Without doubts, Sony DD series walkmans are the most popular of all personal cassette players. This popularity is not only build quality but also because of sound quality that these walkmans have.. or should have?

As most of you know, Sony WM-DC2 or WM-D3 are the best sounding walkmans ever made (if we forgot about heavy D6C). But what about other DD models? Do they also sound so well? No, there is huge difference between them and above mentioned models. Why is that?

• Head used. While DC2 or D3 use amorphous head, other  models use just permalloy head. But I cannot hear well higher frequencies, where these heads are excellent. And anyway, DC2 sounds much better
• Head wear. Those permalloy heads are soft and wear very fast. Most DD walkmans have some hours played.
• Cicruit design and components used. DC2 uses discrete preamp stage while other DD models use IC preamp.

No, there must be also something else.

Frequency response

I measured FR of several WM-DDII to DD33 and result is very surprising. They’re absolutely missing bass, dynamics in lower frequency range is very poor. I am not talking about few decibels. Output level is 20dB lower at 20Hz and at frequency of 63 Hz it is still not in acceptable -3dB level. 

In case of DD30 and DD33, this problem is compensated by megabass circuit. When turned on, signal at lower frequencies is amplified.. but how? This effect is most evident at about 60 Hz. With cheap headphones, it improves sound a lot. But what happens if you use quality headphones/earphones? Sound is dull, it is amplified where it shouldn’t be. With megabass turned on, sound is unlistenable. With megabass turned off, it is missing bass. Frequency response is not flat. 

When you look at graph, you can see it clearly (the blue line marked as 10uF). I used FR tape and oscilloscope to measure level at various frequencies. I tested it on several DD models, all with the same result. All were tested after complete recap. Do you want better results? Buy WM-DC2! Or use the following modification.

Simple modification to get flat response

Note: these modifications will be audible only when high quality headphones are used. Their real frequency response lower-end must be 20-30Hz, not on paper only

All four DD models have one thing common. They use BA3304F as EQ preamplifier. The capacitor connected in series with the gain-setting resistor between the NF pin and the bias pin, is the bypass capacitor. The value of this capacitor determines the bass characteristics. Value of this capacitor is 10uF. In case of DDII and DDIII models, it is tantalum capacitor. As the tantalum capacitor ages, it causes issues in FR response. Look at the blue line, and peak at about 100 Hz. Replacing this capacitor to new one removed this issue but the problem with poor bass response remained. In later models, DD30 and DD33 Sony used electrolytic capacitors instead of tantalum ones as they probably knew about this issue.

Frequency response at lower frequencies can be simply improved by using capacitor of higher value. I used 47uF instead of 10uF with perfect results. See the ref line marked as 47uF.

• WM-DD30 – capacitors C103,C203, buy here
• WM-DD33 – capacitors C103,C203, buy here
• WM-DDIII – capacitors C102,C202, buy here
• WM-DDII – capacitors C103, C203, buy here

The bass frequency response can be improved little more by using higher value output capacitors – the ones connected between headphone amplifier IC and headphone connector. Modern polymer electrolytes are of very high quality and can have high capacity at small size. I recommend to use 560uF/2.5V instead of 220uF/4V. It is useful especially when small impedance earphones are used, e.g. 32 Ohm. This will improve bass at 20Hz about 6 dB.

Belt is critical component of every cassette player. That’s why choosing correct belt for your cassette device is very important.

Belts can be used on various places of your cassette decks, for example main belt, inter-capstan belt, mode belt, counter belt… Some are square, some are flat. The most discussed belts are those that affect audio quality. Old good belt producers are gone forever and current belt producers cannot make belts of quality required by sensitive cassette deck mechanics.

These are three parameters that have to be taken into consideration, when choosing right belt:

• belt quality
• belt dimensions – circumference, width and thickness
• belt material

Using incorrect belt will have negative effect to tape stability, wow/flutter, mechanics noise and parts wear. Each cassette transport is designed to use belt of specific parameters. It is always good to find balance between belt length, belt thickness and belt material. Some mechanics require elastic , some require harder belts. Some are more sensitive to correct belt, some less. It is given by design.

There is wide choice of generic belts of various lengths and types on today’s market. There are suppliers of generic belts, like ASWO here in Europe, or turntableneedles in US. They are very cheap, optically they look very good. The only problem is that their surface quality is not sufficient enough to meet specs given by manufacturers of cassette decks. Some transports, like Aiwa dual capstan transport, are very sensitive to belt quality.

Fortunately, quality of flat belts can be significantly improved using method of sanding. There can be several reasons why the belt is not perfect.

One of them is surface not smooth enough. It does not necessarily need to be visible by naked eye. You can feel it when you move your finger along the belt surface.

It is possible to use this DIY method using very cheap tools, at home environment. The only tools that you need is 600 grit sanding bands rotary tool. As a result, surface of such  belt will be smooth, uniform and flat.

By using this method, I could reduce wow/flutter from 0.07% (10 sec peaks) WRMS to 0.03%, by just adjusting belt A of Aiwa XK-009 cassette deck. This method was tested in various cassette decks with similar results.

Warning! Not all belt materials are suitable for this modification. EPDM, NBR materials can be easily sanded. On the other hand, neoprene is very difficult to sand. 

Another one is thickness inconsistency. 

It means that the belt thickness varies around the belt length. Quality original belts vary less than 0.005 mm. In case of  currently produced belts, it is a question of luck. When I ordered new belts of various dimensions from my manufacturer, in case of some dimensions, the belts are perfect while in case of some other dimensions they are very badly. Sometimes also quality varies within the same size.  I measured some old Sony and Aiwa belts and thickness uniformity was always within 0.002 mm. In case of new belts, I never found better than 0.008 mm. 0.01 to 0.015 is typical. And some belts from crap producers vary within 0,2 mm! So what can we do to improve it? It is not difficult but it requires patience and a time. And you will need the sanding rotary tool described above plus quality micrometer with constant push force.

I use to measure belt thickness cm by cm and I mark place where the thickness is largest. I repeat measurement to confirm it.

Then, I use the sanding tool to adjust thickness around the marked place. I move up and down the rotary tool, around the marked place.

Then I repeat the whole process again until I find another maximum thickness. I repeat the steps until thickness varies within 0.005 mm range. 

It need a little practice but result can be very good.