Hi John, it looks reasonable, but nothing special. And it contains the standard mistake, the vbe mutliplier is built the way that the amp will blow when pot-wiper looses contact...
But, it will never give 100watts into 8ohm. Power output is quite similar to the symasym.
Also, bass might be weak with the caps chosen.
The IRFs will work.
Mike
But, it will never give 100watts into 8ohm. Power output is quite similar to the symasym.
Also, bass might be weak with the caps chosen.
The IRFs will work.
Mike
you CANNOT (well, I suppose you could , but you SHOULD NOT, if you want to be successful 🙂 ) substitute vertical/hexfets (IRFP240 and IRFP9240) for lateral mosfets (2sk1058, 2sj162) unless you do a fair amount of redesign. They have different thermal characteristics. To keep it simple, the hitachi parts have a zero tempco bias point around 100mA, the hexfet/verticals have a zero tempco bias point of several AMPS!!! Since we will not be biasing them at that current (well , maybe Nelson will 😉 ), you need a bias regulator of some sort. The Vgs or Vbe multiplier mounted in thermal proximity to the output devices suggested by some folks is one way to do that.
the biasing arrangements is one area where the lateral fets are simpler/easier to use. Set your circuit to provide Vgs bias on the output devices to set quiescent drain current to apprx 100mA per device (somewhere around 0.5V Vgs). since this is the zero tempco point, you can use a simple variable resistor to set the current. this is not perfect, but it works and is simple.
there is more but let's keep it simple ...
slone's amps use lateral mosfets ...
mlloyd1
the biasing arrangements is one area where the lateral fets are simpler/easier to use. Set your circuit to provide Vgs bias on the output devices to set quiescent drain current to apprx 100mA per device (somewhere around 0.5V Vgs). since this is the zero tempco point, you can use a simple variable resistor to set the current. this is not perfect, but it works and is simple.
there is more but let's keep it simple ...
slone's amps use lateral mosfets ...
mlloyd1
jmateus said:
ilimzn said:
It is not the first time when you perform an attack on personality when you disgree with some technical solutions. I'd suggest you to take some courses in colledge, it helps.
Building amps with FET outputs I've discovered that they work well on microcurrents, so high bias is not needed at all. It is the fact that can't be denied by any moderators.
When I mentioned a belief, I mean belief that class AB is better than class B. It was true for tube amps without feedback (or with low gain in the feedback loop) long - long tme ago, when class AB assumed a common cathode resistor for output tubes that introduced a negative feedback at least for DC current. In case of symmetrical transformerless transistor output, no matter is it FET or BJT output, it is completely different story, and class AB means more idle current and LESS stability of a working point without audible gain in the sound quality.
Well, I didn't take any courses on electronics when I was in "colledge", but the differing tempcos look like a serious issue, at least to my admittedly limited level of understanding.
With micro idle current, what does the distortion spectrum look like?
With micro idle current, what does the distortion spectrum look like?
SY said:
Try 10 MA VS 100 MA and compare. Also, current damping was very popular in England in the beginning of 1980'th, when class A driver with N times of voltage amplificsation of the overall amp supplied output through resistor of N times of a load impedance. The result was excellend sound with near zero idle current of output transistors.
Current dumping (not damping) topology is a completely different story and does in no way correlate to the actual subject: Using IRFs without thermal compensation.
Mike
Mike
But that's a totally different circuit. And bipolar, as well; that doesn't mean I can drop the idle current of (say) an Adcom 555 to 1mA and expect it to sound "excellend."
MikeB said:
Gentlemen, calm down and read my original posting!!!
I proposed LESS idle current WITH thermal compensation! Stop flame please and read carefully, not WHO writed, but WHAT writes!
Current dumping IS corellated with biasing of output transistors DIRECTLY.
SY said:
It is the totally different story. If I drop current in some control circuit of my Nissan Armada I can't expect any good result as well. 😉
The question was, do we need 100 MA bias of output complementary FETs IRFP240 / IRFP9240, or not. The answer is, we don't, despite there is a belief about "class AB sounds better than class B".
I can't help it, but the actual subject was using IRFs without thermal compensation... You should have noted that you changed subject to idle bias in current dumping amps. This way it sounded like you proposed using the IRFs in actual discussed circuit would be absolutely fine, and using zero bias there also. No wonder Ilimzn exploded...
The only one upset right now is you.
Mike
The only one upset right now is you.
Mike
MikeB said:
No, it was not. The original subject was, what to do with complementary paris of FETs, how to use them for the power amp. Pderiod.
If we helped the author of the topic to figure out what to do with them, we did a good job.
At least, now he may assume that he can use them in the output stage of the transformerless amplifier with symmetrical distortions with less idle current than is believed.
The next question is, how.
Hi,
and end up with a blown amp and he will probably blame the whole Forum when he should be blaming one advisor in particular.
Hi Wavebourn,
I'm sure you will have one of two options. Horrible sound, or dead outputs (unless a fuse blows first). If you are really lucky, both will happen.
-Chris
This post shows that you haven't actually built anything with the IR devices. The following comments from many others, including MikeB and djk are both correct and accurate. Please, do some experimenting or build this and report on what your results are.
I'm sure you will have one of two options. Horrible sound, or dead outputs (unless a fuse blows first). If you are really lucky, both will happen.
Yes, and those posts that have you upset speak to those things.
-Chris
AndrewT said:
Sure, that's why I humble asked, does he really need 100 MA, or 10 MA is enough?
In my 500W amps output transistors are starving, as somebody may believe, so small is an iddle current. But neighbours tell that they think it is the real piano playing in my house when I play a CD... 😉
anatech said:
I did, Chris. Both class A and class C amps with IR devices. In case of class A they are used as source followers loaded by CCS, and in case of B class amp they are symmetrical source followers with very low iddle current. However, a driver before them works in class A with 0.5A current supplying current to huge input capacitances of output FETs. The second amp I use now, while the first one stays in garage waiting for cold winters. 😉
--Anatoliy
Hi Wavebourn,
Using IRFXXXX, he does need a bias current close to 100 mA per device. You can get away with less, but there is a sonic penalty for doing so. Bias control is also much more difficult. I've tried, so I know.
When you are outside a house, all kinds of sonic evils can be commited and it may still resemble the intended sound outside.
-Chris
Using IRFXXXX, he does need a bias current close to 100 mA per device. You can get away with less, but there is a sonic penalty for doing so. Bias control is also much more difficult. I've tried, so I know.
When you are outside a house, all kinds of sonic evils can be commited and it may still resemble the intended sound outside.
-Chris
anatech said:
Yes, if he blindly uses them in the topology designed for BJTs.
Recently I tried to solve the problem with biasing of BJT followers driven by FET followers (yes, cheap ones from IR), but I gave up this idea.
- Status
- Not open for further replies.