WM-EX60 dynamic bass boost tuning

Here are the frequency response measurements of the original circuit in 'Normal' EX-DBB mode.
Left channel:


Right channel:


L+R channels:


As we can see, the frequency response is 9Hz – 70kHz +/-3 dB.
20 Hz – 20 kHz +/- 1 dB, which is great.
The channel matching is perfect.

 

Here are the frequency response measurements of the original circuit in 'MID' EX-DBB mode.
All measurements refer to one channel (left) since both stereo channels match very closely, with minimal changes.




Some conclusions:
- The system operates as a “loudness” function with more gain for lower level signals (-55 dBU...-40 dBU) and less gain for higher level signals (-35 dBU...-20 dBU).
- The bass boost can be as high as +15 dB(!) for some input signal levels. Unfortunately, as I expected, the Sony's tuning was odd and leaked too much boost into the low mids, making it sound muddy and bloated.
- Mid frequencies have no gain.
- High frequency boost is fixed at +5..+6 dB for all input signals.

 

Here are the frequency response measurements of the original circuit in 'MAX' EX-DBB mode.
All measurements refer to one channel (left) since both stereo channels match very closely, with minimal changes.




Almost the same result, with the only difference being that the maximum bass boost can reach +20 dB(!) for some input signal levels.
So the difference between MID (max bass boost +15 dB) and MAX (max bass boost +19.5 dB) is +4.5 dB more gain.

 

Thanks for the research. Unless used in combination with bass lean or underpowered headphones, no wonder why max I never ever liked, and mid I can tolerate only occasionally.
Tuning it down a bit would be awesome.

 

Thank you for your interest in my research. The next step is to reduce the bass boost and limit the frequency range of the boost.

You're right when you say that modern headphones and recordings aren't as bass-less as they used to be. So my goal is a maximum bass boost of +5-6dB and a treble boost (fixed) of +5-6dB.

 

I noticed that in this case, the factory specs of the NJM2106 microchip completely coincided with the actual measurements in the device. It's good when the documentation is reliable. The only thing is that this chip allows some adjustment of the bandwidth and gain of low-frequency signals at the discretion of engineers. Sony chose an extreme tuning option, with maximum gain and the widest possible bandwidth. The documentation provides options for more moderate tuning, which I will try to implement.

 

I'm sure back then those specs they had decided on were ok with the current Headphone, etc.

Thanks for the Update's Johny_2000

 

Going to be interesting to see the results. Good stuff!

 

Intermediate tuning results for both modes.
MID mode:


MAX mode:


It's not finished yet. The maximum bass boost level is significantly better in MID mode (+4dB). But in MAX mode it is still too much (+11 dB).
And the gain bandwidth is still too wide (10-320 Hz).
Next, I will continue tuning and reduce the gain to +3 dB/+6 dB in both modes, and the bandwidth to 10–150 Hz.

 

Last night I received new SMD resistors and capacitors in the mail. Soldered them this evening.
Tomorrow morning I will take new frequency response measurements to see the effect.

Here are some interesting electrical characteristics from the NJM2106M datasheet:

As we can see, there are two important frequencies to carry out test measurements: 50 Hz (low frequencies) and 10 kHz (high frequencies).
As we already know, the high frequency gain is fixed and its value is +4.5 dB.
Regarding the low frequency (bass) boost at 50 Hz depending on the input amplitude: -60 dBm = +2.6 dB; -45 dBm = +18 dB; -20 dBm = +3.3 dB.

 

Oh, now I trimmed too much.
MID only +0.8~1 dB:

MAX is about +3.5~4 dB:

I will have to change resistor R305 again to a value between the previous and current.

 

Looks Good John, I have a board ready to send!! When you have your desired results. I haven't really reworked a board this small SMD Components. You must have a nice "AMSCOPE" and steady hands.

Nice Work John!

 

Yes, this is the board I'm working on. I can't promise anything yet. Still in the process of tuning. Electronic component suppliers are holding me back a lot. I didn't have SMD components (resistors and capacitors) that small in size in my inventory. So I had to order. Initially one by one. It was expensive and slow. Then I bought various kits, but they don’t have everything I need in small increments. I still have to order some components separately.

I have access to an electronics repair station with a magnifying glass, lights, and body grounding strap. As well as some other equipment such as power supplies, oscilloscopes, signal generators, spectrum analyzer, etc. Please always store electronic board in a special anti-static bag. There are very sensitive microchips that can be damaged if touched.

 

Yes, I know all about the static! I have recapped my share of PC Motherboards in the Mid 90's! When they had quality control issues! Soyo, D.F.i, Asus. I posted a photo just to keep your Energy Level going. That's all.

Well, off to bed!

 

Okay, I think I got about what I wanted to see in this EX-DBB feature. At least according to instrumental measurements.
The final resistor value of 5.6 kOhm put this in the right place. Now I can replace the coupling capacitors C112, C212 and begin the listening test on music.

MID is +2 dB bass boost max:


MAX is +6.5 dB bass boost max:

 

I put everything back together and did a listening test on known pre-recorded tapes. Verdict: everything is good now! The bass boost is almost unnoticeable in MID mode and slightly noticeable in MAX mode. The treble boost is the same, which is good. Now I like it because it doesn't overload modern recording and headphones that don't lack bass. There is no more distortion or muddy sound in any EX-DBB modes. The recordings sound rich and bright, as I expect from them.

Note. The NJM2106M's bass boost is easily adjustable from 0-18dB, and anyone can tweak it to their liking by adjusting a few resistors and capacitors in this circuit.
This is my final edition:

 

Thanks John for all your testing and research work and final results regarding the EX[DBB] Bass Boost Settings on the Sony WM-EX-60.

 

Thanks, Duey!

I have already started the second part of the repair program: mechanical checks, replacements and lubrication.

The front and back covers, the electronic curcuit board, the tape head - pinch roller assembly, the drive motor, the pushbuttons - tape switches assembly, the frame, the battery compartment and the cassette compartment shield are removed (I didn't plan for the latter, but the old glue fell off on its own).

Inspected the gears, flywheel and capstons, solenoid and cam mechanism. So far most things look good.


The tape select switches were oxidized and had increased resistance. I scrubbed and washed them with contact cleaner. Now they measure well.

P.S. I bought the wrong headphone amp output capacitors, 330uF 2V. They were on the back of the board and I couldn't see (measure) them. So I used the service manual specifications to order new parts. Now the board is accessible from all sides - and the capacitors are completely different .

 

Discman land could sure benefit from similar, but different, tweaks to their atrocious Mega Bass boost switch as well... Worthy top 1990 model D-35 (D-350) and later 1991'ish D-303 "line out specialist king" could receive sudden added value by making that feature usable. Hmm.

 

Headphone amp output capacitors have been discontinued by all manufacturers worldwide. They are 330uF 2V, 4.3 x 12.5mm.

A few closest matches I can find:
220µF 6.3V - Panasonic ECA-0JHG221 (5 x 11 mm) and Rubycon 10YXJ220M5X11 (5 mm x 11 mm)
270µF 6.3V - Rubycon 6.3ZLQ270MEFC5X11 (5 x 11 mm). Non-Stocked, Factory Lead-Time 20 Weeks.
330µF 6.3V - Panasonic EEU-FM0J331 (6 x 11 mm).
470µF 6.3V - Rubycon 6.3ZLH470MEFC6.3X11 (6.3 x 11 mm)

I'll use a caliper to measure the free space under the PCB to see which one I can fit.
I will also measure the capacitance and ESR of the original capacitors, which I have now removed from the board.