Following in the foot steps of the ALPHA amp, I have hybridized the Aksa-Lender input stage with the output stage of the Pass M2 amp. Instead of the un-shielded auto-former that provides the 6x voltage gain (but is susceptible to EMI and hum pickup), I have used the Aksa Lender preamp to serve as the input stage and the voltage amplification stage. The Pass M2 output stage has a self-regulated slow-ramp bias scheme which makes it really nice for easy setup and no turn-on speaker thump. Some people may feel the magic of the M2 is in its autoformer, but I for one had a lot of noise issues and it has never been as quiet as I would like even when using mu-metal shielding on it. I got noise down to about 400uV rms at the output. But if I pressed my ear to the speaker cone, some residual noise was still present. Overall, i think this will be a wonderful match as the output stage has no global negative feedback, and the input stage has very low negative feedback (local). This should make for excellent soundstage, imaging and a natural engaging sound quality.
As the M2 power output stage is already capacitively coupled with a 10uF electrolytic cap per the original M2 design, so it was a simple matter to add the Aksa Lender front end. This has another benefit in that circa 15dB gain of the M2 basically always required a preamp to drive it to achieve the full range swing for 25wrms. Now, setting the gain of the input stage and VAS to 25.4dB, the ALPH-M2 needs no preamp as it is built-in. I do recommend a separate single rail power supply for it which can easily be drawn off the existing +24v rail through a step up DC-DC converter and cap multiplier CRC combo as is customarily done on the Aksa Lender preamp. I may add a custom on-board DC-DC step and cap mx for the front end PSU as that is not that hard to do with just a small handfull of parts.
Here is the LTspice schematic:
Here is the predicted FFT for 8Vpp into 8ohms - signature is pure H2 and H3 like a good SE Class A amp:
Here are the predicted components of the distortion for 8Vpp into 8ohms, about 0.007% THD and mostly H2 and H3. This is with the usual 1.27amps bias current and +/-24v rails:
Here is a louder playing level of 8Wrms, THD is about 0.01%:
Next step is to make a prototype, or perhaps just take my Aksa Lender pre and jumper it right to the spot after the autoformer...
I would like to thank Hugh Dean for the wonderful Aksa-Lender preamp input stage, and Mr Pass for the superb M2 power output stage. Again, we have the best of both worlds.
As the M2 power output stage is already capacitively coupled with a 10uF electrolytic cap per the original M2 design, so it was a simple matter to add the Aksa Lender front end. This has another benefit in that circa 15dB gain of the M2 basically always required a preamp to drive it to achieve the full range swing for 25wrms. Now, setting the gain of the input stage and VAS to 25.4dB, the ALPH-M2 needs no preamp as it is built-in. I do recommend a separate single rail power supply for it which can easily be drawn off the existing +24v rail through a step up DC-DC converter and cap multiplier CRC combo as is customarily done on the Aksa Lender preamp. I may add a custom on-board DC-DC step and cap mx for the front end PSU as that is not that hard to do with just a small handfull of parts.
Here is the LTspice schematic:
Here is the predicted FFT for 8Vpp into 8ohms - signature is pure H2 and H3 like a good SE Class A amp:
Here are the predicted components of the distortion for 8Vpp into 8ohms, about 0.007% THD and mostly H2 and H3. This is with the usual 1.27amps bias current and +/-24v rails:
Here is a louder playing level of 8Wrms, THD is about 0.01%:
Next step is to make a prototype, or perhaps just take my Aksa Lender pre and jumper it right to the spot after the autoformer...
I would like to thank Hugh Dean for the wonderful Aksa-Lender preamp input stage, and Mr Pass for the superb M2 power output stage. Again, we have the best of both worlds.
Attachments
Last edited:
Interesting mod is to run at +/-27v rails, use 60v for front end, and change R24 & R25 to 0.22R for 2.8A bias and 72w dissipation per MOSFET. Then upgrade outputs to larger die IXYS MOSFETs used on Alpha BB amp. Now the amp becomes a 35w Class A. This is a significant improvement in headroom over a 25w amp.
Great idea X! Would be great if it could be configured so that the M2x boards could still be used and the AKSA Lender Pre could be jumper-ed in like you say. Then you could A/B the original M2 sound vs. all the M2x input DB's, vs. the AKSA Lender Pre input stage. So many opportunities await!!!!!
Let us know how the initial experiment goes!
Let us know how the initial experiment goes!
Hi Jwjarch,
I think if you wanted to compare to the other flavors of M2X, simply take an Aksa Lender pre DB and adapt it to fit the 4 standoff connectors of M2X. You could add a jumper to bypass the Edcor. I am not sure if this build is going to be able to accommodate the M2X DB’s form factor. I am sort of thinking of making a board capable of 35w of one desired that. This will give builders the option of a easy path to power upgrade. I will also put the monoblocks CRCRCRC PSU on board so that the only connections needed are to a trafo and the source and speakers.
I think if you wanted to compare to the other flavors of M2X, simply take an Aksa Lender pre DB and adapt it to fit the 4 standoff connectors of M2X. You could add a jumper to bypass the Edcor. I am not sure if this build is going to be able to accommodate the M2X DB’s form factor. I am sort of thinking of making a board capable of 35w of one desired that. This will give builders the option of a easy path to power upgrade. I will also put the monoblocks CRCRCRC PSU on board so that the only connections needed are to a trafo and the source and speakers.
In the same amp board, like AKSA modules (NAKSA 80, SAKSA 85, MAYA 2000...)?
JDS Labs Atom Amp – Official Release & Benchmarks – JDS Labs Blog
Last edited:
Off topic
While I was searching the NAKSA 80 picture I have found:
The AKSA story:
[PDF] http://www.aesmelbourne.org.au/wp-content/media/THE%20AKSA%20STORY.pdf
While I was searching the NAKSA 80 picture I have found:
The AKSA story:
[PDF] http://www.aesmelbourne.org.au/wp-content/media/THE%20AKSA%20STORY.pdf
Having the PSU rectifier bridge, caps and CRCRCRC on board make the overall amp assembly much easier as the interconnects from trafo to PSU and PSU to amp are many. Furthermore, having the supply rail only a few cm's away from the MOSFET drain pins ensure the lowest impedance path possible. With proper star grounding, the noise on this type of layout is very low. Bass authority is superb with the all important ultra-low ESR path from the last capacitor to Drain.
Yes, it is like NAKSA, SAKSA, etc... not a bad way to go. Fetzilla too I think.
Yes, it is like NAKSA, SAKSA, etc... not a bad way to go. Fetzilla too I think.
Here is a 35w variant with +/-27v rails and 2.81A bias current:
Here is predicted FFT for 8wrms into 8ohms, looks much better now:
Here are predicted distortion components and THD is about 0.02% but dominant H2 and lower H3, 8wrms into 8ohms:
For 1W into 8ohms is is now 0.005%THD
Here is predicted FFT for 8wrms into 8ohms, looks much better now:
Here are predicted distortion components and THD is about 0.02% but dominant H2 and lower H3, 8wrms into 8ohms:
For 1W into 8ohms is is now 0.005%THD
Attachments
Last edited:
Hmm... same feeling here.
Look forward to updates!
hehe, having fingers crossed for a Christmas project
I have been using the M2 output stage this way, being driven by various preamp circuits for a pretty long time. I have used the M2 being driven directly from the CD player with just a volume pot inbetween. You get a taste of what the M2 stage can do by iteself versus how every input stage invariably reduces macro dynamics. I find symmetrical input stages perform better in this type of experimentation.
The only difference is that I use a regulated supply for the M2 output stage.
So far I have found the CD player, pot, M2 combination to sound the best, driving Fostex Sigma 168s in a Quarter Wave Tapered Tube.
The only difference is that I use a regulated supply for the M2 output stage.
So far I have found the CD player, pot, M2 combination to sound the best, driving Fostex Sigma 168s in a Quarter Wave Tapered Tube.
Nelson Pass: The Importance of Phase (in H2)
Off course, if the power amp is H2 predominance, the others chain elements must be very clean to do not modify the harmonic profile (amplitude and phase too).
H2 thread by Nelson Pass
[PDF] http://www.firstwatt.com/pdf/art_h2.pdf
=> The Importance of Phase
.jpg)
mh-audio Tapered Quarter Wave Tube
Off course, if the power amp is H2 predominance, the others chain elements must be very clean to do not modify the harmonic profile (amplitude and phase too).
H2 thread by Nelson Pass
[PDF] http://www.firstwatt.com/pdf/art_h2.pdf
=> The Importance of Phase
Fostex FE 168 EZ Sigma Fullrange buy at hifisound.de 94.5 dB
mh-audio Tapered Quarter Wave Tube
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.