap. as previously described to isolate the 2 circuits, avoiding DC coupling between them which could be catastrophic.
Needless to say, you compare the response to that in the good channel to determine what you should be getting. "Earsight" and your tape recorder's VU meters are quite good enough for this type of level testing when you are locating where the problem is,
Thank you Ian, for this info. The output of my little Audio Generators is either direct from the Oscillator - or - through a decoupling 220mF elect. They have variable output levels through that decoupler of: from 0.01v min. to 1.52v max.
I shall start my rebuilding the Pre Amps by first inserting KSC1845's in the now vacant positions. As I said before, Akai thoughtfully marked every Transistor mounting point by using a white rectangle surrounding the three holes. One end of the rectangle is 'whited out' and includes just one pin hole. In this way all transistor positions are 'polarized'!
Earlier in this thread I Posted the result of subjecting each old transistor to the DCA55 and gave the readout of each as I removed it. I also made a second list of the readouts from the new KSC1845's but this wasn't Posted.
Burning question is: is it at all important to attempt matching the Hfe of the new to the old . . . NO that's a stupid idea! That's why there's a problem! If you find that Post of the readouts you will see how wildly the old transistors differ. Apart from about three or so mavericks they are all 2SC458's. Have you any suggestions to make on the procedure?
Having inserted all the new transistors I shall re-assemble the Akai and make an audio test to see if the machine has resurrected itself from this radical surgery. If it works then I shall leave replacement of Caps to another time. "Sufficient unto the day is the Evil thereof" - as they say. Michael
Last edited:
In an earlier post I suggested not bothering with matching and went on to say that Akai specified graded Hfe parts - i.e. the appropriate degree of matching is in the grade specification. The wild difference I saw was the BC109 replacement which was wild for sure, since BC109 spec. is ungraded.
Matching is only needed with crude circuits designed without attention to costs or a few topologies such as differential amplifiers (the long-tailed pair) and parallel power devices which need to share current evenly. Matching to limits closer than in available grade limits, which you could purchase, is something manufacturers and service people detest. It means time, special parts stocks and perhaps special test jigs.
This unit has no individually matched parts, simple single-supply circuitry and the only time you see matched pairs in Japanese consumer products is where they are produced in a dual package pair of adjacent transistor chips. These will already be quite closely matched, which is inherent in the manufacturing process.
If you buy KSC1845/A992 from Mouser or other approved stockists, you'll see they preserve the manufacturer grade code in the part stock number.
http://www.mouser.com/ds/2/149/KSC1845-88467.pdf
Fairchild Semiconductor KSC 1845 Transistors Bipolar - BJT | Mouser
Look at the 3-letter suffix and the first letter, E or F, specifies the available grade. In this case, F refers to an Hfe range of 300-600 and should present no problems with any part of the Akai circuitry from that quarter.
Realistically, you don't have much choice with parts that are now legacy products and only available in the most common options.
Hello Ian, Thanks very much for this! I didn't know the relevance of the suffix to the Transisitor's Hfe. The 80 KSC1845's I bought in October from Digi-Key in Thief River Falls USA cost £0.102 each + shipping. They have the suffix: FTACT-ND so, as you point out, they should be fine for the Pre Amp section. My DCA55 shows all to be in the range 339 - 391Hfe. I will solder them in place and do a test. Fingers crossed!. Michael
Anyway, good luck with replacements.- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.