Help with Cassette Dolby circuit debug

[Jimkarl]

Have a Vintage Kenwood KX-1030 tape deck with noise in the left channel audio. This occurs even when playing back a blank tape, all setting normal or turned off (e,g. normal bias/eq, dolby turned off etc). Using and audio tester (I don't have an oscilloscope) audio was good at the Playback head and through the initial PB circuitry. Then I got to IC 103 (see attached pic) which I assumed was an amp chip of some sort. Audio good at the input (pin 6), then good at pin 5 , 2 and then 3. At the output, pin 7, There is a substantial amount of noise in the signal, definitely not evident on it's right channel counterpart. So I figure either the chip is bad or something driving the chip. So I take some voltage measurements of both left (bad) and right (good) chips. They are:
Left Right
Pin 1 9.4v 1.6
Pin 2 9.6 8.4
Pin 3 10.0 8.8
Pin 4 9.5 8.3
Pin 5 9.4 8.2
Pin 6 9.6 8.4
Pin 7 8.8 15.6
Pin 8 0 0
Pin 9 0 0
Pin 10 9.5 7.8
Pin 11 9.5 2.2
Pin 12 Meter read OL on both?
Pin 13 0 0
Pin 14 15 10
Pin 15 9.6 9.5
Pin 16 18.6 18.6

Obviously something going on here, but what? Further investigation, I find this chip is a NE645B Dolby processing chip, which was unexpected, especially since the dolby circuit is turned of, but apparently their design is to route the audio through it regardless.

So given the relatively high input voltage on pin1 (not the source input to the chip, which is actually pin 5, but one of the stages internally) and low output voltage on pin 7 (even though the output actually is much louder then the right channel one) What might be going on here? Is there come surrounding circuit component I should be looking at closer that is causing the problem, or is the chip simply bad?

 

[dorke]

The first thing that stands out is the different Vref voltage at pin 4.
It should be about 18.6/2=9.3v achieved by the voltage divider of R151 and R153 each 1k and filtered with C151.
Are you sure of bad/good, left/right on the pin voltages columns? looks the other way around regarding Vref.

 

[Jimkarl]

Yep I'm sure. The right side may be low(er) than expected but it is working fine. The left may have the correctvref, but I'm thinking the core problem is pin 1 and pin 11 are the same value. Looking at the alldatasheet for that chip (a TERRIBLE pdf btw - very difficult to read) it would imply either an internal short through the C side chain amp, or the voltage on Pin 1 is opening up the little variable impedance thing in that diagram. Clearly on Right side it's way different - Small voltage on pin 1, double that on the output of pin 11. My gut tells me these high voltages are playing a factor in all the distortion coming out of EX (Pin 7). In the diagram (attached) you can see there is some relationship from those pins and EX, though I don't fully understand what it is doing. But I would bet that if one or both of those pin voltages were reduced to what the righ side is seeing, the distortion would go away (unless the chip is fried internally of course)

 

[dorke]

The datasheet you posted is of the NE645 not the NE645B they are slightly different.
The info in the datasheet is minimal and enigmatic.
Look up the datasheet of the NE545 (an older Dolby IC) and the LM1112 (an improved version),
they both are pin compatible but slightly different and have a lot more info in their datasheet.

When you say "audio tester" ,what is it exactly?
You should use a proper scope for this,and a software one is also possible.

"noise in the left channel" and "distortion coming out of EX (Pin 7)" are different things.
can you elaborate on that.

An idea:
record the sounds of "good and bad" with a PC and post them here as "sound files",
it might help to understand what you are getting.

 

[Jimkarl]

I cannot find a specific NE645B datasheet. Are the the pinouts different? The Audio tester is a handheld transistor amp. As mentioned in the 1st post - I don't have a scope but I've learned to be able to tell pretty well with audio tester when there is a issue, my ears are pretty good. The audio signal is present in the left channel (all relevant pins, including pin 7). However when it comes out pin 7, there is a intermixed sporadic but constant short bursts of "noise" (distortion was the wrong word to use) in the signal that is not present in prior stages. Relatively loud as compared to the source signal, definitely easy to hear and obviously present. I'm not setup in a position to record this however, no pc or other recording device is present. However, it's like a pop and crackle, not unlike what crunching bubble wrap would sound like. It is only present when the tape is playing and it's clearly obviously when playing a blank tape, since it's not competing with the music as hard, though obvious even then.

 

[dorke]

I couldn't find the "B" part datasheet either.
The 645 is a direct replacement/ pin compatible with the 645B,the datasheet states that clearly.
Are the audio levels the same at pins 5,2 and 3?

If you look at the LM1112(page 1-121) you can see that bursts can occur in the IC operation at pin 7.
Are these what you are hearing?

 

[Jimkarl]

Audio levels are not the same, They get progressively louder and are equal to their counterparts on the right channel. I would ecpect them to get louder, as the chip diagram shows gain amps there.
The burst could be what I am hearing. I don't know that can say for sure though and it is not present on pin7 on the right channel chip.

 

[dorke]

I meant,
the same levels between left and right channels on pins 5, 2 and 3.

 

[Jimkarl]

5&2 seems to be about the same level wise across both channels, of course the volume is so low on both it's a bit hard to distinguish, certainly something an oscillosope would be better at detecting for certain. Pin 3 definitely seems slightly higher on the left channel. Not 50% louder, but maybe 20% (guess). My gut tells me they would all be slightly higher on the left, given the slightly higher voltages present.

 

[dorke]

Different audio levels play a roll in the way the Dolby IC performs.

Can you lower the audio volume at pin 3 of the left bad channel by adjusting the PB trimmer (VR1) to, any effect /improvement?

Before you do, mark the trimmer position.

 

[Jimkarl]

Trimmer movement has no apparent effect to the noise issue, though it has provided an additional piece of information. I can completely dial out the source audio, and yet still hear noise on Pin7 - a constant white noise with sporadic crackles and such, which I assume is what I am hearing under the source material when being played, though it seems to be much more prominent when there is source material present, then just by itself. It's the same effect if I don't even play a tape at all and just listen to the audio that is present at pin7. To make things worse now, the Right channel is doing the same thing. It also has all the same voltages on that chip (within a 1/10th volt) as the left channel, whereas before (when it was working properly), the voltages were radically different. This definitely was not happening when I started this thread and I'm glad I had captured all the "good voltages" on the right before it has now failed too. But I think I'm still stuck back with the original issue/question - Bad voltages causes the chip to create the noise, or the chip damaged causing noise and bad voltages? Which is the chicken, which is the egg?

 

[dorke]

You do need a scope to get to the bottom of this.
What is the load your "tester" puts into the circuit?
Maybe it is too small and caused damage?
I think it isn't wise poking around with it.

Crackles may be produced by bad caps or even a combination of dirt and corrosion .
How clean is the board?

What do you get at the output of the Unit not testing at pin 7 (noise and crackle)?
Are you getting getting the noise and crackle with the PB trimmers (VR1 & VR2) at minimum?
What effect does the Dolly NR on/off have?

The Dolby circuit involves a lot of filters in it, as well as a rectifier.
Lots of caps /resistors in there(area between pins 4 and 15).

Do you have a cap to shunt/replace C38(100uF/16V)?
Any effect if you do?

Damaged ICs are possible the question would still be what caused that if at all.

P.S Looking at the schematics.
ICs 103 and 104 have their pin 4(Vref) wired together,
therefor they should be the same value always.
A 1v difference, like in your original list makes no sense at all unless the connection of the pins is "bad"/broken.
What are you getting at pin 12 ?

 

[Jimkarl]

Since the issue was occurring before I ever popped the case lid, I don't think my tester has caused any problems. I have used it on numerous amp/receivers without issue.
Circuit board is super clean, both sides
The noise/crackle etc is present regardless of trimmer position (off, mid or other) The only thing that changes is the source audio level. With a blank tape, it's basically high end hiss that I am removing or adding in, the noise is unchanged.
Turning dolby on or off doesn't remove the noise, it just tones down the high end on it.

I assume you meant C308 not C38, Pulled and checked. A bit out of tolerance (reading 134uf) but esr measured good. Replaced for good measure anyway - no change.
Pin 12 measures 9.3 on both, though the first time I measured (today) I saw 8.3 on the right, could not reproduce that since.

More info - Upon power on the unit, the Left VU meter goes completely nuts for several seconds (over 100 on the scale many times and wildly moves around. After about 15- seconds or so it settles down to about 10-15% of that movement but never disappears and is constantly moving. This occurs with no tape inserted or playing. With nothing inserted or playing, Although very difficult to hear because of the extreme low volume, even when connected to an external amp turned way up, the noise is present. It's not until you press play on the tape that the noise becomes amplified and is easily audible. The Left and right are somewhat different. After listening to it for an hour now, I would better describe the left as a hurricane or crashing waves. Constant but a moving noise source (moving in meaning changes but in volume and in frequency, though not radically) There are also underlying clicks/pops that occur. When that a louder pop happens, the VU spikes up to about 30-35%, whereas it's normally moving around the 10-15% range with the noise only. On the right side, there is much less of the hurricane sound, if any, and much more pronounced clicks/pops (also seems to be at a higher frequency then the clickpops on the left, but hard to quantify. The VU meter rarely moves on that channel unless a pop happens then it spikes up to about 30% momentarily. It is pretty infrequent but is occurring.