Looking at updating my vintage Sony TA-E88, arguably the best preamp ever made ;-)
The phono headamp is AC coupled to the cartridge with two parallel 470uF/3.5V electrolytics. They need to go, but what to replace them with?
Anything other than electrolytics is unobtanium for that value.
But why does that need to be AC coupled anyway? It's not that there is a chance that a phono cartridge gives off DC, is there? And the input is referred to ground already.
Jan
The phono headamp is AC coupled to the cartridge with two parallel 470uF/3.5V electrolytics. They need to go, but what to replace them with?
Anything other than electrolytics is unobtanium for that value.
But why does that need to be AC coupled anyway? It's not that there is a chance that a phono cartridge gives off DC, is there? And the input is referred to ground already.
Jan
Attachments
Maybe to keep the base current of the input stage out of the cartridge? It also prevents loud thumps when you switch those switches at the input.
Yes that base current could be the reason. Especially with all the parallel input transistors. The TA-E86, a lower-cost version, has only single transistors here and the coupling cap is only a single 470uF.
But since both sides of the caps are at essentially ground, there would not be a switch-thump, would there?
Jan
But since both sides of the caps are at essentially ground, there would not be a switch-thump, would there?
Jan
".. with all the parallel input transistors"
Eight transistors act like one current wise if nothing else is changed.
It is done for noise reduction only, with a theoretical gain of 9 dB here -
but you know this.
Eight transistors act like one current wise if nothing else is changed.
It is done for noise reduction only, with a theoretical gain of 9 dB here -
but you know this.
Yes but I was reacting to Marcel's mentioning of the base bias current. That flows through the cartridge if DC coupled. With 8 // transistors it is 8x the individual bias current, unless they are all conducting 8x less current than a single which I believe is not the case. The tail current is about 40mA.
Jan
Jan
Last edited:
R101//R102 ~= 98.82 ohm. With 20 mA per side and a wild guess of 200 for the hFE, the base current is 100 uA. 100 uA times 98.82 ohm is 9.882 mV. Connect a 10 ohm cartridge in parallel, and the voltage drop becomes 0.9081 mV, so there is a voltage jump of 8.974 mV. Assuming a low-frequency gain of 10000 times, that is an 89.74 V jump at the output of the MC amplifier. That's quite a thump!
"Noise wise" I have to correct myself probably, because total
current set by the tail resistor does not change with the number of
transistors and while noise may be current dependent the 3 dB per
factor 2 rule may not apply. Joachim can tell us.
current set by the tail resistor does not change with the number of
transistors and while noise may be current dependent the 3 dB per
factor 2 rule may not apply. Joachim can tell us.
Assuming constant hFE and neglecting any changes in the 1/f corner frequency, paralleling transistors without changing the tail current only reduces the noise related to the base resistance.
Paralleling transistors and scaling up the tail current proportionally reduces all contributions to the equivalent input noise voltage, which means mainly the base resistance and the collector shot noise related components. It also increases the equivalent input noise current.
Paralleling transistors and scaling up the tail current proportionally reduces all contributions to the equivalent input noise voltage, which means mainly the base resistance and the collector shot noise related components. It also increases the equivalent input noise current.
Marcel, as usual, you hit the nail on the head.
Which means I need to find a very clean 1000uF. OTOH, the voltage across that electrolytic will be minuscule so maybe it really doesn't matter at all ...
Jan
Which means I need to find a very clean 1000uF. OTOH, the voltage across that electrolytic will be minuscule so maybe it really doesn't matter at all ...
Jan
The volume pot is 3k for lowest noise. It consists of several concentric track sections to get the right law. An engineers' wet dream ;-)
Volume at the left, output amp and stepped level control at right.
Jan
Volume at the left, output amp and stepped level control at right.
Jan
Attachments
Last edited: