Recording/Playback Level Instability

Sony TC-K61 Recording/Playback Level Instability

For some time now, I have noticed that the deck's ability to record and playback at stable levels is questionable. Quite often I would have to recalibrate the playback levels (via 400Hz Dolby Level ref tape), and re-configure the record levels for my Maxell UR tapes. These random deviations from the target values were not small, but in the order of 2dB-4dB.

Frequently, it was the left channel that began to vary with time - was an hfe or hFE (ac and dc current gain) transistor parameter drifting as the deck warms up in the playback circuit? - if so, which one? Often small variations in hfe in a circuit with negative feedback have little effect. However, with reference to the playback circuit below - are the signal inverting transistors Q104/Q204 exhibiting problems?

New Components: So far I have replaced every capacitor in the audio and control boards, including all bias frequency generator capacitors. The bias frequency is stable at 100Khz, and generates around 30v (peak-to-peak) for Type I tapes at the record head terminals. Type II setting yields 34v, and about 65v for Type IV tapes.

Signal Path Relay: I even changed the Record/Replay Head signal relay to eliminate any possibility of a potentially and slowly oxidizing contact?

Old G2V-2 relay replaced with OMRON G5V-2. Note: old caps were still in circuit at the time of this image.

However, this full capacitor and relay replacement did not cure the problem, so what was it?

Playback Circuit: It was only recently that I began to study the playback circuit which consists of three transistors in an apparent Current Shunt negative feedback configuration - not Voltage Shunt as I originally thought? They are: 2SA836 (PNP), and the 2SC1345 (NPN). These are low power, low noise-figure (NF) transistors, which are no longer made.

There may be several schematic errors in the service manual?
Example: above Q103 and Q203 there are a biasing resistors in circuit, not capacitors.

2SC1345 Replacement: After searching the  internet for suitable replacements, I decided to 'chance' a KSC1845 Fairchild transistor which I had in stock. And which offered similar specifications - in particular dc hFE characteristics. As mentioned above, the basic purpose of Q104/Q204 is to act as a signal inverter, current-shunt negative feedback is later fed from Q105/205 to Q103/203.

From voltage measurements, the default 2SC1345 dc quiescent collector current for Q104/Q204 was approximately 310uA. Replacing both Left/Right channel 2SC1345 transistors with KSC1845s worked well - producing the precisely the same quiescent collector currents!!! Wow!

Board Modifications: Note - I have reversed the record and playback level potentiometers, re-soldering them so they are easily accessible. The 'new' KSC1845 NPN replacements are marked in red. (long pins)

Assessment: Currently, the replay amp is working well. After playback and record calibrations, the TC-K61 seems to have stabilized for both left and right channels.

Time will tell if I have found the source of the problem, fingers crossed!

Article subject to revisions, and correction of mistakes. 18/03/2020.
+10th minor revision, 31/03/2020.


26/03/2020: I have now replaced the Q103/203 and Q105/205 (2SA836 PNP) with Fairchild KSA922FBU PNP transistors, and recalibrated the TC-K61 - the deck's playback circuit is very stable! The dc quiescent collector currents Ic, and Vce quiescent voltages differed from the original 2SA836 configuration by between 2% to less than 5%. Virtually identical - good!

Finally, the buffer amplifier at the input line-in stage appears to be a simple emitter-follower configuration. The old (but not faulty) 2SC1345E was replaced with a Fairchild KSA1845FCK NPN transistor.

 The TC-K61 is working superbly well for Type I tapes!