AX84 SEL from AmpMaker N5X Build

Update: I have updated the schematic with a few changes made: added a humdinger, and changed the grounding scheme to a star layout.

FURTHER Update: I found the sound of the amp a bit shrill and lacking some low-end. Might be my speaker cabinet. However, after searching the AX84 forum I tried a couple of changes others have made.

  • I increased the dropping resistor R73 to the preamp to 15k. This reduces the preamp voltages by ~80V and seems to make the distortion smoother.
  • I increased the 1st stage bypass cap by putting another 0.68uF cap in parallel to C51.

With these changes (shown on the updated schematic below) things sound really good.

My first build was sacrificed to make a ‘rehearsal studio’ amp – something to leave there between rehearsals with a bit more clean headroom that works well with pedals.

I had thought to make a 15W version of my L2L/UHG build but chickened out – time pressures meant I had to get something built quickly. So I gutted my 1st build, back to the chassis, and bought a 15W output transformer and beefier power transformer from AmpMaker, and decided to upgrade the N5X power board and build a new preamp board based on the AX84 SEL design.

It’s stock apart from the following:

  • Variable bias mod
  • Lead / rhythm switch
  • 47k (well, two 100k in parallel so 50k) instead of 100k cathode follower resistor R11
  • Switchable power tube grounding – the ‘Alex’ mod meaning the power tube is grounded through the output transformer (although I’m not sure if this is implemented correctly and I can’t hear any difference)
  • Switchable bypass on the 2nd and 3rd stages – 1uF, 10uF or unbypassed – taking inspiration from the original N5X
  • Tone defeat switch – again, from the N5X
  • VVR scaling of the poweramp – again, from the N5X
  • A few components in the power section as I kept things closer to the N5X, but upgraded caps for the higher voltage
  • Using a 6L6 initially, as I have a spare JJ and two spare Svetlanas knocking about.

The build was easier than my 2nd and the amp worked 1st time. My measured voltages are very close to the SEL schematic values. Note, I only used shielded wire on the input to 1st grid stopper, making things massively easier time-wise. In the pictures, the wires I was expecting to have to replace are the long-ish pink wires crossing the preamp board. There was a bit of squealing with everything dimed, but moving the wire from the board to the 2nd grid (purple) about 2mm upwards fixed that. Minimal hiss, and only slight 100Hz hum. Brilliant.

The amp sounds great – a bit lower gain than I was expecting based on a couple of audio clips / youtube videos I’d heard, but maybe that’s all in my head. It will be great for most uses. Highly recommended. The switchable bypass is of great use – in unbypassed mode things are nice and smooth, in 1uF mode it’s nice and bright, and 10uF mode gives an overall boost.

Contact me if you want the layout file to make your own board.



Scope Traces

I took some scope traces at each grid and at the speaker connection with a 500Hz 1Vpk-pk input signal. See them here.

Hum & Harmonics

I took some FFT plots while I was scoping. Below shows a 200Hz signal, 500Hz and 1kHz injected. Reading taken at the 6L6 grid (I think).



I took the following reading at the speaker output – 500Hz input signal all gains at 100%

Speaker - gains dimed

The below shows the background noise taken at the speaker output – the 100Hz hum is clear.

Mains-gains dimed

Power Output

With the scope, I injected a signal (500Hz, 1Vpk-pk) and adjusted the controls to get the sine wave on the power tube grid to be just on the point of flattening. The trace taken at the speaker output (going in to an 8ohm load) indicated 9.7Vrms. This givesP=V^2/R = 11.76 W clean.

Turning everything to max I measured 11.2Vrms, giving 15.68W.