1980 Soviet Radio Cassette amplifier problem.

[73's de Edd]

Sir Richard 9025 . . . . .


Here is the audio flow path starting at the tuner as RED arrows and then after the RADIO-TAPE switch, which I have now being "switched" to RADIO position, where the signal then proceeds as BLUE arrows..
Walk your good injected signal progressively backwards from C14 to see where differences onset.



73's de Edd
.....

 

[Richard9025]

Walk your good injected signal progressively backwards from C14 to see where differences onset.

At every point I inject, there are no differences. If I insert the audio at point 9 it sounds very well, and the tone control works very well.
I measured every resistor and capacitor (by inserting audio thru it and see if it distorts the sound) and there aren't any problems... So the yellow box is perfectly fine.

Because I get good audio when injecting from my phone anywhere, and also good audio if I hook the audio signal to another amp, I think that there is a problem regarding the audio levels.

Your input signal and hiss is too strong so I would highly recommend reducing the gain by adding to 150 to 220 to R32(56 Ohm) this reduces the gain so you can get cleaner volume.

I added a 200Ω resistor in series with R32 and no difference.
600Ω makes the distortion appear at a higher volume (70% instead of 50%) and also significantly decreases the loudness of the sound.
At 1.2kΩ the distortion is present only after 90% but the audio output is very low. So low that 100% volume on this unit is like 20% on my other radio...

The audio signal may be strong, but also maybe weak? but also if I grab the signal and run it thru a separate amp, it sounds good.....

There definitely is some discrepancy between the audio levels

Look at the red squares that indicate the audio level in mV:


How do I measure the audio levels in mV? with a multimeter in AC mV range? where do I connect the probes?

Conclusion: The audio getting in the amplifier is badly distorted by it, but it sounds very good thru another amp. The amplifier is distorting the audio signal from the radio, but if I inject sound in it, it amplifies it extremely well...

 

[73's de Edd]

Sir Richard9025 . . . . .

Looking at post 45's schematic you show the " X " breaking connection and confirm that injection of audio into C14 produces flawless sound at higher sound levels. You also seem to infer that the radio signal at the volume center tap is also good when you feeding that into a separate audio amp . .WITH A SERIES DC BLOCKING CAP . . ..
providing DC isolation. Now is this other signal tracing / evaluating amp being within the other older version of this radio ?
How do I measure the audio levels in mV? with a multimeter in AC mV range? where do I connect the probes?
Of course the RED box specs indicate that you will be experiencing a 2-3 times attenuation by the inherent loss produced by the passive tone control circuitry.

You CAN guesstimate *** at those AC millivolt voltage levels if you are using a sine test signal with the DVM metering in AC millivolts range and using a series DC isolating capacitor. ( *** Because each circuit is reflecting from a different impedance level.)
BLACK meter probe to ground . . .RED meter probe to the point of the pure sine signal being tested / traced.

Now I really want to physically eye C14 capacitor . . .but I can't . . .any chance that it is being one of the four sickly green rectangular block type of caps, that are being visible in your Post #1 on the next to last photo ?

Which would not be an electrolytic, but a paper . . .possibly poly . . . of 0.15 ufd at a 200V rating.
If you have another .15--.22-.47 , can you sub it in, after testing for it having no DC leakage thru it.

73's de Edd
...

 

[Richard9025]

Now I really want to physically eye C14 capacitor . . .but I can't . . .any chance that it is being one of the four sickly green rectangular block type of caps, that are being visible in your Post #1 on the next to last photo ?

C14 is a 1ufd electrolytic cap

If you have another .15--.22-.47 , can you sub it in, after testing for it having no DC leakage thru it.

With C14 replaced with a new: 0,1ud/250V poly, 150nf and 470nf polys and with a 47nf ceramic, the results are all the same, so C14 clearly isn't a problem.

Of course the RED box specs indicate that you will be experiencing a 2-3 times attenuation by the inherent loss produced by the passive tone control circuitry.

You CAN guesstimate *** at those AC millivolt voltage levels if you are using a sine test signal with the DVM metering in AC millivolts range and using a series DC isolating capacitor. ( *** Because each circuit is reflecting from a different impedance level.)
BLACK meter probe to ground . . .RED meter probe to the point of the pure sine signal being tested / traced.

Ok so here is how I measured the audio levels:



The measurements look ok, I inserted the sine wave and got good results.
But how about measuring the audio level that the radio outputs at point 29? (I don't know and I don't think I have the tools...)

You also seem to infer that the radio signal at the volume center tap is also good when you feeding that into a separate audio amp . .WITH A SERIES DC BLOCKING CAP . . ..
providing DC isolation.

Exactly

Now is this other signal tracing / evaluating amp being within the other older version of this radio ?

Nope, it is another radio
The audio is inserted via the phono input (just a 50kΩ input to the amplifier)
The separate amplifier schematic:

 

[Sunnysky]

1) Measure the AC voltage across C35 500uF when output is 900mVac, Any mV here is a loss in output power. This can be increased to x mF. Use battery V- as Gnd for AC measurements. ( more later)**
2) Also measure DC R of speaker. It should be ~ 1/2 of rated speaker impedance. e.g 4 ohms(DC) for a 8 Ohm speaker or 2 ohms for a 4 ohm speaker.
A lower speaker impedance can generate more power but often at a tradeoff of THD

however depending how thermally coupled the output transistors are.... when output power increases in audio, Vbe rises and so does crossover distortion since the Vdc is fixed at 1.1V difference of 5.0V(23)-3.9V(20)=1.1V = 2 Vbe diode drops at low current.

This can be increased if possible with R38 at the expense of idle current and a noticeable drop in 9V battery to 8.7V which may or not be ok. ...10mA results in low drop 50mA larger drop.

** The choice of output DC blocking Cap is critical for low ESR.
here are some good choices for 9V operation

 

[Richard9025]

1) Measure the AC voltage across C35 500uF when output is 900mVac, Any mV here is a loss in output power. This can be increased to x mF. Use battery V- as Gnd for AC measurements. ( more later)**

Exactly 100mV AC across C35.

2) Also measure DC R of speaker. It should be ~ 1/2 of rated speaker impedance. e.g 4 ohms(DC) for a 8 Ohm speaker or 2 ohms for a 4 ohm speaker.

Half? I found on the internet:

Typically, the resistance reading should be roughly 15% less than the nominal impedance on the label.

With the 8Ω speaker disconnected, I get 7.3Ω resistance across it.
With another 4Ω speaker, it measures 3.2Ω.

** The choice of output DC blocking Cap is critical for low ESR.

That means that the original cap must have been a low-esr one? Good thing that I still keep the caps. Shall I sub in the original one to see what it does?

Look at post #19, where duke37 told me to remove T9, T10 and C35 and connect some headphones in series with a 100Ω resistor and a capacitor. The result was the exact same distortion... So I guess that these aren't a problem.

 

[Sunnysky]

Exactly 100mV AC across C35.

Half? I found on the internet:
With the 8Ω speaker disconnected, I get 7.3Ω resistance across it.
With another 4Ω speaker, it measures 3.2Ω.


That means that the original cap must have been a low-esr one? Good thing that I still keep the caps. Shall I sub in the original one to see what it does?

Look at post #19, where duke37 told me to remove T9, T10 and C35 and connect some headphones in series with a 100Ω resistor and a capacitor. The result was the exact same distortion... So I guess that these aren't a problem.

OK Thankyou for the quick results. Good luck. I missed that. Without a scope, one needs a high impedance listening device to sample each stage to see where distortion occurs. From my simulation, your DC values seem near normal range, so I assume your pots are set to optimum.

 

[Richard9025]

I disconnected C14 to test the audio coming in the amplifier, and it sounds perfectly good...
Maybe the amplifier is over-amplifying the signal?
I'll try to get a scope, they're kinda hard to get. I hope that we can find the problem without a scope.

I recorded how I gradually increase the volume, maybe you guys can identify the type of distortion?
https://vocaroo.com/i/s1rCWbrH45LQ

Thanks!

 

[73's de Edd]

So you had C14 out of circuit and that stopped the radios audio flow thru the power amp. So the circuitry from the rotor of the volume control is now feeding your different audio test amplifier.
That present test audio could now either be an FM or AM signal and my ears are hearing a compressed mushy sounding audio at the highest level you recorded..
If your units amplifier amplified that a bit more than your present test amplifier is doing, that could sound quite a bit more choppy.
We now have two more audio source tests to evaluate, the AM radio, considering the fact that this units FM detector is using a FM discriminator circuit which does not incorporate the pesky electrolytic, that the ratio detector circuit does.
A rat det summing capacitors deterioration, gives rise to distortion and lowered audio output.
The AM detector is almost bulletproof in its audio quality, plus you ADDITIONALLY have the option of using a cassette tape audio source.
See how the other two audio options sound like.

 

[Richard9025]

That present test audio could now either be an FM or AM signal and my ears are hearing a compressed mushy sounding audio at the highest level you recorded..

The audio from the previous post was the radio on AM, with the tone at max and bass at minimum. (thru its amp, not another one)

Listen to this recording (I gradually increase up to 100%), it is taken with the radio on AM, with the base at 50% and tone at 50%, it should sound perfect but it sounds... take a listen...

https://vocaroo.com/i/s1wtSwU8LFk7
The final second is at 100%, you can't notice how loud it is, it was so loud that I immediately stopped it so it wouldn't "explode".

plus you ADDITIONALLY have the option of using a cassette tape audio source.

I tested a tape at 100% volume, then at 80%. Also, I put the tape on another much newer radio for comparison:

100% - https://vocaroo.com/i/s1BscRf2eOL5
80% - https://vocaroo.com/i/s02gqDCLKq0n
On new radio, for comparison: https://vocaroo.com/i/s1g1YEJjQlyd

You can't notice, but on tape, it is MUCH quieter than on radio, on radio is like 3 times more louder.

When playing the cassette, I saw that this communist radio doesn't like capitalist cassettes:

I luckily managed to unscramble the cassette and put it back.

 

[Sunnysky]

I've lost track what you are trying to accomplish because you cannot get more power when it is clipping with too much gain. You can only get more power with a lower impedance speaker. That's Why I said reduce gain or just dont go past clipping. try an external speaker.

 

[73's de Edd]

Can you provide a 4khz sine for testing ?

 

[Richard9025]

try an external speaker

With this big speaker unit (on the back it says 8Ω 40W, it measures 15.4Ω) it does the same, take a listen:
https://vocaroo.com/i/s1ZbznRFFs9t
The speaker unit:

Can you provide a 4khz sine for testing ?

Yes
I inserted the 4khz sine wave, 7mV into point 29:
(the red X)

Audio results (phone on top of the original speaker)
https://vocaroo.com/i/s0u1cy0KBY89
I can't notice if it distorts the wave or not?

Reading on a russian forum, some guys talk about this radio and say the following about the transistors:
"Probably half-dead transistors in terms of gain and reverse current "
"We have a transistor ... when a multimeter passes, it behaves normally, when measuring hFE too, but in the circuit, it does not work - that's a half dead transistor"

 

[Richard9025]

I think I traced the problem.
First thing I did, is to replace every transistor in the amplifier section, to clear the idea of bad transistors:

The results were... the same. (I labelled every old transistor that I desoldered so I know which is which)

Next, I inserted audio thru point 29.
As usual, it sounded excellent and loud.
But, I increased the volume level of my phone to the max.
What I observed was the exact same distortion that the amplifier produces when connected to the radio...

Also, I connected a DMM in series with the batteries to measure the current draw. I saw that the louder the phone audio signal is, the higher the current consumption is. This explains why the radio draws so much current when it is distorting the sound.
So one thing is clear after these tests: The input signal of the amplifier is too high.

I also changed VT12 with a BC557B to see if it changes anything and it doesn't.
One thing to note is that the distortion is present on AM and on FM wavebands.

 

[Richard9025]

Measured some voltages present next to VT11/VT12. They're 0.5V less than the value on the schematic.
I also replaced VT11 but the results are the same.

full unedited schematic download (jpeg,2.4mb) : https://lensdump.com/i/8RR0x2

 

[Richard9025]

Hooking the signal that comes after the detection & preamplification with T11 and T12, (point 29) to another amp, I get quality audio only at low volume, if I crank up the volume, it goes into clipping and makes the same distortion as the original unit's amplifier does.
So the signal present at the input of the amp is too high, causing it to clip and overdrive the audio.
I can improvise an oscilloscope using the PC's soundboard.
What can I do next? Awaiting your replies.
Many thanks!

 

[Sunnysky]

What isn't clear to me (no pun intended) is the harmonic distortion from non-linearity at various stages when distortion occurs and confirm it is not from clipping or saturation ( by audio coupling high impedance to external headphone). Once the node where distortion occurs at a node then parts connected that node can be suspected for non-linearity , voltage controlled leakage, etc.

The efforts to describe the efforts, but the skill to isolation the fault requires a scope to make it easy, and must be frustrating when normally it is not this hard to debug.

I would use a similar method of sound injection and sampling with a high impedance pre-amp and speaker or headset if I did not have a scope, to zero in on the faulty node..

If both Left and Right , that ought to make it easier to locate the common fault.


I assume you are not expecting it to be louder than it should because any radio can distort when the input signal is too high.

 

[Richard9025]

I assume you are not expecting it to be louder than it should because any radio can distort when the input signal is too high.

I'm expecting it to be louder when operated on cassettes (it uses a separate preamp, with one IC and one transistor) , not the radio of course, on radio it's overdriving the speaker

I would use a similar method of sound injection and sampling with a high impedance pre-amp and speaker or headset if I did not have a scope, to zero in on the faulty node..

So you say to use a high impedance input, (oscilloscope is perfect for this), like some old headphones or a separate high impedance amplifier.
Unfortunately, I only have some generic pc headphones (they have an impedance of ~30ohms).
I don't think I have some high impedance amps, only amps from radio receivers and car amplifiers.
I think I can make one using some audio transformers and some driving circuitry?

As a side note, I measured some components in the preamp stage, they all look good:



To clarify some things:

The problem of this radio is that at volumes over 15%, it begins to distort the audio, at 50% its very bad, the speaker has so much base present that it just makes aggressive back and forth moves (like a trampouline), and extremely high volume. (on AM and on FM)

If I hook audio at its input, it amplifies very good the audio (and loud), but if I increase a lot the audio input, the amplifier makes the exact same distortion as above.

If I hook the audio present at its input to another amplifier, it does the exact same distortion at high volumes...

When playing cassettes it doesn't distort the sound, only the head needs cleaning (along with pinch rollers etc)

So either the amplifier is not working as it should, overdriving, with the audio present at its input (which I don't believe, because, with 3-4mv at its input, it is outputting 900mv, as it should) either the signal present at its input is too high, overdriving any amplifier.
If I attenuate the input signal with a potentiometer, will it do something?

I'll go tomorrow to see if there are any oscilloscopes left in an old deposit.
Many thanks Sunnysky and 73's de Edd !!

 

[Richard9025]

Good news, I found a scope (0-10MHz), and it works perfectly (all original components from '85):

I'll take its probes later, they are in a box and I didn't have the time to grab them. (It works perfectly for signals in the 20-20.000hz range without any probes)

Now I need advice on where to use it and what transistors to pull out, etc.
I'll start out tomorrow by inserting a 1khz sine wave to the audio amp and connecting the oscilloscope to the speaker terminals to see how it distorts the signal.
Many thanks! Awaiting your replies.

 

[Richard9025]

I'll start out tomorrow by inserting a 1khz sine wave to the audio amp and connecting the oscilloscope to the speaker terminals to see how it distorts the signal.

Did that:
I inserted a 1khz sine wave in point 29, from a tablet at 70% volume, the oscilloscope is connected to the speaker leads, the speaker is disconnected:

Volume pot at 50%:



At 70% :


At 80%


100%



This surely is audio clipping. Over 70% it begins to clip and at 100 it does it so badly that it becomes a square wave...
The scope is set at 2 AC volts per division.
Its output should be ~700mV with an input of 3-4mV at 400hz
The tablet, even though it is outputting a 1khz signal, at 70% I'm sure that it produces a much greater than 3-4mV signal.