F4 low power
Hi,
I'm thinking of building a F4 that max out at around 12W, because 10W is enough for me and my source+preamp gives only 14V anyway (2V DAC out, x7 preamp gain), and I want to avoid unnecessary heat dissipation.
I modified the component values a bit, got the attached and it simulated OK in LTSpice, but I would like some comments from the experienced ones, to see if I made some stupid changes.
Hi,
I'm thinking of building a F4 that max out at around 12W, because 10W is enough for me and my source+preamp gives only 14V anyway (2V DAC out, x7 preamp gain), and I want to avoid unnecessary heat dissipation.
I modified the component values a bit, got the attached and it simulated OK in LTSpice, but I would like some comments from the experienced ones, to see if I made some stupid changes.
- Q1/Q2 are 2SK170/SJ74 as I don't have 2SK370/SJ108;
- MOSFETs are 2SK2013/SJ313 as I have some, and they seem to perform better at lower bias than the IRFs;
- Preamp is LuminAria, it can swing 60V with 50mA current, but the gain is only x7 and I don't need/want to add another gain stage;
Attachments
Ofcourse you can lower voltage rails and bias current to have lower power amplifier, but performance will be worse. Why? Because F4 has no error correction loop and MOSFETs are more linear at higher currents and higher voltages (less non-linear capacitance), especialy at rail voltages lower than 20V non-linearity starts raising quite a but, just look at datasheets. So yes, your amplifier will work and can sound good, but it will not sound as good as original F4.
Last edited:
If "better than" means "not as good as" you are right.
I tested k2013/j313 thoroughly and built many circuits with them (F4 included).
With three pairs, supplied with +/-20V and 400mA through each pair it sounds noticeably better than original F4. Lowering the PS voltage to +/-16V changes nothing except the output power.
I used 0R33 source resistors.
k2013/j313 MOSFETs have lower capacitance, but also lower transconductance. So you have higher output impedance (lower damping factor) than with IRFP240/9240 MOSFETs. Apart from DF, Toshiba MOSFETs are more linear and they do not have IRFP P-channel problem.
Thanks all for your comments.
Hi Juma, this is very reassuring, thanks
. Given your comments I've decided I'll build it, instead of just "thinking of" in my earlier post.
Hi kacernator, the simulated THD1k 8R@1W is around 10ppm, and output impedance is around 0.66R which gives a DF of 12 - I think it's not too bad, though it's true that the original F4 have a higher DF of 40.
Hi Juma, this is very reassuring, thanks
. Given your comments I've decided I'll build it, instead of just "thinking of" in my earlier post.Hi kacernator, the simulated THD1k 8R@1W is around 10ppm, and output impedance is around 0.66R which gives a DF of 12 - I think it's not too bad, though it's true that the original F4 have a higher DF of 40.
Well, I think Q2 is in wrong. Generally though, these "short channel" JFETs work the same in both directions
When I had my build of F4, I tried cap multiplier PSU too. Tried it with BJT and MOSFET and I prefered sound with just CRC passive filter by a big margin. I was suprised that the cap multiplier affected sound so much. But after some more experiments I discovered, that lot of parts in the circuitry (not just PSU) affected sound somehow. I gues it is because F4 has no error correction. Don't get me wrong F4 sounds quite good, apart from low power and high heat, F4 lacks resolution especially on complex music (for example rock, metal). F5 beats F4 in this regard.
Hi there.
I have two very basic questions (because I cannot cope with some 420+ pages of info in this thread):
Would my McIntosh C48 pre fit well with a FirstWatt f4 diy build?
Do I need to use a Pass BA-2 VAS stage in order to avoid problems?
Thanks for your patience!
Kind regards,
Peter
I have two very basic questions (because I cannot cope with some 420+ pages of info in this thread):
Would my McIntosh C48 pre fit well with a FirstWatt f4 diy build?
Do I need to use a Pass BA-2 VAS stage in order to avoid problems?
Thanks for your patience!
Kind regards,
Peter
No, the C48 would not work well with the F4. It doesn't have enough gain or produce enough voltage.
You're better off building one of the other First Watt amps, or one of the Burning Amps.
Why did you choose the F4 (which has no voltage gain)?
Thanks Mr Pass. I have made a separate thread for my build with this question as the humble beginning, so I will butt out of here. I got another useful answer there which indicates that I would not be able to use the full power of the F4 with the McIntosh C48. Thanks also to ArthurDK = the f4 parts supplier! Kind regards, Peter
Last edited:
There's an easy fix -- get a pair of Jensen JT-123-BLCF transformers. They can be wired for 1:2 and are very high quality. The McIntosh should be more than adequate for the task.
You may be able to find less expensive transformers of excellent quality on EBay, like UTC, Cinemag...
I just checked the specs of the C48. It has 15dB voltage gain.
If you use a cd player with 2V rms that is certainly enough gain for F4. You may as well try that first before adding extra gain stages for another 6dB.
If your speakers are 90dB or higher you probably won't need the extra gain.
Last edited: