Troubleshooting white noise with oscilloscope

Hi, one of my Yamaha CR620 developed some minor white noise on the left channel. You can acknowledge its existence only when nothing is playing, otherwise it's completely covered by music.

I hooked my scope to che outputs and tried to see if I could recognize the noise, but the only difference I've noticed is that the left channel shape shakes slightly, while the right channel one is firm. Is the movement a representation of the noise? Is there any specific strategy to find the source of the noise?

Thanks

Andrea

At times there will be a slight 'shake' on the display. Check your connections. Reverse X & Y on the scope. If it follows check your probe. Double check that all connections are tight including Ground @ the scope & amp.
Vertical shake on the scope display? What are your scope settings? AC or DC coupling?
Measure DC offset at the speaker terminals.

If you have Pre Out/Main In:
To check the pre section get a pair of BNC male, RCA female adaptors. This will convert your scope BNC probes to RCA so you can connect your RCA cable to the Pre Out terminals and scope X or Y.

You're dealing with very small signals and may or may not see it on the 'scope. By looking at the schematic or block diagram, and turning volume and tone controls, you can see what does and doesn't affect the noise. Use headphones. If it goes away when you crank the balance control to full L or R, and is affected by the loudness control knob, I'll give you a hint- it's likely the class A transistor amp/buffers right before the power amp. 2SC1213A (TR671 and TR672) and come in the same odd angled front package as the infamous low noise 458 parts that so often go bad. Pinout is ECB looking at the angled side and going left to right. Emitter will be closest to the front panel.

You can look at the pins marked MR and ML (E is ground) on the board with the scope to narrow things down. The trick is always finding a point, seeing if the noise is there, then moving upstream or downstream to isolate where it originates. If it's after TR671/2, that puts it in the power amp and it will be much harder to find because feedback will make it appear everywhere in the circuit.

It could well be a problem elsewhere, but that's where I'd start, only because I fixed this exact same problem about an hour ago!

jblmar said:

At times there will be a slight 'shake' on the display. Check your connections. Reverse X & Y on the scope. If it follows check your probe. Double check that all connections are tight including Ground @ the scope & amp.

Click to expand...

I did, and the shaking follows. So I'm assuming it's really there.

Vertical shake on the scope display? What are your scope settings? AC or DC coupling?

Click to expand...

AC coupling. And yes, the shake is vertical. I have both signals (L and R) on the scope screen now, and only the left trembles a bit.

Measure DC offset at the speaker terminals.

Click to expand...

right is steady at 10.3-10.5, while left oscillates from 2.something to 5.something.

If you have Pre Out/Main In:

Click to expand...

unfortunately this model doesn't have them. Thanks for the feebdack.

Andrea

ConradH said:

If it goes away when you crank the balance control to full L or R

Click to expand...

it does

and is affected by the loudness control knob, I'll give you a hint- it's likely the class A transistor amp/buffers right before the power amp

Click to expand...

moving the loudness reduces the noise, at 5 is still present, very low, at 8 I cannot hear it anymore

2SC1213A (TR671 and TR672) and come in the same odd angled front package as the infamous low noise 458 parts that so often go bad. Pinout is ECB looking at the angled side and going left to right. Emitter will be closest to the front panel.

Click to expand...

I will give it a try. I have to remove the tone board I guess... probably I won't have time till next weekend unfortunately.

It could well be a problem elsewhere, but that's where I'd start, only because I fixed this exact same problem about an hour ago!

Click to expand...

Hehe, just in time to give me some good advices ;-)

I will keep you updated!

Thanks for the hints,

Andrea

You don't have to remove anything other than the case to get at those, as they live on the tuner board and are cabled over to the tone board. One is close to a metal bar, but easily removable. The other is right in the open nearby. Many common 50V NPN devices will replace them, or (it's a bit of a cheat) you could carefully remove and exchange positions with them. If the noise moves to the other channel, you've found the bad part. Caveat- heat of soldering could "fix" a bad part. Usually you can find this sort of problem by heating and/or cooling the suspect part, but mine didn't respond to that. I'd still give it a try.

ConradH said:

You don't have to remove anything other than the case to get at those, as they live on the tuner board and are cabled over to the tone board.

Click to expand...

You're right, I didn't see them the other evening, guess I was too tired.

And you're right also about the source of the noise: it's TR671, 2SC1213A. That's where the trembling starts, and since the trembling was both on base and emitter I also checked the transistors on the tone board and everything was hunky dory there, so I swapped TR 671/672 and the noise moved with them.

Thank you so much for the help and the information!

Andrea

You get a can of freeze mist and isolate the noisy part NOT on a scope but use a signal tracer. (so you can hear it)
Then replace the same part for both channels.

Good job! We're lucky that the same things tend to go wrong with any given piece of equipment, so we develop a bit of history to help with the next one.

qsaudio said:

You get a can of freeze mist and isolate the noisy part NOT on a scope but use a signal tracer. (so you can hear it)
Then replace the same part for both channels.

Click to expand...

Before using the scope I tried with a rudimentary signal tracer, basically a shielded cable with a cap attached to the tip of the inner cable (the tip is used for signal inspection, while the other side of the cable goes to the RCA inputs of another receiver) but could hear no noise through it. Maybe my tracer is not so good or maybe the noise was too faint.

What's your opinion on this gadget? Should I get a more sophisticated equipment?

Thanks
Andrea

That should be an effective technique, but it depends on the gain. Can't say why you didn't hear the noise. Noise can be tough using a scope because you can hear incredibly small amounts of it. I've often thought of building a signal tracer consisting of an old speaker, chip amp and a front end with tons of gain. Probably some limiting too, so it doesn't blow me out of the room.

As I mentioned previously, heat and cold usually isolates these sorts of problems very quickly, but in this case it didn't affect the noise much. No idea why. On removal I measured both the good and bad part on a curve tracer. The good part looked perfectly normal. The noisy part also had low gain and non-linearity, so it was probably affecting more than just noise floor.

The typical scope goes down to 5 or maybe 1 mV/division, and that's not enough for many purposes. One can build a scope preamp to improve the sensitivity, but unless you do a lot of low level stuff, it's probably not worth the effort. My secret weapon for low level work is an ancient Tek scope with a differential plug-in that goes down to 10 uV/division, with adjustable filters.

If you didn't see my other post, how far does your FM tuning meter swing when you go off signal? Mine moves about 1 tiny division, but none of my signals are very strong.

ConradH said:

If you didn't see my other post, how far does your FM tuning meter swing when you go off signal? Mine moves about 1 tiny division, but none of my signals are very strong.

Click to expand...

I'm sorry I think I'm missing something here, I can't understand your question. Can you reformulate it? I'll be happy to answer.

You should have two meters, the left one for signal strength, the right one to show when you're perfectly tuned. As you turn the dial away from a signal (detune, say to go to another station), the right meter will go off-center. What the maximum amount you've ever seen it go away from center?

spaghetti said:

Before using the scope I tried with a rudimentary signal tracer, basically a shielded cable with a cap attached to the tip of the inner cable (the tip is used for signal inspection, while the other side of the cable goes to the RCA inputs of another receiver) but could hear no noise through it. Maybe my tracer is not so good or maybe the noise was too faint.

What's your opinion on this gadget? Should I get a more sophisticated equipment?

Thanks
Andrea

Click to expand...

Sometimes you can cheat a little by using an older high impedance headphone set with test prods on the lead ends and then you can move through the audio chain quickly. It does take a little patience and somewhat of a learning curve to pick up the white noise on those headphones.

Also, with the transistors out of the circuit, you can identify the noisy one from quiet ones by using the bipolar transistor testing method listed in one of the stickies, and the lesser the difference between forward and reverse test lead polarities- the noisier the transistor will be. Usually a silicon transistor with a difference less than 1:7 will be noisy. (either base collector or collector emitter measurements.) A higher ratio will usually by quieter. Germaniums get noisy at about 1:5 or lower.

ConradH said:

You should have two meters, the left one for signal strength, the right one to show when you're perfectly tuned. As you turn the dial away from a signal (detune, say to go to another station), the right meter will go off-center. What the maximum amount you've ever seen it go away from center?

Click to expand...

Ok, now it's crystal clear. Sorry I didn't get the first time. The average movement is slightly less than 1 division here, very similar to what you experienced.

Thanks! Now I have three data points, mine, yours and one other. It appears that about 1 small division is normal.

OT- the tuner alignment on mine seems good, but I'll probably touch it up anyway. Yamaha seems to assume that the RF front end should remain perfect forever, as their manuals for the CR series completely ignore several of the adjustments that are typically made. Worse, other than getting the dial pointer aligned correctly, the CR-620 service manual is missing an entire page on setting up the rest of the FM section. Fortunately it's not hard to figure out, especially if you look at other CR service manuals. I notice that the up-model CR-820 has a bit better tuner circuit, with dual gate FETs in the front end. IMO, the CR-620 is probably best in areas of strong signal, rather than out in the middle of nowhere.