• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Akai 4000DS right channel playback issue

D

dg_diyer

New Member
I am the original owner of an Akai 4000DS. Deck was idle for years with a dried out pinch roller that I just fixed. I have a lot of out of print music that I would like to recover. Just before the pinch roller dried out, everything worked fine.

Now, the right channel playback has a problem. When I turn on the deck the right VU meter gyrates hard to the right limit and then settles down after about 5 -10 seconds. After that, when I play a tape the music from the right channel is very faint and mostly overridden with hum that sounds like a grounding sound.

The left playback works fine. And when the deck is set to source, both VU meters act normally when a source sound is fed into the deck through the RCA input jacks. The playback sound is the same with headphones plugged into the headphone jack on the deck as playing the tape back through to speakers. So I don’t think the right output RCA connector is a problem.

I have looked at a schematic and have read other related threads. It seems like the signal is not getting amplified. So this points to the pre-amp board. I have read about the 2SC458 transistors and also the LD 3141 integrated circuit.

Based on these symptoms, which of these is likely to be the problem, the transistor or the IC?

Thanks for any help on this.
 
Register to hide this ad
Clean and exercise the record / play switch thoroughly first, also the track selector switch and tape / source switch.
 
Greetings from RojoLand!

If the switch cleaning and exercise doesn't help (and it certainly can't hurt!): Since you have normal function in Source mode, that clears all the LD3141 ICs (whew!). The issue is confined to TR3 and TR4 stages in the right channel head amp. Both are 2SC458s, but don't assume it's the transistors themselves. If you have a scope, trace the signal through the head amp from terminal H (amp input) up to terminal T (head amp output to Tape/Source switch). There are electrolytic capacitors in the signal path that could be at fault here.

You are lucky. My 4000DS had both channels dead, and I had to replace two LD3141 ICs (I used discrete components rather than trying to find those rare and troublesome ICs).

Take care,
—
J. E. Knox "The Victor Freak"
 
Thanks for the advice on this. I need to get some Deoxit. After I apply it, I will report back.

It is good to hear that the problem doesn't involve the LD3141s. It's interesting that the transistors for the left channel aren't both the same like in the right channel. I don't have a scope but I can solder if I needed to replace components on the board. Thanks for the advice on the path to trace the signal. Now I understand that the circles with letters on the diagram are terminals. It was the main thing I was not understanding about the circuit diagram. Obviously, I am only a beginner but trying to learn enough to keep this deck from turning into a doorstop.
 
Greetings from RojoLand!

"…the transistors for the left channel aren't both the same like in the right channel…" — Huh? The left and right channels are identical. I looked at the schematic and I think I see where you might be confused. You see what looks like two channels because the record amp (at left on the schematic) and the playback amp (at right) are quite similar. What you are seeing at the top of the schematic drawing is the left channel. The two empty rectangles below the left-channel drawing represent the right channel, and that circuit is identical to that of the left channel above (thus Akai chose not to repeat it).

In the service manual, note "VII. COMPOSITE VIEWS OF COMPONENTS" and the layout of the Pre-Amp. P.C. Board (LE-5022). The terminal letters (the ones in circles on the schematic) are hard to see in the copy I have from HiFi Engine, but what you're looking for are letters marked on each circuit trace right by where the wiring connects. Look closely at the board layout and you'll see what I'm getting at here. The top half of the circuit board has the head preamps and output amplifiers; topmost is left channel, then the right. The right-channel components are suffixed 'b' (as in TR3b, TR4b et al). The component locations are also visible on that layout; carefully identify each component terminal as you trace through the circuit. Some study beforehand will save you much frustration when you're working with the deck itself. Patience is a virtue here. You'll get there. Good luck!

Take care,
—
J. E. Knox "The Victor Freak"
 
Thanks so much for taking the time to correct me about that. Like I said, I'm no expert at this. What you explained makes much more sense. I am working from the same Hifi Engine diagram. Its very helpful that you pointed out that the right channel transistor will have the "b" appended to them on the board. Previously you mentioned that the key path is terminal H to T which has capacitors in addition to the TR3b and TR4b transistors. I tried to trace the path and concluded it would include the C19, C26, C28 and C29 capacitors. Does that sound about right?
 
Greetings from RojoLand!

C19 and C29 are the primary ones of concern. C26 is a 100-picofarad (100 pF) part not likely to go bad. C28 is a power-supply decoupling cap; it should eventually be replaced (along with the rest of the 'lytics) but for now don't sweat it; you are looking for what's killing the signal at this point. Worry about a full recap once the deck is more-or-less functional. Before you warm up a soldering iron, did you clean and exercise switches as previously suggested? (Or are you still waiting for the DeoxIT?) If you find audio at terminal 'T' but not at terminal 'M,' the Source/Tape switch is gimp and needs attention. Note too that the two channels have the same terminal letters; that is, there's a terminal 'H', 'T', 'M' (etc.) for each channel. Be sure you're looking only at the right-channel part of the preamp board.

You say you're "no expert at this." What sort of test equipment do you have? (Please bear with me; I am a retired technician and am used to having test gear at hand, and I do realize not everyone does.) Don't worry; it's really not as hard as it may seem. By the time you're done, you will stand back and say to yourself as the machine plays a tape loud and clear, "Hey; I just fixed this!" And that is a great feeling.

Take care,
—
J. E. Knox "The Victor Freak"
 
Thanks for the continued advice. I am waiting for the Deoxit to arrive later this week. When it does my first step will be to address the switches as you described. If that doesn't work then I suppose the soldering iron is next. As for equipment, nothing high end. I have a mid range capability digital voltmeter.
 
Greetings from RojoLand!

You are hamstrung not having an oscilloscope; even a moderately-priced unit would be adequate for the job. But lacking one, if your DVM has a low-AC volts range, you can get by for now (not ideal though, as some of the AC voltages you'll be looking for are quite low level). Another signal-tracing help would be a second, functional recorder of any type (believe it or not!). The mic and line inputs of the second machine can act as "signal-sniffers" when checking for signal in a device-under-test (DUT). Connecting a DUT's test point through a 0.1-µF 630 V film capacitor (value isn't critical) to a second machine's mic or line in (depending upon expected signal level) should give an indication of signal being present or not (audibly, as well as visually on the second machine's meters). Alternatively, any audio amplifier with a line input can be used the same way. (The capacitor is important, to prevent DUT's DC voltages from being applied to the external recorder or amplifier.)

Soldering iron: This job will call for a good-quality medium-sized pencil-type iron (I recommend the Weller WCTP-series units such as this one: https://www.amazon.com/WTCPT-Temperature-Controlled-Soldering-Station/dp/B00004W463). Smaller irons may not work well except on the smallest connections, and larger ones are more likely to damage fragile printed-circuit traces. (And NEVER use a soldering *gun* on PC boards!) I also recommend use of 63/37 "eutectic" solder, or 60/40 if you can't get 63/37. A small-gauge solder is best for PC-board work.

I realize all this is getting kinda hairy, perhaps confusing. It'll all make sense in the long run.

Take care,
—
J. E. Knox "The Victor Freak"
 
Well, no Deoxit yet. It is restricted to ground transportation which has been held up by snow. It's supposed to be here Monday. I am seeing oscilloscopes with a wide range of price and no doubt capability. You mention that a moderately priced one will be adequate. Will a $100 budget cover it? I am seeing some positive reviews from hobbyists on Amazon for some in that range.

Coincidental that you say another deck could serve the purpose through a capacitor. I have a nice Marantz cassette deck but it has a mechanical issue that will be my next project on a different thread. I also have a receiver with a right channel issue and an old Yamaha class A amp with a channel that cuts out occasionally. I wouldn't mind tinkering around with these too. As you can probably tell, I like to fix things.
 
Greetings from RojoLand!

"Will a $100 budget cover it?" Hard to say; it will depend on the specific unit. This is one just like my scope: https://www.ebay.com/itm/254089457279
— it's a Hitachi V-212 two-channel 20-MHz scope and works well for me. Problem is, with this eBay item, one probe is missing and the screen image doesn't show the unit to good advantage (though it is lit up, the trace is way off to one side). Other specimens of this scope are available on the Dock-o'-the-'Bay but for more than $100. Search carefully.

Take care,
—
J. E. Knox "The Victor Freak"
 
Greetings from RojoLand!

The LD3141 ICs aren't the problem here. OP stated he has source audio on both channels, and if one or more of those ICs were borked, that wouldn't happen. Transistors TR3 and TR4 in the right channel (the head amplifier) are suspect, but I suggested that before doing such surgery, he should signal-trace the head amp and find the point at which signal stops; that will isolate the faulty component. Something as simple as an open capacitor can cause the same malady, and there are two electrolytic caps here (C19, C29) that could be culprits. That's exactly what happened to me with a recently-acquired 8-track player; the input coupling cap on the right channel (4.7 µF 16 V) had gone open, nearly muting the audio, and replacing it cured the problem. No transistor surgery required. (A full recap, however, is gonna happen!)

Take care,
—
J. E. Knox "The Victor Freak"
 
Thanks for the continuing advice on this. The Deoxit finally arrived last week. It did a great job of restoring the nearly frozen left channel pot to normal. But applying it to all the other switches, including the record/play one, did not fix the main problem. I took the deck apart and got a look at the 5022 board. There are no obvious culprits. Nothing looks burned up nor are there any swelled capacitors.

I can appreciate what you said about being hamstrung without an oscilloscope. I watched a number of you tube videos of oscilloscope demos and some people doing audio diagnosis mostly on amps. I saw the value of the oscilloscope in identifying root problems which were not necessarily the apparent one at first glance. I am shopping for an entry level option with an eye toward caution as you advised re: Dock O the Bay. One thing I am considering is a PC based unit instead of an independent bench instrument.

Do you think I will be able to test the deck while it is plugged in or will I need to get a signal generator to put a signal in to be tested through the path of the circuit? And if the deck is plugged in, should I have a variable power supply? I have seen those used but they seem to be applied more in situations where a device may have been idle for quite a while.
 
Greetings from RojoLand!

Not sure I'd recommend a peecee-based "oscilloscope." Bandwidth and response is usually inferior to a real one. There's a reason Tektronix digital scopes cost around $3500 or so. (I used a Tek 4-channel digi-scope at work for a few years before I retired. I miss that rig.)

You will need to have the deck powered up (plugged in) in order to test it, of course, but you needn't worry about a variable power supply. That's more a concern if you're working with high-voltage (i.e., tube) gear, or stuff that's not been powered up in some time (and probably should be recapped anyway).

(I call eBay the "Dock o' the 'Bay" because it's often a waste of time. Think Otis Redding. [LOL])

Take care,
—
J. E. Knox "The Victor Freak"
 
Update: Snowblower fixed so back to the Akai. Symptoms are the same as before,

1. Right VU meter maxes out when deck is turned on.

2. The right channel makes a whooshing sound like wind blowing hard for about 30 seconds after the deck is powered up and then it settles down leaving a grounding sound (60hz).

3. Normal sound can only be heard faintly in the right channel with the ongoing grounding sound when music is played

4. Left channel remains normal on playback.

5. VU meters for both channels act normally when a source is input to the deck through line in.

New findings:

1. Probing the right channel Tr 3b base pin causes the right VU meter to jump hard to the right and max out like it does when first turning on the deck. I found that the pinout is a bit different on these transistors (from the front going L to R it is E, C, B) so I am pretty sure I am probing the base pin.

2. Probing the C19b and C20b capacitors on the pin that directly leads to Tr 3b also maxes out the right VU meter but probing on the other pin of those capacitors does not affect VU meter. On the good side of those capacitors, voltage on C19b is same as J6 input (expected given the trace) and C20b is about 0.4 volts. Because the board is not marked and the picture of component items on the board in the service manual does not exactly match my board I think these are the capacitor numbers but not 100% sure about that.

3. Touching the chassis removes most of the grounding sound but regular sound is still very faint in the right channel.

4. Probing at J6 for both channels, where the signal enters the 5022 board, the voltage is very low. It looks like this is the signal coming from the heads so the low voltage looks reasonable.

5. Probing both left channel transistors, Tr 3 and Tr 4, each produce various voltages at the different pins with no problem at the base Tr 3 pin. Except for the Tr 3b base pin, the other right channel transistor pins for Tr 3b and Tr 4b act like the left channel transistors with various voltages. For the purposes of all probing the oscilloscope is grounded to the Akai chassis.

Does the inconsistent behavior of Tr 3b at the base pin indicate that it might be the problem? Because the signal from music is very irregular is this the point where diagnosis would be easier with a signal from a signal generator input to the board?

Thanks again for any help on this.
 
Greetings from RojoLand!

The meters usually jump *a little* when this deck is powered up. Does your right-channel VU meter "peg" if the deck is powered up when in Source mode (the mode in which you apparently have normal operation)? The VU meters get their signal from IC2 (LD3141) outputs, via C35 (22 µF 25 V). That output normally operates at half the Vcc (pins 6 and 7 are noted at 13.2 VDC on the schemo); that's why the meters jump a bit normally on power-up. If the right-channel VU meter pegs in either Source or Tape mode, I think C35 is going leaky* and allowing DC through to the meter. Replacing C35, as well as C36 (10 µF 16 V) should make a big difference (and be sure to do BOTH channels). (*"Leaky" not meaning physically leaking electrolyte [which *could* happen], but instead electrically leaking — passing DC, which capacitors should not do.)

Again, since you have good audio in Source mode, the LD3141 (IC2) isn't the issue with the nasty playback audio. Let's look at your other findings:

I think you got the correct terminal for the base pin. What you describe makes sense. The 0.4 V reading on one side of C20b is consistent with the 0.46 V emitter voltage on TR3b; that's fine. The reason you got VU-meter pegging is that there is a leaky capacitor farther up the circuit which is upsetting the DC levels (and may well be contributing to that power-up pegging too).

Item #4 — agreed. Head signal is normally quite low.

The collector of TR4b should show 13.1 V per the schematic. It connects via C28b (4.7 µF 25 V) to circuit board terminal 'T' and from there to the Tape side of the Tape-Source switch. The positive side of C28b should show 13.1 V but the negative side should be much lower, close to 0 V if not at 0. (There is a resistive path to ground via R23b and R21b.) If there is significant DC present at terminal 'T' then C28b is leaky and must be replaced. As you can see, electrolytics are the cause of many problems like this, which is why it's so often suggested to fully recap these decks. Any DC on terminal 'T' gets passed through the Tape-Source switch back to circuit board terminal 'M' and then via R32b and C30b to the input of IC2 which will amplify that DC and output it to the line out and to the VU meter, which will promptly peg. Does that make sense?

You could disconnect the wire at terminal 'T' and then test-fire the deck. Chances are that the right-channel VU meter would "blip" a little on power-up, then settle down to rest (there would be no tape audio indicated because of the disconnection). That would pretty much clear IC2 and associated parts. But changing out all those electrolytics would be a good thing. If you disconnect the 'T' wire, power-up and *still* get the heavy VU pegging, then the 'lytics associated with IC2b should be changed out (see first paragraph above).

I don't believe TR3b or TR4b themselves are at fault here, as there is amplification going on (DC!), but because they are both 2SC458s they could be a problem later on if not now. My feeling here is that C19b, C21b, C22b, C27b, C28b and C29b (all the 'lytics associated with these stages) should be replaced first. One or more of those caps has definitely gone leaky (C19b, C22b and C28b in particular) and is passing DC, which is upsetting normal circuit operation. Do that first before changing the transistors. If those 2SC458s appear to have "black legs" — that is, the leads are black instead of silvery — then they are at risk and should be replaced (use KSC1845). Compare circuit connections at each pin to the schematic to help identify which pin is which.

When replacing electrolytic caps, be certain to observe polarity. The schematic symbol shows the negative terminal (can) as the larger line that partially surrounds the smaller one.

Take care,
—
J. E. Knox "The Victor Freak"
 
Wow, thanks for such a thorough diagnosis. To answer a couple of your questions 1) the right VU meter does not peg when the deck is turned on and set to source and receiving an input signal from an amp through line in. 2) The transistor legs are black except at the top and bottom of them where they are shiny just before going into the transistor body and at the bottom before going into the board.

I have a couple of short questions about the capacitors.

1) The service manual lists C19b and C21b as “noiseless”. What is a noiseless cap? I have not seen them listed that way on internet searches. If I am looking at the board correctly, those two caps are the light yellow ones with black lettering in the right channel playback part of the board.

2) I assume that even though there are two caps in the right channel playback with the same spec, 47µF 25V, they look different from each other because one is noiseless (C21b; the yellow one) and the other is not noiseless (C29b; gray barrel shaped and shorter with black lettering). Is that correct?

Thanks again for the help on this.
 
So, I probably should have searched other threads on this forum before asking about noiseless capacitors. I can see that the question has been asked and answered.

For those reading this, I noticed what looks like an error in the service manual on page 30 where the capacitor specs are listed. C30 is shown as 4.7µF 25V. But based on the one actually in the board and, though a bit blurry in the circuit diagram, it looks like this cap is 1µF 25V.
 
Greetings from RojoLand!

Good catch on the parts-list error. Schemo definitely sez 1/25 as you note.

"Noiseless" capacitor: I see the parts-list note; don't think I've ever heard of that before. To change out C30 (a 4.7 µF 25 V 'lytic) I would recommend a film cap (Wima makes a suitable part). Any 'lytic or tantalum under 10 µF in a signal path, I would change to film; did that with my Tandberg 10X. Larger 'lytics don't have film equivalents that would fit the available space, or are not available at all.

Good news on the meter not pegging when in Source mode on power-up. That should clear everything after the Source/Tape switch (particularly the LD3141).

Black on those 2SC458s spells impending disaster, but since the black hasn't come up to the transistor case, I wonder if you might be able to pull them, scrape the black off the leads and reinstall them. Could stave off the inevitable, or cure it completely (?). But if you've got 'em out, putting KSC1845s in their place solves the issue and (so I hear) improves sound quality (lower hiss). Your call.

Take care,
—
J. E. Knox "The Victor Freak"
 
You must log in or register to reply here.
Back
Top Bottom