It was set at 10mV over the gate resistors
OK, that would mean around 40mA on each resistor. Fine. I had overbiased it to 59mA.
Anyway, I'm getting this glitch on the sinusoid using 1KHz. This is not the first time I see glitches on the LTP out, and it usually gets right after the VAS.
I can send the asc file for you to check if I'm doing something wrong.
Attachments
And I have no doubts it should sound good, because many people are building Apex amps instead of buying unreliable amp kits on eBay.
My intention is not to criticize Apex's designs, as I find them very elegant and simple, which can be misleading on their sophistication.
I'm doing this research on several amps, old and new.
If I intended to criticize something I would PM Apex, and no put my questions here. He's a generous person who is sharing his knowledge with us, and that's not easy to find in this world.
My intention is not to criticize Apex's designs, as I find them very elegant and simple, which can be misleading on their sophistication.
I'm doing this research on several amps, old and new.
If I intended to criticize something I would PM Apex, and no put my questions here. He's a generous person who is sharing his knowledge with us, and that's not easy to find in this world.
mosFET output stage do not have the equivalent to BJT optimal ClassAB bias.
MosFETs start with lots of crossover distortion when bias is very low, reducing to some crossover distortion when bias is high and gets to very low crossover distortion when the bias is getting near ClassA. Only at full ClassA do BJT and FET output stages have zero crossover distortion.
Summary.
BJT - bias to ClassA, or to optimal ClassAB
FET - bias to ClassA, or as hot as the heatsinks will allow in ClassAB.
That's exactly what I said and mean.
To come to a forum and get into a brawl for matters of ego, and not just talk out your opinions in a nice conversation, where everybody wins, is plainly foolish.
BTW: the FH12 mod2 runs very well and specs are exactly as said.
THD @ 20KHz was a bit less: 0.0099%. You can increase input level a bit more, like 1.5v, before the amp clips.
The asc sim schematic is a bit different from the pdf one.
Andrew: you're right on how bias affects distortion on this FET amp and others. You just play with the bias and see that.
To come to a forum and get into a brawl for matters of ego, and not just talk out your opinions in a nice conversation, where everybody wins, is plainly foolish.
BTW: the FH12 mod2 runs very well and specs are exactly as said.
THD @ 20KHz was a bit less: 0.0099%. You can increase input level a bit more, like 1.5v, before the amp clips.
The asc sim schematic is a bit different from the pdf one.
Andrew: you're right on how bias affects distortion on this FET amp and others. You just play with the bias and see that.
BTW, I had another look at the pdf file, and it does spec 148w @ 8 ohms. Which is what I got on my sim when I increased the input level.
For those that know LTSpice: how is it possible that output result when the supply is just +/- 60v?
I ask this because I did get similar results with other amps. Is LTSpice showing fake results?
For those that know LTSpice: how is it possible that output result when the supply is just +/- 60v?
I ask this because I did get similar results with other amps. Is LTSpice showing fake results?
I probably sounded like I'm ignorant,that wasn't my intention,I've built a lot of Mile's designs and I respect him sharing his knowledge with us,I answered like I did because I can't check that simulation(don't know how to do it),and that's why I referred to Mile.
I'm sorry if I offended anyone,wasn't my intention.
I'm sorry if I offended anyone,wasn't my intention.
If you did not like opened your mind, it is OK
I like every thing that bring new idea for me. Although, if it is not their original idea, i don't care. I learned from many member in this forum: Mile (Apex), Hugh Dean (Aksa), Nelson Pass, Dadod, Bonsai, Keantoken, Bob Cordell, Douglas Self, etc.
Sometime, I share my idea, too.
Bimo,
What does it exactly mean "THD at 148w/8ohm"?
If that output is not possible with the actual amp, which should be clipping with that input level and such a supply, how should it be interpreted? A fantasy output?
As I said, this is not the first amp I found that shows such results.
a 230:dual 40Vac transformer operating on a UK supply of 240Vac gets a 3pair output stage to ~ 37.3Vac (52.7Vpk or 105.5Vpp)
This is equivalent to 174Watts into 8ohms, from a PSU+quiescent amp that idles @ ~ +-58.5Vdc
This same amplifier (Leach Lo Tim) managed 49.9Vpk into 4r0, clip free.
With 60Vdc you should be able to get even higher clip free output.
Last edited:
In simulation, I use power supply with zero output impedance or in reality it use regulated power supply.
P = V x V / R, P = power, V = Voltage, R = load impedance
148 = V x V/8
V = 34,4 rms
V peak = 34,4 x 1,414 = 48,6V.
Because it is not using bootstrap at VAS then V peak max for PSU 60VDC is about 52V. Then 48,6V is below clipping.
Of course if you use non regulated PSU, clipping voltage must below 52V. Maybe clipping voltage is around 47V because voltage drop on PSU.
Here you are RIA phono pre by Apex.
Attachments
Thank you.I found them all.MM28 and MM22 use many parts.Is there a PCB layout for MC10 and MM4?
Which one do you prefer among all the designs?Which one sounds better?
