There has to be something wrong with the LM317s I'm using. Bought them surplus several years ago. I originally calculated for 16v, but got 14v. I recalculated for 18v and got 15v. After the bridge I get 19.34v. Could it be that I'm asking the LM317s to work too close to its minimum requirement of 1.5v difference?
Good news is the loud hum went away at 15v and got some distorted audio out of both channels...NEED MORE POWER!
Good news is the loud hum went away at 15v and got some distorted audio out of both channels...NEED MORE POWER!
Carpenter,
It's a Vishay 3 amp 200v bridge, nothing fancy. Compared to others, it had a lower Forward Voltage drop @ lower currents. The spec sheet graph explained it.
Might not even be special in the above mentioned. Just what I found.
GBPC102-E4/51 Vishay Semiconductors Bridge Rectifiers
It's a Vishay 3 amp 200v bridge, nothing fancy. Compared to others, it had a lower Forward Voltage drop @ lower currents. The spec sheet graph explained it.
Might not even be special in the above mentioned. Just what I found.
GBPC102-E4/51 Vishay Semiconductors Bridge Rectifiers
A simpler alternative to a negative supply is to lift the Source with a resistor between 270-470Ohm. This will lower gain, but then you need to use a cap on the input. Very nice sound...
Skorpio, with such a high value of Rs (and still keeping the single rail PS) you'll get the Id of 1mA or so and your Drain will be only 2-3 V far from the PS voltage, leaving you with a very small output voltage swing. So, you'll have to use higher value of Rd to compensate for a such a low Id and to get you further from the +rail. That will cause a higher gain again, but this time you set the JFET deep into non-linear region of transfer characteristic and your THD explodes.
Anyway, this might sound great if you use it as an electric guitar preamp (you'll have that trash-metal guitar sound), otherwise, about 200mV is the max input signal you can bring in without causing the clipping (max output swing would be about 1V rms).
Why do you tink that input cap is needed in this case? If your idea is more complex than just changing the Rs value, please attach the schematic.
Last edited:
Hi,
It is not changing the 10ohm Rs, but connecting another Rs in series with the original. The 1M gate must still connect to the 10 Ohm, so gate is no longer reffered directly to 0V, therefore AC coupling the input.
It is like biasing a tube when you have a large kathode resistor. You split it in two and refer the gate to the junction.
My values are: Supply=18V, Rd=1K, Id=8mA and using 2sk147 jfets with IDSS=11mA
It is not changing the 10ohm Rs, but connecting another Rs in series with the original. The 1M gate must still connect to the 10 Ohm, so gate is no longer reffered directly to 0V, therefore AC coupling the input.
It is like biasing a tube when you have a large kathode resistor. You split it in two and refer the gate to the junction.
My values are: Supply=18V, Rd=1K, Id=8mA and using 2sk147 jfets with IDSS=11mA
Now it's clear what you mean - here is the picture for those willing to try.
This way you can extend the input signal level up to 1Vpeak and output up to 3Vpeak. Compared to neg. rail version this is still an inferior solution - you still got higher THD and lower signal level headrom, but yes, it will perform better than simple single rail version
Input cap shouldn't be a problem, with Zin=1M this cap can be as small as 47-100nF, so a good one shouldn't be expensive.
Anyway, changing the JFET to J310 (Vishay) is better solution with simple signal rail version (no input cap, no level shifting, no neg. rail, 6-10dB gain, large signal headroom - 5-6Vpeak at output).
This way you can extend the input signal level up to 1Vpeak and output up to 3Vpeak. Compared to neg. rail version this is still an inferior solution - you still got higher THD and lower signal level headrom, but yes, it will perform better than simple single rail version
Input cap shouldn't be a problem, with Zin=1M this cap can be as small as 47-100nF, so a good one shouldn't be expensive.
Anyway, changing the JFET to J310 (Vishay) is better solution with simple signal rail version (no input cap, no level shifting, no neg. rail, 6-10dB gain, large signal headroom - 5-6Vpeak at output).
Attachments
Last edited: