I want to choose whether to go with BJT or MOSFET OPS. I want to know what is the sound difference between BJT and MOSFET output stage. Also difference between lateral and vertical MOSFET.
Thanks
Thanks
It's not just about the output stage, the rest of the amp needs to be designed in accordance with what output devices you want to use. So you need to think at a system level what are your goals - cost, size, efficiency, performance. It's a trade.
Mosfets tend to like to operate at higher temperatures but have higher input capacitance.
Mosfets are easier to drive too, but it doesn't mean you dont need driverstage...
Lateral mosfets are quite stable and in some designs Vbe multiplier is unnecessary
And I'm definitely not an expert on this but I like BJT as much as MOS FET designs.
Mosfets are easier to drive too, but it doesn't mean you dont need driverstage...
Lateral mosfets are quite stable and in some designs Vbe multiplier is unnecessary
And I'm definitely not an expert on this but I like BJT as much as MOS FET designs.
Thanks for the replies. My question is assuming the VAS and the bias spreader is optimize. Just concentrating on the sound of difference between BJT and MOSFET.
I am not asking about technical design difference, just the sound.
Thanks
I am not asking about technical design difference, just the sound.
Thanks
As far as the sound difference is concerned, that is down to the amplifiers designer and this question cannot be successfully answered. There are too many permutations.
If the rest of the amp is designed appropriately to suit, there will not be any sound character that is attributable to MOSFET or BJT, it will be the overall design that produces the signature.
There are good and bad examples of both MOSFET and BJT amps. You would be best to think about your overall goals then look at designs that meet them and ask for opinions on sound from users of them.
There are good and bad examples of both MOSFET and BJT amps. You would be best to think about your overall goals then look at designs that meet them and ask for opinions on sound from users of them.
I had some test BJT & MOSFET output stage with same front-end circuits.
The sound of MOSFET a bit better (clear) around midrange frequency (500 ~ 2KHz). BJT better at bass. Hi-mid & hi are the same. MOSFET have sounding too cool.
My MOSFET is Hitachi/Renesas Lateral MOSFET and BJT is Sanken LAPT.
Regard,
The sound of MOSFET a bit better (clear) around midrange frequency (500 ~ 2KHz). BJT better at bass. Hi-mid & hi are the same. MOSFET have sounding too cool.
My MOSFET is Hitachi/Renesas Lateral MOSFET and BJT is Sanken LAPT.
Regard,
Thanks, this is the kind of comment I am looking for. Everything else being equal.
Can you tell me the part number of the power MOSFET and BJT you use?
Thanks
My MOSFET i used is 2SK1057/2SJ161 pair, driving by 2SD669AC/2SB649AC. Only an output pair.
BJT version is 2SC2922/2SA1216, driving by 2SK2013/2SJ313. One output pair too.
Regard,
BJT version is 2SC2922/2SA1216, driving by 2SK2013/2SJ313. One output pair too.
Regard,
I use the same BJT front end no matter what output stage is used, even tubes though for tubes higher voltage transistors are selected. This is more a question of phase margin and how much feedback. 90° of phase margin and 50dB of feedback with mostly correct any changes of the output stage.
Hi,
The sound differences are all about technical differences
and nothing else, and ignore all subjective opinions,
as they are totally meaningless waffle.
BJT is the best done properly, and is simpler.
rgds, sreten.
The sound differences are all about technical differences
and nothing else, and ignore all subjective opinions,
as they are totally meaningless waffle.
BJT is the best done properly, and is simpler.
rgds, sreten.
Sounds like there is very little to mostly no difference on the sound between BJT and MOSFET power transistor output stage.
But there should be a big advantage with MOSFET because the input capacitance goes up on BJT a lot more than MOSFET as beta of the BJT goes down when swing high and drive large current. That change of beta creates distortion. MOSFET doesn't do that until you get the Vgd very low.
But there should be a big advantage with MOSFET because the input capacitance goes up on BJT a lot more than MOSFET as beta of the BJT goes down when swing high and drive large current. That change of beta creates distortion. MOSFET doesn't do that until you get the Vgd very low.
Member
Joined 2009
Paid Member
From what I have read, built and heard there are no significant difference if you design your amp to suit the devices you choose to use - they should all work well. They have different strenghts and weaknesses. Choose the device type that you know best how to design for.
Subjectively, I have found that I prefer the sound of BJT based amps.
By the way, my best amplifier uses both BJT and FET in the output in order to take advantage of what they both offer ( search the forum for TGM8 )
Subjectively, I have found that I prefer the sound of BJT based amps.
By the way, my best amplifier uses both BJT and FET in the output in order to take advantage of what they both offer ( search the forum for TGM8 )
Last edited:
On the other hand, the BJT has more transconductance than a MOSFET, and because it draws base current, you end up needing a beefier driver stage that is more capable of driving the input capacitance as a side effect.
With the latest audio BJTs I think it is a wash. A MOSFET output stage will have less dynamic crossover distortion at high frequencies but more static nonlinearity, and will dissipate more at idle than a BJT one. The overall THD figures end up about the same. However, with Hawksford error correction you can trade the extra speed of MOSFETs for lower distortion and get very low THD indeed.
One problem I noticed with the new MJL3281/1302 etc. is that they start to slow down drastically when they get close to the rails. For Vce less than about 5v things get pretty sluggish.
Last edited:
Yes, you definitely need a strong driver to drive the MOSFET to avoid loading by the input capacitance. But one advantage is the input capacitance does not change with MOSFET until VDG gets to less than 10V. For BJT, as the voltage swing closer to the rail, current draw from the power BJT goes to a few amps. Beta drop significantly and input capacitance increase drastically, and this is regardless of the headroom to the rail. With MOSFET, if you give enough headroom, you don't have capacitance variation problem, but you always have capacitance variation problem with BJT. So this create more distortion like the crossover distortion that is swing dependent.
I don't know what is Hawksford error correction!!
According to Bob Cordell's book page 233 about using source degenerate resistor of 0.15 ohm for IRFP240 and 0.22 ohm for IFRP9240, the crossover distortion gets much lower if you run about 200mA and up as shown in Fig.11.11. It looks a lot more gentle compare to the Wingspread diagram of the BJT even using the right emitter resistor to get 26mV drop.
It slow down because the beta drop upon high current drawn and cause the input capacitance to go up tremendously regardless of headroom to the rail. That's where MOSFET shine, if you give enough headroom, the CGD can be kept low.
Anyway, I am glad you join in. I am getting a little frustrated with the OPT of the tube amp design. I feel I am at the mercy of the OPT. I never consider Tube amp is superior nor SS amp is superior. I always plan to build one of each. So I might as well spend some time here.
Another thing about MOSFET is the output impedance is higher. from the data of IRF240. At mid point where bias current is about 200mA, gm is about 1.5 to 2S. So the source impedance is 1/1.5 to 1/2 which is about 0.5ohm. Adding 0.15 ohm degenerate resistor, the output impedance is 0.65/2= 0.33ohm. ( both sides on). But if you look at BJT running 200mA, the degenerate resistor to get 26mV is 0.026V/0.2A= 0.13 ohm which is about 0.15ohm. re=26ohm/300=0.087. total output impedance =(0.15+0.87)/2=0.119=0.12ohm. So output impedance of BJT is less than half of MOSFET.
Last edited:
Fwiw, i have never heard a Mosfet amp with the type of bass i like. Elsewhere they may be ok, even very good, just not my favorite type of Krellish bass 🙂 No matter if lateral, vertical, massively parallel or class A.
With me,
Class AB, go to modern, fast BJTs. Or Lateral MOSFETs.
Class A, AAB, go to vertical MOSFET or HEXFET.
My formular: JFET at LTP + BJT or MOSFET VAS/TIS + MOSFET at driver + BJT output = Perfect.
Class AB, go to modern, fast BJTs. Or Lateral MOSFETs.
Class A, AAB, go to vertical MOSFET or HEXFET.
My formular: JFET at LTP + BJT or MOSFET VAS/TIS + MOSFET at driver + BJT output = Perfect.
Not sure why you assume a sub would mask the bass quality of the main amp...
Is the Velodyne bjt or mosfet?
- Status
- Not open for further replies.