DIY Class A/B Amp The "Wolverine" build thread

Now all working perfect .... offset 0.100mv -0.200mv, bias 40-44mv , output ac measure input sort 0.040 mv ac.
Trafos 2X42 Vac gives me about 56Vdc input ac voltage 223Vac.
 

Attachments

  • IMG_20230721_075800.jpg
    IMG_20230721_075800.jpg
    541.3 KB · Views: 154
Last edited:
Note: Slightly off topic but need a bit of a direction taking into account multiple interpretations of course.

I'm driving the Wolverine with an S.M.S.L Sanskrit MK-3 with below specs. It sounds pretty good I've say but effectively I can only control the volume. I'm using the 22K impedance input on the Wolverine. Is it worth/fun to explore other pre-amp solutions or may be only upgrade the DAC?
Any advice or direction (on this forum) appreciated.
1689944086519.png


thx in advance

w.
 
The Wolverine has an AC gain of approximately 27.4 with the 33k configuration. output level of 2 Vrms is ~2.8V peak. This will drive the output to around 76 volts, which is beyond clipping for any of the recommended supply rail voltages. I'm thinking that the DAC and the wolverine match up well. You may need a bit of attenuation on the output to avoid having full scale outputs clip.

From my limited testing, and if I remember correctly, visible distortion (of sine wave on 'scope) shows up around 3.5-4 volts below the rail voltage with no load

You can do software amplitude and equalization. That may eat into the noise margin a bit. You could also use/make a pre-amp based on some of the latest TI audio opamps that have distortion of the same order of magnitude as the Wolverine. See TI 16012 and others See Self "Small Signal Audio Design" for an excellent starting place. His book is based on older opamps, but opamps are usually drop in parts except for possibly the pinout.
 
Hello,
I have constructed the Wolverine amplifier and i am currently making the power supply. I have a 2x42 V transformer and 2 capacitors 20000 uF at 100 V and 2 capacitors 10000 uF at 100 V . Can i connect those different capacitors (10000 uF and 20000 uF) in parallel or the capacitors must have the same capacitance for everything to work properly?
All the capacitors have the same ESR.
 
@danieljw Thanks for the correction. The number didn't look quite right. There is a whole series of very low distortion and relatively quiet opamps. I use the OPA1664 or 1662 in the low-distortion oscillator and for active notch filters. The specs for this series are amazing. Good replacement for the NE5532 if cost isn't an issue.
 
  • Like
Reactions: danieljw
Finally worked through all my late night DIY "issues" I caused for my self and been listening to some music for a few hours or the last three days as I finish up the chassis. Does sound very (extremely) good, so my questions are more for me to gain knowledge and increase my "wheelbase" and not a knock or con for the Wolverine design in any way.

Question: was not expecting this amp to run so warm, so I want to see what others impressions of the heat generated is.
Set-up: I am running a EF3-3 at +/-69Vdc with bias at 42mV, ~0.001V of offset and CCS is ~5.02mA on both boards.
Transformer is a second-hand 800VA Antek 2x46, I was expecting +/-65 but actually getting +/-69 because the no-load AC on secondaries are closer to 52V, so running a little "hot" for a 3-pair Wolverine, but I always live on the edge..

After about 30-45 mins of normal music while working in a small office, the heatsinks are getting 115-120-degree F. So not hot, but they are quite warm, so I wanted to make sure that's normal?
Is this just the combination of the driver current and output current and totally normal? Would hurt the sound a lot to lower the bias down to 40-50mA?

For context and my heat comparison: I have a different amp with 4-pair of the same outputs MJL4281/4302 running at +/-75Vdc on the same sized 4U heatsink (6.25"Hx1-3/8"Dx10"L) and runs about 65% of the temps. The other amp is running 35mA (~0.3A total) on the outputs and this Wolverine is running ~85mA (0.5A total) - unsure of the driver current on this other amp because they are on the heatsink with the outputs, so its hard to know how much heat they generate in that amp.
 
The EF3-3 is running about 50% more power at idle than the other amp, so it's not surprising that it is warmer. Also, the AB amps run maximum temps around 35-40% of maximum power output. 120 F is just under 50C, so it is probably normal. 60C is the usual design limit to prevent operator burns. Also, be sure you are testing with the heatsink orientation normal (both fins and base in vertical planes) and there is space below the heatsink for air flow. Often bench testing has the heatsink in a sub-optimal orientation.

The driver contribution with those output transistors is somewhere between 1 and 2% of the total output, generally closer to the 1% value.

Ideal bias depends on the value of the emitter resistors. You are looking for around 22mV across each output emitter resistor at 25C. That said, the Wolverine seems to be fairly insensitive to bias errors according to testing done by @fireanimal

You didn't say what the speaker nominal impedance is. The 3 pair should be fine with 8 ohms and 69 volts. This works out to around 8 amps total output current across the 3 pairs, so under 3A each. The place to be wary is with that voltage and a 4 ohm or lower load.

You might check your mains voltage, as usually transformers regulate a bit better than what you are implying.