Friday, August 7, 2026
Update to the Pleiades IRFS450 single MOSFET zero feedback power amplifier in class A operation
At your risk. Please take all safety precautions. A suitable fuse must be in series with a battery for injury or fire hazard protection.
It is almost one MOSFET, an output transformer and gate biasing arrangment.
Today the cute little IRFS450 MOSFET was mounted on a big black anodised aluminium heatsink about 20 times the area of the transistor itself.
Vb is 12V from a 1.2Ah lead acid battery.
Bias is created as follows. From Vb a potential devider is created with one 47Kohm resistor in series with a multi turn 50Kohm potentiometer. Bias is taken from the wiper of the potentiomer and directly applied to gate.
1st test with a 100 ohm load resistor. Vd=7V at Vg=3.3V
2nd test with 6R8 load resistor. Vd=7V when Vg=3.3V
3rd test with only the primary of the Pleiades 1:1 output transformer with air gap, connected between Vb and drain. The primary DC resistance is 2.4ohms. For given conditions best sound is when Vg=3.7V.
Signal path:
Brothers is Arms - Dire Straits on vintage compilation CD - SONY D-EJ758CK CD Walkman line out - Pleiades IRFS450 - Marantz Imperial 5-G
Yes it worked and sounded surprisingly nice.
Initaly there was some clipping distortion as no volume control had been connected. A potentiometer should always be connected for ear protection anyway. The nearest potentiometr found on the table was a 220Kohms. It worked attenuating the output of the CD walkman but once there was attenuation there was severe treble cut. Possibly form high impedance feeding the high? gate capacitance.
When poentiometer (which must be changed soon to a low K value) was at max, at the begining of the song which is quiet the sound is phenomenal, bright, clear with great bandwidth.
As bias is increased sound is not getting better. There seems to be voltage drop at the small battery.
Heat sink is running hot.
Things look promising. Thank to Panos for giving me the inspiration to try a MOSFET by letting me know of the Suzan Parker zero feedback schematic.
Thursday, August 6, 2026
In Music the bass line or melody line are both one note at the time
At your risk.
For example listening:
On exaple below Melody is the top line, Bass the bottom line:
Anothe bass example, bass going in octaves:
Another example:
Classic example of just voice and bass at 0:38, True - Spandau Ballet:
The rainbow makes you feel happy
At your risk.
It can be seen while watering with a hose thirsty plants on a field, while sun rays are scattered by the water expanding beam.
Beginning of the Pleiades IRFS450 single MOSFET audio power amplifier in class A operation (electrons flowing all the time)
At your risk.
Possibly the first time experimenting with a MOSFET in the Pleiades lab.
I think it is same to IRF450 on a different case.
It is placed on a relatively small heat sink.
In order to check the MOSFET is in order, a digital mulimeter was used at resistance measurmnet. Drain to source resistance reading was infinite.
But when an exhauted 9V battery giving 5.6V was connected from source to gate, the MOSFET had tuned on and Source to Drain resistance was of the order of 0.3 ohms.
Thia condition was preserved even after removing 5.6V to gate. Gate keeps the charge.
Promising. I am curious how will it sound with an output transformer driving the Marzntz Imperial 5-G loudspeaker, MOSFET driven by the SONY D-EJ758CK line out.
Update on the Pleiades derivative of simplified no negative feedback Musical Fidelity A1
At your risk.
Refering to the original Musical Fidelity A1 schematic.
The idea is to try just a BD135 driving a 2N3055 driving an output transformer and no negative feedback.
A BD139 was used instead, a Mitsubishi 2N3055, a bifilar 1:1 output transformer istead of perhaps a more suitable 2:1 output trasformer. Air gap is made by sticking a layer of 3M scotch invisible tpae on the stacked Es.
Vb is 12V.
BD139 is mounted on a small heatsink. 2N3055 is mounted on a larger audio cassette case size heatsink, black aluminium. It run bare touchable hot at 24V.
4R7, 100 ohms, 300ohms resistors are used as on the Musical Fidelity output stage schematic.
Bias is adjusted by a 220Kohm in series with 1Mohm variable resistor from Vb to the base of BD139.
2N3055 Ic was set to about 400mA.
Signal path:
various CDs, Dire Straits, Barbara Streisand, Year of the Cat - Al Stewart - Sony CD Wlkman line out, Pleiades simplified striped down Musical Fidelity A1 - Marantz Imperial 5-G
It worked, it sounded loud, even louder at Vb=24V, Ic=800mA.
Changing 4R7 to 6R8 made for a somewhat more open sound.
Quick changes are posble thanks to using Wago snap connectors.
But there was lack of treble, as if one or two high octaves were missing.
Going back to such a single 2N3055, the sound so softer, less power output but better sound quality, a bit more treble.
Going back to the Pleiades Darlington BD650. It sounded much more open with extended high frequency bandwidth and deep bass too, as if 2 more high octaves were added. Loudness was between the 2 above extremes.
Why does it sound much better? Is it because there are no resistors around the transistors? Should a BD135 driving a 2N3055 be tried in this way with least possible resistors?
The Darlington BD650 at Vb=12V sounds great at just Ic=100mA, saving charged battery time.
On a small heatsink it run luke warm.
Moving closer to the spekaer at abot 1m from 3 m the sound is even more impressive.
WHat makes this so simple Darlington amplifier stand out is how alive drums sound, hi-hat sounds for example as if there is a hi-hat inside the room, for example of Sultans of Swing - Dire Straits, or On the Border - Al Stewart.
Subscribe to:
Posts (Atom)