# Mosfet based ripple filter for tube amp

I need some hint to fully understand how this circuit works (credits to Audio Design Guide):

simulate this circuit – Schematic created using CircuitLab

My doubts are in the operation of the MOSFET. I tried to gather some information from my books and online, and the summary of what I know about it is:

• R2, R3, R1, C1, R5, D1 and M1 work as something called "Capacitance multiplier"
• The above mentioned components work as a ripple filter
• The B+ Out voltage is determined by the R2 - R3 Voltage divider.
• B+ Out takes circa 110 Seconds to reach its maximum

From my understanding the ripple is minimized by the R1 - C1 Lowpass. What I don't understand is what the Mosfet does, how it operates. Can you give me a hint?

• Check your M1, both part number and connectivity. – WhatRoughBeast Apr 26 '18 at 15:03
• @WhatRoughBeast I swapped the original MOSFET with one I had at home. In the original design the MOSFET is not properly placed. Drain and Source are swapped. – Enrico Apr 26 '18 at 15:08
• A safe design would use a MOSfet at M1 whose drain-to-source voltage rating is greater than the maximum DC voltage at C2. – glen_geek Apr 26 '18 at 16:40

The MOSFET is acting as a source follower and this generally means that whatever signal is placed on the gate, is also seen on the source AND it is quite resiliant to ripple or movements on the drain.

So the source is the output and the gate is at DC level of 82% of the drain voltage BUT it's filtered (by R1 and C1) to remove any ripple seen on the drain.

R2, R3, R1, C1, R5, D1 and M1 work as something called "Capacitance multiplier"

It is generally called a gyrator circuit and in this configuration, the MOSFET acts as a series component resembling an inductor.

B+ Out takes circa 110 Seconds to reach its maximum

This is dictated by R1 and C1 - they have a time constant (RC) of 22 seconds. The rule of thumb is that it takes about 5 RC to reach the final value hence 5 x 22 = 110 seconds there or thereabouts.

• Great explanation. Thank you. Can you please just elaborate a bit the sentence "The MOSFET acts [...] resembling an inductor"? Why is it so? Does it oppose to variation of currents? – Enrico Apr 26 '18 at 15:05
• @Enrico I say "acts like" because in reality that is all it does. If it were acting like a resistor then a change in voltage at the drain would result in a change of current flow into the drain. Because of its configuration in the circuit, a slow change of voltage at the input (drain) is passed through but, a fast change is not (simple terms). Mathematically it can be shown to resemble an inductor. – Andy aka Apr 26 '18 at 15:22
• And this fast voltage change rejection capability seen at the source of the MOSFET is given by the lowpass R1 - C1, right? – Enrico Apr 26 '18 at 15:45
• @Enrico correct – Andy aka Apr 26 '18 at 16:29

For good filtering, you'll need adequate headroom for the FET; you need to avoid the TRIODE region where Vds is < 5 volts; examine the drain plot of your chosen FET, and find the flat-output curve region. Increasing R2 increases the headroom.