I used a 5W zener diode from Mouser. I wasn't concerned about brand. 5W was also overkill for power capability. I think the zeners I bought were by Onsemi.
I currently use a DC-driver like post 513, based on a what is called B1 buffer - a pair of 2sk170/2sj74 biased to -7v5, with a cap in front. It works.
I have just made a version with BSP129 that will allow a transformer coupling with a low DCR that should abosorb the smallish or largish gate leakage too (indeed across 10K Rg there is drift)
And yes you also do nee a large cooler:
I have just made a version with BSP129 that will allow a transformer coupling with a low DCR that should abosorb the smallish or largish gate leakage too (indeed across 10K Rg there is drift)
And yes you also do nee a large cooler:
How many 2SK180 parts do you have? If you have enough, you might consider making this a balanced amp so you can get rid of the DC blocking cap in the output. Also, cross-coupling the two choke loads on the same core will improve the linearity. See SuSyLu etc.
A balanced amplifier is not necessary to avoid the DC blocking capacitor at the output. A balanced amplifier also reduces or even cancels even order harmonics, not something that many would appreciate.
Here is something that I modelled in LTSpice that eliminates the output capacitor. Instead of choke loading, it is CCS loaded with a bipolar power supply.
Here is something that I modelled in LTSpice that eliminates the output capacitor. Instead of choke loading, it is CCS loaded with a bipolar power supply.
Ben,
Exactly. Having said that the F4, with it's very low distortion still contains some H2 sweetness. When I switch from my SE tube amps to the F4 I don't miss the distortion at all. Amazingly, it just sounds sweet and clean at the same time (of course): Papa's power follower is be the best SS thing I've ever built, it's just brilliant sounding. But yeah, it's never going to sound like SE, but it's not far off, unlike other SS PP designs I've built or listened to.
Don't worry, I'm just going to build it the way you originally designed it with the choke. As a tube amp builder, I love using chokes!
Cheers,
Greg
Exactly. Having said that the F4, with it's very low distortion still contains some H2 sweetness. When I switch from my SE tube amps to the F4 I don't miss the distortion at all. Amazingly, it just sounds sweet and clean at the same time (of course): Papa's power follower is be the best SS thing I've ever built, it's just brilliant sounding. But yeah, it's never going to sound like SE, but it's not far off, unlike other SS PP designs I've built or listened to.
Don't worry, I'm just going to build it the way you originally designed it with the choke. As a tube amp builder, I love using chokes!
Cheers,
Greg
two different beasts
common source, to investigate and enjoy in Triodish characteristic of Biguns
common drain ..... well, I'm not so sure there is strong enough technical reason** to use them that way - you can practically get everything the same (or even better) using damn IXYS puck
**if pure wish is to have SIT in amp, that's certainly proper reason ....
common source, to investigate and enjoy in Triodish characteristic of Biguns
common drain ..... well, I'm not so sure there is strong enough technical reason** to use them that way - you can practically get everything the same (or even better) using damn IXYS puck
**if pure wish is to have SIT in amp, that's certainly proper reason ....
That's a great reason!
There are many people enjoying their diyAudio Sony VFET followers! Also the SissySIT.
I have rebuilt my Classical Choke Loaded 2SK182 Amp (a champ) to run with a SMPS. It is great. I have a 6Amp common mode choke on the input, 0,1 ohm each winding, and some mhenries.
The drain-loaded circuit "distorts" like hell, but is very stable at that. It is 2nd harmonic, and the distortion is stable at a very wide range of bias.
I tested 7 Volt RMS out at 1.000 Hz and it was 4,6%. Aalso 4,6% at 1,8 Amp bias, 2 amp, 2,5 amps: as I change the operating point it stays the same. That means that the 2SK182 is very stable in the distortion it gives, that the distortion is independent of the actual signal.
At 5 V RMS it is 4%
At 1,5V RMS it is 1,2% which means the First Watt is less than 1%.
And yes, would I have an impedance converter (a transformer that is) then the distortion decreases a lot because it was seen that at 4 ohms the distortion rises (see the sawtooth: the blunt saw is a 4 ohm trace). I have drain loaded, a very heavy-duty choke. Finding a gapped transformer with the high current capacity is very very hard. So I will look for one without DC. So a transformer would be a simple 64/8 ohm transformer 'for transistor' that is, without a gap, that can be used after the output capoacitor (coupling cap). I have to look for one.
The drain-loaded circuit "distorts" like hell, but is very stable at that. It is 2nd harmonic, and the distortion is stable at a very wide range of bias.
I tested 7 Volt RMS out at 1.000 Hz and it was 4,6%. Aalso 4,6% at 1,8 Amp bias, 2 amp, 2,5 amps: as I change the operating point it stays the same. That means that the 2SK182 is very stable in the distortion it gives, that the distortion is independent of the actual signal.
At 5 V RMS it is 4%
At 1,5V RMS it is 1,2% which means the First Watt is less than 1%.
And yes, would I have an impedance converter (a transformer that is) then the distortion decreases a lot because it was seen that at 4 ohms the distortion rises (see the sawtooth: the blunt saw is a 4 ohm trace). I have drain loaded, a very heavy-duty choke. Finding a gapped transformer with the high current capacity is very very hard. So I will look for one without DC. So a transformer would be a simple 64/8 ohm transformer 'for transistor' that is, without a gap, that can be used after the output capoacitor (coupling cap). I have to look for one.
Attachments
Yes, drain loaded common source will have those distortion numbers.
Are you running a fairly low Vds?
I have found that changing Vds changes harmonic distortion. Higher Vds gives higher maximum power output and has lower distortion at a given power output.
On another note, the weather has warmed up so this morning I took the THF-51S common drain Mu follower out of the system and in went the common drain choke loaded 2SK180. It is sounding real good.
Are you running a fairly low Vds?
I have found that changing Vds changes harmonic distortion. Higher Vds gives higher maximum power output and has lower distortion at a given power output.
On another note, the weather has warmed up so this morning I took the THF-51S common drain Mu follower out of the system and in went the common drain choke loaded 2SK180. It is sounding real good.
Hi Ben,
What dose the power supply rejection ratio PSRR look like referencing the GND (load across the inductor) vs making the load across the SIT?
Thinking about this for a summer rig.
Thanks,
Hi Botte,
I don't have numbers but in theory with SITs, connecting their drain to ground is the quietest.
In practice, my 2SK180 follower with the power supply connected to the drain is very quiet. With my ear to my 103dB sensitive speaker I do not hear any noise. I believe that is because of the very quiet CLC power supply and also because of the attention to keeping loop areas very small (attention to wire routing, twisting of wires etc), and keeping AC power away from DC and low level signal.
So by keeping the power supply and power delivery quiet, PSRR becomes much less of a factor.
One noticeable difference between grounding the drain versus grounding the inductor is that the grounded drain seems to have a bit of power-up thump whereas the grounded inductor does not. I have a 2SJ28 inductor loaded follower with the drain grounded and it has a quiet power-up thump. However even with my sensitive speakers it is not loud so it is not objectionable to me.
I don't have numbers but in theory with SITs, connecting their drain to ground is the quietest.
In practice, my 2SK180 follower with the power supply connected to the drain is very quiet. With my ear to my 103dB sensitive speaker I do not hear any noise. I believe that is because of the very quiet CLC power supply and also because of the attention to keeping loop areas very small (attention to wire routing, twisting of wires etc), and keeping AC power away from DC and low level signal.
So by keeping the power supply and power delivery quiet, PSRR becomes much less of a factor.
One noticeable difference between grounding the drain versus grounding the inductor is that the grounded drain seems to have a bit of power-up thump whereas the grounded inductor does not. I have a 2SJ28 inductor loaded follower with the drain grounded and it has a quiet power-up thump. However even with my sensitive speakers it is not loud so it is not objectionable to me.