EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: dankinzelman on November 29, 2024, 11:37:24 am
-
Hi there,
this is my first post here. I am a beginner with a little bit of experience repairing and modifying tube guitar amplifiers.
I am currently restoring a Technics RS-M17 cassette player. I have repaired the transport and rebuilt some switches, and now it makes music, but there is some really bad noise in one channel (actually in both, but its much worse on the left so I want to fix that problem first).
I've attached a complete schematic as well as a zoom on the area where I think the noise is originating. I have very little to no understanding of solid-state amplification circuits so please be patient with me.
The heads feed into a preamp IC called the QVITA7122BPB (the chip itself just says TA7122BP), apparently no longer available. I put a scope on the output with the motor muting switch shorted and found the RCA output noise is highly correlated to the ouput signal at pin 6 of the IC. I've attached a sample of the noise, with transport stopped and motor muting switch shorted (ie no signal arriving from the tape heads).
As a first step I tried swapping the ICs between left and right channels and the noise stayed on the left channel, so I presume there is a noisy passive component somewhere in the network surrounding the chip. There is no perceptible noise on pin 2 of the chip, which I presume is the input pin (right?). Shorting pin 5 to ground seems to significantly reduce the noise, but I did it by accident and don't want to damage anything so I've only done it a few times briefly, and don't know if the chip still amplifies with that shorted to ground, just that the noise is significantly lower with pin 5 grounded. I tried swapping C11 with a new part and the noise did not go away.
I'm not sure where to go from here. What should I look at next?
I'm happy to simply follow orders, but If you feel like explaining your reasoning, it would be much appreciated. I'm also happy to read articles/watch videos if you have any relevant suggestions. I'm here to learn.
Also, any suggestions for easy modifications I can make to improve the noise/distortion performance of the player, especially easy drop-pin replacements for outdated ICs or transistors would be appreciated. The transport is in good shape and seems well built, so I think it could be a nice project to try to get it sounding as good as possible.
-
have you check all the supply lines ??
the head wires grounding ?? had some who simply cutted out ver the years
i redid a old high quality Sony deck "Esprit series", redid all the psu capacitors, and the noise floor drastically dropped
if inside you have looong slide switch for record playback etc .... clean them with deoxit .... check pcb groundings screws / posts
-
have you check all the supply lines ??
the head wires grounding ?? had some who simply cutted out ver the years
i redid a old high quality Sony deck "Esprit series", redid all the psu capacitors, and the noise floor drastically dropped
if inside you have looong slide switch for record playback etc .... clean them with deoxit .... check pcb groundings screws / posts
Thank you. I already disassembled the looooong slide switch and cleaned it. I also lost one of the contacts, but it's the one that turns the recording led on, it could have been much worse.
I will check solder joints for the head grounds and pcb groundings. The chassis is floating (no earth ground). Should I install a 3-wire cord and give it an earth ground reference? The case is mostly plastic, metal parts like top cover, face plate and part of bottom cover are connected by aluminum grounding straps. What point should I use as my main ground reference for measuring/checking ground? Negative termanl of first filter cap?
-
take your line in ground as reference, but you should have a good ground line on the pcb, test points etc ...
i did saw once ground shield of the heads wiring cut ....
to my knowledge i never saw the 7122 going bad, mostly connections / solder / the famous rec play switch, and supply issues
-
A very common problem with cassette decks is a dirty rec/play switch.
Best to turn the unit so the switch is vertical, put paper towel under it (switch) and spray a contact cleaner with PPE like DeoxIT. DO NOT GET ANY SPRAY ON THE BELTS, CLUTCHES, IDLER WHEELS ETC. ANYTHING RUBBER.
Then, move the switch back and forth 20 times to scrape off the oxide.
If you don't have contact cleaner, you can move the rec/play switch back and forth by hand a zillion times. Not with power on and a tape in, that will cause erase blips on the tape and wreck any recording.
Back in the day, we would press the record tab sense button up, and move the transport play/stop many times as another way to move this switch.
edit: added board pic, it's the long switch with the connecting arm.
If someone is aggressive with a tape head demagnetizer, the head's output voltage can blow the head preamp IC. This is rare but does happen due to transformer action and the high level magnetic field.
I would not fault the IC until the switch is confirmed clean. In play mode carefully measure the DCV on each pin to look for trouble.
Other things that go wrong (rarely) is the head coax cable breaks due to flexing. But an open-circuit there will crackle of cause hum pickup.
-
Hi everyone, I'm realizing I don't exactly know how to properly check grounding. My multimeter reads only down to 0.x ohms, I presume this is not enough resolution to tell me unless a ground connection is really terrible. I can perhaps use it in DCV mode, which reads down to millivolts, but I'm suspicious of this not being sensitive enough either. So what's the procedure? Should I clip a test lead to a known good ground and go around with the other end and manually ground various components that according to the schematic should be grounded? What's the suggested approach?
Also, in order to mess around inside the unit, I have to open it up and this lets in all kinds of EFI garbage which gets mixed in with the noise I'm trying to fix. Is there a technique or a procedure to handle this? I'm often uncertain if I've made an improvement because taking off the cover instantly raises the noise floor considerably.
-
Why is your workbench so (electrically) noisy? LED lighting? Messy power cords lying all over?
If the head preamp (input) has an open circuit, it will pick up hum and noise. This is normal. You think it's a bad ground but not so likely.
You can carefully move your finger around inside and listen to the hum and noise (you are the generator) and see if there is a sensitive spot.
Like I said, it's likely the rec/play switch making poor connections. You can also gently tap things with a plastic pen to look for an intermittent connection.
Multimeter test leads have resistance and under an ohm is fine.
Just connect (-) to anything convenient - at the RCA jacks, transport frame etc. when you go around making voltage measurements.
-
Like I said, it's likely the rec/play switch making poor connections.
That's been my experience, too.
-
My noisy workbench is probably a combination of LED lighting, overhead transformer for additional halogen lighting and my cheap soldering iron.
As I mentioned, I have already disassembled and cleaned the rec-play switch, and to make doubly sure, I just ran through and bridged the relevant pins with solder from the under side of the board and the noise did not improve. I'm pretty stuck now, poking around with my chopstick and my finger did not reveal the source of the noise, and an LED desk lamp I have which is good at injecting noise did not allow me to zero in on the culprit (just that it's noisiest when close to the IC, which seems obvious since the gain is very high there). I did find one microphonic disk capacitor, but replacing it did not improve the noise. I also re-flowed most of the solder joints around the ICs, still no joy.
Voltages are all about 10% above manual spec, but very consistent throughout on both channels. Resistance to ground on all pins of IC seems to be symmetrical on L and R channels. Any suggestions what to do next?
-
I would try to inject a signal at the MIC jack J1 and see if that makes it in (REC only) to IC1. It might be easier to try that. Use a mp3 player or music source at low level and see if that works.
In PLAY mode, power off, measure ohms across R7/R9 100k I expect the tape head to get connected there but the schematic might have a mistake because the head is shown not connected in PLAY mode. How many ohms did you get?
With power on and unit stopped, carefully measure and write down DCV at each pin on the TA7122's. To compare them and look for problems.
I would probably check continuity with connections to the tape head. They are low resistance and you should be able to see that at the head preamp IC. This would be down stream of the head's shielded cable wiring (look at the back of the head if the connections are OK, you can measure its resistance for L and R coils there) and it's after REC/PLAY switch.
-
Thank you. I am traveling for work ATM but I did measure across the tape head (at the plug-in connector to the board). Each channel shows ~210 ohms of DC resistance. Does this sound reasonable?
I will check your other suggestions when I return home, and measure and write down DC voltages at both ICs. As I stated, I did measure them (without taking precise notes) and found all are about 10-12% higher than stated in the service manual, but both channel voltages match each other very well (the same is true for the pins of the Dolby chips).
-
I had already replaced R7. It reads 99.7k. I haven't managed to find the equivalent resistor on the R channel - the R8 stencil on the board is next to the rec-play switch, with no equivalent resistor nearby.
Voltages are pretty much exactly the same between the two chips, within a few millivolts.
Pin 1 2.05V
Pin 2 0.612V
Pin 3 .006V
Pin 4 0V
Pin 5 0.623V
Pin 6 5.838V
Pin 7 10.96V
Continuity from the board to the tape head seems good.
I tried your suggestion and sent a 440hz sine wave into both mic inputs from a signal generator. I've attached a shot of the oscilloscope output from pin 2 of each IC. Top trace is L channel (IC 1), bottom trace is R channel (IC 2), both at 10mV/division. Seems like slight phase shift, assymetric distortion and loss of signal. Could there be a dying capacitor on L channel? How would I identify the faulty component?
Floobydust, it seems like you have one of these, for my personal curiosity can you tell me where on the board R8 is?
-
Thank you. I am traveling for work ATM but I did measure across the tape head (at the plug-in connector to the board). Each channel shows ~210 ohms of DC resistance. Does this sound reasonable?
I will check your other suggestions when I return home, and measure and write down DC voltages at both ICs. As I stated, I did measure them (without taking precise notes) and found all are about 10-12% higher than stated in the service manual, but both channel voltages match each other very well (the same is true for the pins of the Dolby chips).
Never measure the resistance of a tape head unless you have a demagnetiser to demagnetise it afterwards.
-
Good to know, thank you. I do have a demagnetizer and will demagnetize. Is there a risk of damaging tapes, or is it more about allowing it to return to maximum performance?
-
Good to know, thank you. I do have a demagnetizer and will demagnetize. Is there a risk of damaging tapes, or is it more about allowing it to return to maximum performance?
A magnetised head gives higher replay noise and slightly erases existing recording each replay.
-
OP, it seems weirder because the head preamp IC voltages are all reasonable and the same between channels.
Pin 2 is the input and that sick scope waveform there could be capacitor C1 10uF 16V or others maybe at pin3 that are bad, or the 100R resistor to pin 2 went high value, or the IC has some strange failure.
I thought you can use a multimeter on diode-test and check between pin 2-3 look for an E-B junction voltage drop - but we already see that when power is on.
Pin 1 is a common bias for both channels which can cause oddball problems if that cap below is open.
The IC is only three transistors inside: TA7122 schematic (https://www.petervis.com/electronics%20guides/tc152sd/Toshiba%20TA7122.html)
-
I've done some more digging.
R7 was not actually 100k from the factory, but rather 100R (I mistakenly followed the schematic instead of reading the resistor I had removed). So I think that explains the waveform, once I replaced it with a new 100R resistor the waveform came out normal, but the noise persisted.
The power at pin 7 is noisy, and corresponds roughly to the noise on the outputs. I traced it back towards the power supply. Noise is worse upstream of R27. I can't find R402 on the board, and measure no resistance between C404 and the emitter of Q401, so it seems like R402 is not present in my player (can anyone verify this?).
I replaced C401, 403 and 404 and noise reduced significantly, but is still there, and presents more strongly on the left channel, especially some low frequency intermittent noise, with peaks at 100, 150, 200hz and 2khz. Should I replace Q401? What would be an appropriate modern equivalent?
This noise is present at pin 1 of IC1, but not on the other side of R13. It's also not present on the right channel (or much less so in any case). This made me suspicious of R13, but changing that didn't help either. I can't find R21 on the board, although I do measure ~10k of resistance between R13 and R17. I replaced C7 (connecting to ground between R13 and R21) and C13. Still no improvement.
-
If the 12VDC rail is noisy, I would first fix that. But it supplies both channels with common feeds, so why is one channel a problem? Hmm.
It might be a problem with the voltage regulator (zener D406 MA1150LF 11V? or Q401). I would look at TO-126 2SD882 or 2SC669 etc. but lately I prefer to use a TO-220 part instead (flipped), if Q401 is cooked or damaged.
Noise can also be caused by too much of a load on the voltage regulator. I would look at the voltage drop on R402 1V2? for 36mA?
The switched rail Q403 supplies muting transistors but I can't see it drawing much current if a muting transistor is bad.
R7 (tape head to GND) should be 100k, not 100R that will short out the head and give no signal. Unless it's R1 (in series with tape head) you changed.
-
Thanks for your response.
A few observations:
1. I can't find R402 on the board, and measuring between Q401's emitter and the positive leg of C404 shows no resistance, so I think I have no R402 in my unit. Voltages are a bit low all around (~10-15%).
2. Not sure if this adds any useful info, but the unit takes almost a second to unmute when I press play, sound is initially very distorted as it fades in. Could there be some problem in the muting circuit? It mutes immediately when I stop, but unmuting takes some time.
3. I will look again but I'm pretty sure the resistor marked R7 was a 100R resistor.
4. Can you please tell me a specific part I can get from Mouser to replace Q401? What about a replacement for D406? Sorry, I'm really a beginner and don't want to get this wrong.
-
I guess the schematic does not match. They might have added R402 later. Another way is to see if the rail is overloaded is to see if Q401 runs too hot.
For Q401, 2SC1846S regulator transistor, it has a built-in insulator TO-126 (SOT-32) package which is quite hard to find, I've never seen that and Panasonic has them as obsolete. "TO-126B package which requires no insulation plate for installation to the heat sink".
It's on a heatsink? Then I'd just use an insulating washer and consider BD139-16 (https://www.mouser.com/c/?q=BD13916STU) (cheaper than lower voltage BD135 or 137) or 2SD882.
D406 is MA1150 Zener 15V 400mW. (so it's a +14.3V rail I guess, not 12V).
I would use https://www.mouser.com/c/?q=BZX55B15[BZX55B15 (http://) or BZX79-C15 (https://www.mouser.com/c/?q=BZX79-C15) for $0.10
Check voltage across filter capacitor C401 is more than say 16V I hope, enough extra for the voltage regulator input.
Muting transistors sometimes go bad and cause distortion or a very quiet channel. I'd leave it for now, until the power supply is working OK.
-
...
For Q401, 2SC1846S regulator transistor, it has a built-in insulator TO-126 (SOT-32) package which is quite hard to find, I've never seen that and Panasonic has them as obsolete. "TO-126B package which requires no insulation plate for installation to the heat sink".
...
BD13916STU from onsemi is full epoxy body(no exposed heat sink on the back), doesn't require insulating plate.
-
Q401 has a heatsink, here is a photo. The transistors you are suggesting look like they won't be compatible with the heatsink I have, so I presume I will need to purchase an appropriate heatsink as well? Suggestions?
Also, some more info:
R401 measures almost exactly 2k (in circuit) and I measure 18.6V on the diode side and 15.4V on the downstream side. So I guess the regulator transistor is not getting enough voltage to operate? Is this a reasonable current draw?
Also I presume there is some uncertainty measuring resistors in circuit due to interaction with other components. In this case it's parallel with the collector and base. Do I need to lift a leg, pay attention to DMM polarity, or is there something I can do (discharge caps, etc) to improve accuracy? Or is this close enough for our needs?
I have been unable to locate R7 or R8 anywhere on the board and forgot to measure resistance to ground from the play head. As it turns out, on closer inspection, the 100R resistor I changed (and thought was R7) was actually R9. Both R9 and R10 are 100R (as seen in the schematic). I have taken some photos, the board printing has R7 and R9 at the same spot, and on the other side of the rec-play switch, I see R2 and R8 printed very close to each other, but I see no 100k resistors nearby.
-
Q401 has a heatsink, here is a photo. The transistors you are suggesting look like they won't be compatible with the heatsink I have, so I presume I will need to purchase an appropriate heatsink as well? Suggestions?
...
Sorry for the late reply, here's my take: re-use the original heatsink, attached with a screw(M2.5-M3) and heat transfer compound to the new transistor; don't forget about transistors different pinouts.
-
The Q401 you have looks different than the transistor in the RS-M17 posted schematic. A little cheaper.
It has a clip-on heatsink that slides up and off, it's put on a common transistor in a plastic TO-92 package. Easy to trash the leads pulling on it. No need to yet...
At R401 you measure 18.6V and 15.4V which tells me that is fine - the zener is alive just over 15V.
So what is coming out of the Q401 regulator? Measure at R402 (both sides) or C404. R402 looks 100Ω not 33Ω on the schematic.
Resistors rarely fail (unless in a power circuit) and lifting a leg to test them can do more damage to a board. I suspect most resistors last. You can troubleshoot quite a bit without doing that.
Low value parts you can check in-circuit like R402 33Ω or 100Ω - know you should never read higher than that resistor's value unless there is stray voltage on a capacitor somewhere.
R2 is the 10Ω part coming off the tape head. I see BRN-BLK-BLK-GOLD which is 10Ω not 100Ω. Check your colour codes you might be mixing them up.
-
Thanks, I'm traveling for work now but will check when I get back.
I can't read resistor color codes (without a cheat sheet), I usually just measure them. R2 is indeed 10R, the resistor I was speaking of as 100R is R9, which is between C1 and pin 2 (the schematic is missing the R9 label). I am attaching another version of the schematic here, this one is from the service manual.
As I mentioned, I have not managed to find several resistors on the board, the photos were intended to show where the 'missing' resistor names are printed on the board, but are not actually present. R402 is among the resistors I have yet to find. Do any of you have one of these units to look at?. When I get back home I will see if I can at least measure 100k between C1 and ground, and 33R between collector and C404's positive terminal. What does (1V2) mean next to 33 on the original schematic I posted? 1/2 W?
-
I suggest replacing all tiny electrolytic capacitors, generally they don't last long compared to larger ones. As of noise, could be something sensitive powered through resistor + smoothing capacitor with that capacitor being dead.
-
I see no R402 (on the output of Q401) on your board pic. You'd have to find somewhere else to measure the (guess around +13.7VDC) rail, at C404 or on R27/C13 by IC1.
Right now if the voltage is too low or unstable, Q401 could be damaged or overloaded. I've seen Sony let these parts run very hot for a short lifetime, intentionally would not heatsink and repair shops knew their games.
The regulator also powers the LED VU meters IC502, so they would add load if the LEDs are all on I think. So the rail should be stable even if the VM meter LED's are blinking.
One thing about this board is the use of carbon-print resistors, a little black rectangle on the PCB that makes a resistor. This can be extra confusing as not all resistors are through-hole parts.
-
Wow yes, I had no idea about carbon print resistors. I've spent literally hours looking for those resistors, and they're just those little black squares on the board?
Anyway, I measured R27 to be 268R, and there are 12.4V before and 10.7V after it. Is that too much current draw? Upstream is the same as on schematic, but voltages at IC 1 and 2 are about 10-15 percent lower than on the schematic. I checked again for R33 and resistance between emitter and R27 is 0.5R, so I think R402 doesn't exist. After disconnecting the tape heads, I measured 91.7k for R7 and 95.5k for R8. Is it worth scratching through them and replacing them with new, through-hole parts?
Also, I noticed that the rail gets really noisy when I activate the transport, and the noise appears on the scope together with the noise on the headphones. At idle, the upstream side is fairly smooth (with headphones muted), I've attached a video where you can see what happens when I press play. Could the headphone amp be drawing too much current or something? Should I try disconnecting it?
Here's the video link https://youtu.be/q1yeR-zryws
Top trace is before R27 and bottom trace is after R27, at maximum scope gain, 1mV/div. I tried disconnecting the motor to see if that helped and it becomes very noisy even at idle with the motor disconnected.