Refurbishing a vintage Technics SA-5370

PS -

Missed this earlier:
Yesterday evening, while taking a look at the tone control resistors, I noticed that almost all of them were off the specs by margin bigger than the suggested +- 10% . Could this be the source of the tonal unbalance? Do the carbon resistors degrade with time?
They do in fact have a tendency to drift upwards with time (if not as bad as carbon composition). A reduction in value is observed only when severe overloading has occurred, which is unlikely in this case. Maybe verify one out of circuit.

So no modern Yageos then after all. Those would make a decent replacement though.

BTW, I have done some noise analysis, and the results are not spectacular but adequate. Estimated output noise level is ca. 270 µV, +/-, or ~65 dB below 50 mW/4 ohm. This is actually a good value for the late 1970s - even around 1980, you would still see a lot of devices that would post 59, 60, 61, 62, 64 dB. Only a few exceptions with different gain distribution would post 72 or 75 dB.
 
The OP say in your first post :

" Hi,

Two weeks ago I purchased a vintage Tecnics SA-5370 in great condition. Minimal usage signs, all original under the hood and everything working as it should.
It is a unit produced in '77 as the hiper popular Marantz 2238 and 2252 that have similar specs just its sels for one fifth of their price.

Last weekend I replaced the electrolytics on the psu, pre and power amplifier and now it produces extra clean sound. In the next opportunity I plan to replace the elcos also on the tuner and phono board and to replace the rectifier with four ultra fast diodes."


Here we are again then, in the same place at the beginning?


As someone well mentioned, the experience should serve as something, right ?


I am not against reconditioning of any amplifier, but against snake oil.

What I see wrong:

First it was to replace capacitors just because we suspect that the new generation will offer audible improvements, then it was magic cables, and now it would seem that any resistance that moves away from the predetermined tolerances in the circuit is to blame for all evils.

Well, each one spends his time and money as he pleases, that's a fact.
 
Last edited:
What I see wrong:

First it was to replace capacitors just because we suspect that the new generation will offer audible improvements, then it was magic cables, and now it would seem that any resistance that moves away from the predetermined tolerances in the circuit is to blame for all evils.

Well, each one spends his time and money as he pleases, that's a fact.

I find replacing the electrolytics logical thing to if you want an 40 year amplifier to work according its original specs. If you were a bit familiar with the inside of the SA-5370 you would have noticed that the power supply and main amp. capacitors are situated just under the old style lights (they didn't had LEDs back than) that produce considerable amount of temperature. Literary all the original elcos I took out had a yellow burn mark around the bottom. I didn't found any leaked capacitors because they were all dried up.

Liking or desliking the specific sound of any audio equipment is indeed very subjective matter. Maby the tonal unbalance that I hear from the unit that is the subject of this tread is only in my head. But where did you see the snake oil in all this? Yes I am a bit adventurous, propobly because my knowledge in electronics is very limited. I came here asking for help from the people that know a lot more than me in order to solve my problem, to learn a thing or two and most important to have fun.

Criticism is ofcourse welcome, if you have the urge to get rid of your negative energy please do it at least in a constructive manner.

Totally Friendly,

B.
 
The only reason for putting in a 22k that I can think of is that someone had to replace the bass pot. (Do bass and treble pots look different?) This is a 2x 100k reverse log pot (100kC), and optimum FR flatness occurs at a wiper setting of about 0.1475 in "normal log" terms, or -16.62 dB. I don't know much about antilogs, but from what I've seen of normal log pot measurements, midpoints tended to be around -26 dB, or 0.05. Guess where the flat position with the 22k is? Yep, 0.05.

Still, it should have worked like this. Maybe they did put in the wrong resistors at the factory after all? Otherwise, why would someone even touch these? Putting in modern metal film resistors? They don't look like modern-day blue 1% Yageos, do they?

Now on to the misbehaving measurement setup... OP, what kind of computer (motherboard) and sound device even is this? The results sure look lousy. Those are not from using the headphone out though, are they?

Sgrossklass you are the Man, I really appreciate your input and working up all the theory behind my issue with the Amp. The graphics that you made do show what I have heard befor and after replacing the capacitor. Yes I know 1uf is to much, I hear it 🙂 but that is the smallest value that I have at the moment.
I checked the pots and they seem original as does the 22k resistor. Finding a mistake from a Japanese manufacturer is like finding a needle in a stack of hay. Who knows, I might have found one after 42 years.

I tryed to take the measurements with an old bitten up Dell that my wife bought in Brazil as a second hand some years ago. Long story short, it's simply not capable of taking the needed measurements.

I've ordered replacement resistors for the two positions that are off, together with replacements according the service manual specs for all the capacitors from the tone controls circuit. If all goes according the plan I'll come back with the result somwhere during the weekend.

Gr.
B
 
I find replacing the electrolytics logical thing to if you want an 40 year amplifier to work according its original specs. If you were a bit familiar with the inside of the SA-5370 you would have noticed that the power supply and main amp. capacitors are situated just under the old style lights (they didn't had LEDs back than) that produce considerable amount of temperature. Literary all the original elcos I took out had a yellow burn mark around the bottom. I didn't found any leaked capacitors because they were all dried up.

Liking or desliking the specific sound of any audio equipment is indeed very subjective matter. Maby the tonal unbalance that I hear from the unit that is the subject of this tread is only in my head. But where did you see the snake oil in all this? Yes I am a bit adventurous, propobly because my knowledge in electronics is very limited. I came here asking for help from the people that know a lot more than me in order to solve my problem, to learn a thing or two and most important to have fun.

Criticism is ofcourse welcome, if you have the urge to get rid of your negative energy please do it at least in a constructive manner.

Totally Friendly,

B.


Uhhhh, this gets hard .....🙄

I go in parts, (nothing more than for you not to believe that "the silent one grants"), I will take a little of my time to respond.
First and foremost, notice your contradictions in your comments, who is the one who needs to break free from your negative energy?
If you ask questions, you have to accept the answers, ask if you do not understand them, or continue on your way until you realize that you are wasting time taking wrong directions.
That kind of comment denotes that you have felt upset, I'm sorry, but the negative is ultimately you.
Are you throwing the first stone and then you say "totally friendly"?
Snake oil is attributing - as you admit - problems of subjectivity in your listening to an insignificant difference in tolerances in the value of resistors, source filtering capacitors, etc.
Absurd. You know more than those who designed the amplifier to question values ​​without tone or are they? You don't understand electronic, (- do you know what the value 1.41 means in a PSU?) - such a small difference in the tolerance of a tone control's resistance can never cause audible differences. Yes the capacitors do it in the path of the signal, given that they block or not, - depending on their value - the passage of the alternating currents (they are the audio ones, as they are the ones that you have surely seen in the post attached (which I have not read, because I know what it is) in a Baxandall tone control, which are typically used in these topologies of the time of analog controls, but should not be modified in their values ​​(mainly XC, capacitive reactance ) because you modify the behavior in a catastrophic way, you can make the treble very bright or turn them off completely due to a comma error, for example 0.01 Microfarads instead of 0.10)
And besides, it is very difficult to give you useful advice when you do not name for example the speakers that are attached to your amplifier!
Lucky with your experiments, I've done some, (very few, although I made many circuits already designed by qualified electronic engineers, I don't pretend to be, I'm just a practical technical scholar) because I first studied the theory and understood the meaning of what I was to do.
If it doesn't work for you .....

PD: I'm going to have a little disharmony here, thanks for giving me the opportunity.
On one occasion an upstart company acquired a very complicated machinery to improve the performance of its products, which stopped working and caused them significant damages in profitability. So they called a technician who solved the problem in five minutes.

Very happy, they asked him what the fault was and how much they should pay him for his services.

Yes, it was a loose screw, it's $ 10,000, he told them.
How ? So much money for five minutes of work? For adjusting a screw? Give us a well detailed invoice please!

The bill read:
For adjusting a screw of the machinery 1 U$S.
For the knowledge of which screw to tighten, U$S $ 9999. 😀
 
Last edited:
All squabbles aside (which I find to be (a) fruitless and hence (b) pointless), I think I just found something:

Remember the misplaced resistor, 22k instead of 120k? Well, I found that frequency response will be flat for identical positions of the bass and treble pot (0.18 each) with a value of - 220k. The plot thickens. 22k instead of 220k does sound like a plausible mixup.
So then both the service manual and the actual resistor value may be incorrect after all!

OP, please measure bass / treble pot resistance at midpoint (as outlined earlier).

BTW, I cross-checked with other models of this series, and they use different part values altogether.
 
All squabbles aside (which I find to be (a) fruitless and hence (b) pointless), I think I just found something:

Remember the misplaced resistor, 22k instead of 120k? Well, I found that frequency response will be flat for identical positions of the bass and treble pot (0.18 each) with a value of - 220k. The plot thickens. 22k instead of 220k does sound like a plausible mixup.
So then both the service manual and the actual resistor value may be incorrect after all!

OP, please measure bass / treble pot resistance at midpoint (as outlined earlier).

BTW, I cross-checked with other models of this series, and they use different part values altogether.
 
Sorry, I was busy replacing the caps and the resistors the last days. The tonal balance is back to normal now.
No there are no mistakes in the service manual or in the manufacturing process, ill explain the specifics once I find some time.
While I had the amp still opened I also did some small mods by bypassing couple of capacitors and I did replaced the rectifier after all.
I'm letting it burn the new parts in for now and I'll come pack over a week or something with proper feedback.
 
Sorry, I was busy replacing the caps and the resistors the last days. The tonal balance is back to normal now.
No there are no mistakes in the service manual or in the manufacturing process, ill explain the specifics once I find some time.
While I had the amp still opened I also did some small mods by bypassing couple of capacitors and I did replaced the rectifier after all.
I'm letting it burn the new parts in for now and I'll come pack over a week or something with proper feedback.

I ............. But where did you see the snake oil in all this?


😀
 
Finally found some time to leave a feedback on the now completely refurbished SA5370.

As I mentioned I replaced R505/6 and R513/4 which were showing bigger discrepancies than the marge of 10% and all the polyester tone controls capacitors with 5% Panasonic metalized polypropylens.
While changing the resistors a weird thing happened. R513/4 were original resistors which wile on board were measuring 22k but when I took them out they were measuring their original 120k. Is anybody familiar with this kind of phenomenon, because I am not. Did the old soldering caused such behavior or did the original resistance got reactivated by the application of heat?
How ever they are now replaced, and as it appears the amps tonal balance is back to its original curve.
At this point I was happy with the development of the project. The next day I also replaced the old rectifier with four 15A Fairchild "stealth" technology (fast switching, soft recovery) diodes. The tonal change after this replacement at first wasn't huge but I immediately noticed that all the sharp edges around the midrange and high frequencies were gone.
Now after a 10 days and and close to 100 work hours I feel that the receiver is still settling in but the the sound it produces has taken its final form in very pleasant way.

In order to have a reference I took out from the basement the Marantz 2238 that also waits for its turn to get reconditioned (in generally it's very bad condition discourages me every time to make the first step) .
In generally the Marantz is heavier mostly because of its chassis and trafo which actually isn't bigger than the one in the SA but it has a big metal shielding around it. The Technics is engendered in very smart and economical way in terms of space and costs. One example of this is that the shealding to its trafo is provided by the chassis bracing and by one of the transistor cooling bodies on the other side. The top cover on the Technics is nice and made of plywood and the bottom cover is made of cheep thin pressboard (which I replaced with 4mm HPL for extra weight and stability) whereas on the Marantz boath are metal.
Other essential differences between these two receivers is that the Technics , in contrary to 2238, is direct coupled and all the electronics lay on two PCBs whereas in the the Marantz the PSU, preamp, phono, tuner and main amp are partitioned on different PCBs.
Qua features, I find the tuner sections equally good, just the SA doesn't have a Dolby filter and the stations airing in Dolby will sound brighter (I use the luidness buton as compensation in such cases ) . Where the 2238 also excels is the tone controls, it's midrange is also controllable whereas the Technics has only controls for the bass and the treble.
Sound wise in their original state they are very different. The SA is lean, transparent generally on the bright side of the neutral with very good resolution. The Marantz tends to sound on the warm side also transparent but with a lot more texture and less resolution.
When I started refurbishing the Technics my goal was to tilt its sound signature towards the wormer side, to add some texture but to retain its transparency and resolution. I think I achieved this by the combination of using a mix of Elna silmic II and Nichicon FG electrolytics ( for C902 a used a Nichicon muse bipolar bypassed with a Vishay MKP 18370.01uf ), Polypropylene capacitors for the tone controls and the rectifier replacement.
Now the tonal balance of the SA 5370 is almost on pair with the 2238. The used combination of parts adds some smooth texture in the midrange that wasn't there before. The soundstage is now darker but still very transparent and the resolution stayed unchanged.
One last change I made is replacing the old type lights with some LED type. This reduced the heat in the insides a lot.

All in all I would say "mission accomplished" and therefore I thank all of you guys for sharing your thoughts regarding my tonal balance problem.
 

Attachments

  • IMG_0022.jpg
    IMG_0022.jpg
    349.2 KB · Views: 259
Last edited: