With laterals you need to use TL431 with 1.25V reference (not the 2,5V reference version)
Personally if it were me using laterals I would use 4 pairs not 2 and Increase bias to around 1.6A to 2A.
Personally if it were me using laterals I would use 4 pairs not 2 and Increase bias to around 1.6A to 2A.
Last edited:
Why in a f5 the laterals can take 1A per pair but in f4 you need 3 pairs for 1A total? Can't I build the f4 with 1 pair and bias it to 1A maybe reducing psu voltage I need only 5w output
F5 has feedback
F4 does not have feedback.
Laterals have lower transconductance values.
If you want similar performance specs ( mostly damping factor) it takes more mosfet pairs.
You can throw a single pair in there if you want.
F4 does not have feedback.
Laterals have lower transconductance values.
If you want similar performance specs ( mostly damping factor) it takes more mosfet pairs.
You can throw a single pair in there if you want.
Yes damping factor without feedback is lower.
But unity gain flower has higher damping as the equivalent no feedback f5?
But unity gain flower has higher damping as the equivalent no feedback f5?
Not sure why are you asking about a hypothetical zero feedback F5?
Source follower is a better conguration if you don't need voltage gain.
Source follower is a better conguration if you don't need voltage gain.
Last edited:
Pass M2 is also a unity gain buffer stage but with only one mosfet pair...
My previous post question was:
Will an f4 with one output pair havea better damping factor than an hypothetical f5 with zero feedback?
Another question, I will use the TL431 bias schema of the ba1 buffer, can i omit source resistors? With laterals it is convenient as the source is connected on the case, no need to use mica insulator, the heats ink will be the output
In any case thanks for the helpfull answers, I am quite new in Amp building
My previous post question was:
Will an f4 with one output pair havea better damping factor than an hypothetical f5 with zero feedback?
Another question, I will use the TL431 bias schema of the ba1 buffer, can i omit source resistors? With laterals it is convenient as the source is connected on the case, no need to use mica insulator, the heats ink will be the output
In any case thanks for the helpfull answers, I am quite new in Amp building
F5 sans feedback is pretty high gain thingie , so you can't compare left sneaker and right boot
yes , with TL431 you can skip source resistors , but only with laterals , due to their thermal behavior
verticals will go poof! in most cases
yes , with TL431 you can skip source resistors , but only with laterals , due to their thermal behavior
verticals will go poof! in most cases
F5 is not source follower.
The drain resistance is almost always higher than source resistance speaking from my limited experience.
So source follower will have better damping.
But a single lateral pair will not have better damping than M2 output stage using the original devices.
The transconductance of laterals are much lower.
The drain resistance is almost always higher than source resistance speaking from my limited experience.
So source follower will have better damping.
But a single lateral pair will not have better damping than M2 output stage using the original devices.
The transconductance of laterals are much lower.
You can omit source resistors.
I already answered that question.
You can also do it with irfp devices using tl431 but it takes more careful and thoughtful planning. Not recommended for new builder.
Laterals are easy.
I already answered that question.
You can also do it with irfp devices using tl431 but it takes more careful and thoughtful planning. Not recommended for new builder.
Laterals are easy.
thanks for all enlightening answers.
just 2 more questions:
is the correct F4 schema the one where the Drain of the input Jfet is not connected to the bias resistor network of the IRFP ?
Lowering the PSU to 10 + 10v will increase the damping factor? reduce sound quality?
I need only 2w so 10 + 10 v is plenty...
just 2 more questions:
is the correct F4 schema the one where the Drain of the input Jfet is not connected to the bias resistor network of the IRFP ?
Lowering the PSU to 10 + 10v will increase the damping factor? reduce sound quality?
I need only 2w so 10 + 10 v is plenty...
do not go bellow 15V rails , when dealing with mosfets
less to worry about capacitances , in whatever arrangement they''re connected
less to worry about capacitances , in whatever arrangement they''re connected
ok for 15V rails then
I plan to build only the output stage like the BA-2
will single 30V psu and output capacitor work?
I could reference the input and output to -V
no midle point (0) needed ?!
I plan to build only the output stage like the BA-2
will single 30V psu and output capacitor work?
I could reference the input and output to -V
no midle point (0) needed ?!
Last edited:
Well to answer your question. if you are using stock pcbs the jfets won't correctly bias like that so I would advise against it.
You would need to modify the circuit somewhat. I don't think the modification would bring improvements so it's a wasted effort in my opinion.
You would need to modify the circuit somewhat. I don't think the modification would bring improvements so it's a wasted effort in my opinion.
They help define where GND is in respect to the rails.
The bias circuit, which at first glance appears to be centered at GND is actually is a few volts below, and needs to float. So the outputs need another reference to show them were GND is, and that, with the diodes, do the job.
The bias circuit, which at first glance appears to be centered at GND is actually is a few volts below, and needs to float. So the outputs need another reference to show them were GND is, and that, with the diodes, do the job.
in the BA pushpull output buffer these are not present but still ofset near 0 ?!
How does M2, F4 and BA pushpull output buffer sound compare?
How does M2, F4 and BA pushpull output buffer sound compare?