sony: Sony ICF-2001 random resets due to the 3V battery issues

There are plenty of reports publicly available concerning recurrent issues with the 3V battery system powering the D/A Board.

Symptoms include failure to power ON, blank LCD, random resets and total lockouts requiring to remove and reinsert the two 1.5V dry cells to recover.

I acquired two of these radios, and I am getting courage to dismantle them piece by piece for inspection and further testing.

One of them is in need of a loudspeaker replacement, because when I was opening the cabinet, I had to force the speaker a little to be able to move it away from the front panel while removing the front panel from the chassis.

You know, the speaker is maintained attached to the chassis with one screw that no one can reach before separating the front panel. 

It turned out the speaker was glued (!) to the front panel. Really strongly glued. So the paper cone was pulled and torn at one side. Nice start of mine. I should've known better, as I saw this approach on other models.

But I am digressing. Back to the subject.

After looking into the schematics, I was surprised to learn that Sony believed that common spring loaded battery terminals, prone to oxidation leading to potential power interruption due to voltage bouncing due to sound vibration or human operation resulting in digital circuits failing to work properly, was not a point of failure in the design.

I was expecting to see a large capacitor in parallel with the 3V battery supply as a minimum power backup protection against voltage fluctuation due to mechanical vibrations affecting the battery contacts.

But no, there is only a tiny small bypass capacitor C208 0.022uF.  

Sony states a current consumption of 75uA from the 3V batteries. Consumers are IC203 CMOS 4011 NAND gates and the so called LCD/Control Module. 

Granted, one does not know what is inside this LCD/Control Module as it is not documented as far as I know. Maybe there is a large capacitor or backup battery inside, who knows. If this is the case, it must be defective after all these years.

Even if there is a filter component inside the LCD/Control Module and is not defective, the IC 203 CMOS NAND gates are far away from that filter and therefore are not properly protected against battery voltage bouncing.

Anyway, assuming 2VDC as a minimum to maintain the MOS 4011 and LCD/Control Module processor running, the small bypass capacitor C208 with just 0.022uF. over a load of 40KOHM (calculated as 3VDC / 75uA) would result in a time constant of just 88uS and reach the 2V minimum threshold level in just 36uS.

36 micro seconds.Basically there is no debouncing of intermittent, random, battery contact failures caused by sound pressure from the speaker or when moving or operating the radio.

Even if there is a filter component inside the LCD/Control Module, I see this as a design weak spot.

Fortunately the solution to this issue is quite simple. Just add a small electrolytic capacitor, something like 100uF to have 1.6 seconds until reach the threshold voltage which is much more than enough to filter random voltage drops due to bouncing, or else a 2200uF cap to have 35 seconds running time, enough to replace dead batteries.

 

 

To thank the Author because you find the post helpful or well done.

A forum member brought to my attention that the use of electrolytic capacitors in parallel with batteries meant to last long life, as is the case here, may reduce the battery life due to electrolytic capacitor leakage current.

This is a fact. If a capacitor have a current leakage of 50uA, for instance, added to the average consumption current of 75uA, will reduce the battery life in half. 

Also, let us not forget that all batteries also shows self-discharge, despite some being better than others.

The leakage current is well documented and usually increases with the working voltage and with the capacitance value. So a 2200uF will show higher leakage than a 100uF of the same brand and type and working voltage of 3V. 

That said, not all electrolytic caps are created equal. Some are better quality showing lower leakage current.

Even in the 80's, when the radio was designed, such good caps did exist.

I would recommend to check the manufacturer data sheets before committing to a specific capacitor type, and choose the lowest possible leakage current using 105°C types 

For example, one manufacturer,  among others, Zonkas, offers the LLR series, where a 2200uF would typically show a low 13.2uA leakage current, and a 100uF would present just 0.6uA at the 3V working voltage.

 

EDIT: Forgot to mention, the common formula to advert the maximum leakage current for several manufacturers of current electrolytic parts, among the well known Nichicon and the mentioned Zonkas, is:

• 0.002*C*V or 0.4 (µA), where C is the capacitance in uF, and V is the working voltage in Volt.

 

To thank the Author because you find the post helpful or well done.