MJR7, a low distortion MOSFET power amplifier

Important note: The construction of DIY amps requires knowledge of the electrical safety standards to prevent electric shock and the risk of fire. Failure to heed the applicable safety rules may result in serious or fatal injury.

I’ve been looking for a while to find a DIY amplifier that can be suitable for my loudspeaker projects. And this is what I’ve found.

Mike Renardson from the UK has designed an ingenious power amplifier, for which detailed DIY construction instructions are provided on his site: http://www.renardson-audio.com/mjr7-mk5.html. There is also a lot of design information and measurement data there.

This amplifier is really a revolutionary one, simple circuit-wise, yet at the same time has so low distortion that it can be said that the MJR7 probably doesn’t have any distinctive sound of its own. One thing that may actually be heard is that the MJR7 is an inverting amplifier and people may be sensitive to phase at low frequencies, so a difference may be heard between the MJR7 and a non-inverting amp. But the phase inversion problem is very easily cured: reverse the loudspeaker connection at the amplifier or at the loudspeaker terminal! (It’s also worthy to note that many music recordings don’t preserve absolute phase in the recording chain, so perhaps we shouldn’t worry about phase inversion anyway.)

Since this is a low power amplifier, I recommend it primarily for my Vivace 100 speaker. My other loudspeakers can't be driven to full power with it, though that much power is not necessary for normal listening levels in a normal living room.

The MJR7-mk5 has been successfully built according to Mike’s original board layout more than 30 times. Some people made their own board version of the MJR7, but those were not so thoroughly tested as Mike’s original board layout.

The current revision level of the amplifier is MJR7-mk5. I exchanged some e-mails with Mike Renardson before I made this recommendation of his MJR7 amp on my home page. I had the goal for the design to be totally problem-free for DIYers. Perhaps Mike will include some of the following recommendations in his next revision:

1. Mike has already pointed out on his homepage that the MJR7-mk5 quiescent current can’t be set properly, if the Vgs value of the 2SK1058 and 2SJ162 MOSFETs is near the top end of their specification. Although seemingly nobody who’s built the amp so far has run into such high Vgs samples yet, but they may exist. Mike has advised a simple circuit modification to solve this potential problem.

2. Separate fuses for the left and right channels (since there are two channels on one board).

3. About 30% more gain, because at the highest recommended supply voltage (80 V) the input voltage needed to reach full output is nearly 2 Volts.

4. The use of multi-turn trimmer pots might be recommended, too.

Further notes:

There are a lot of 2SK1058 and 2SJ162 fakes on the market. Buy them from a reliable source.

The maximum undistorted output sinus power depends on the supply voltage and load impedance. The following calculations were made on some assumptions. Namely that the supply voltage sags 8 Volts at the highest output power (this value depends on the stiffness the power supply) and that the output can go as close as 10 Volts to the supply rails without being distorted (which is realistic in this amp). Then the max. power would approximately be as follows:

	Supply Volt.: 60 V	Supply Volt.: 70 V	Supply Volt.: 80 V
Load: 8 ohms	16 W	27.6 W	42.3 W
Load: 6 ohms	21.3 W	36.8 W	56.3 W
Load: 4 ohms	32 W	55.1 W	not recommended

The recommended slow blow fuse rating needed for different supply voltages for two channels (decrease the value to about 70% for one channel) are as follows:

	Supply Volt.: 60 V	Supply Volt.: 70 V	Supply Volt.: 80 V
Fuse	T2A	T2.5A	T3.15A

Main page