NVMOS amplifier

Status
Not open for further replies.
Jump to Latest
#41
Re: Re: Not that smart .....but

Workhorse said:


Hi Quasi,
Bootstraping is used to boost the drive signal to get maximum swing. This is how it works.
in no signal conditions the cap will be charged at a refernce voltage between the resistor and output. When the lower mosfet starts conducting the the out put swings negative and then the Voltage across cap will be Vref+Vnegswing and then it gets charged to high voltage and again when positive signal appears, the the gate of mosfet sees an elevated voltage due to charge stored by cap and it gets an extra voltage at gate by the virtue of virtual supply emulated by the cap between resistor and output , thus the output swings to rail.

This technique is always used in Class-D drivers to get max High side mosfet swing.

hope you understand it.
Click to expand...

Thanks Kanwar. So the negative swing of the output sinewave is added in part to the available gate drive voltage via the charging capacitor. I got it. Can you please edit my cct with indicative values in the "Another quasi-complimentary amp" thread to show me how it would fit in that? For that I would honor you forever etc....

Cheers
 
#42
Sorry quasi at this time,

This type of arrangement is not valid in your design, because the driver stage is itself act as a differential also.
But you could very well adopt my circuit in yours to get this desired effect.
regards,
Kanwar
 
#43
lumanauw said:
Hi, Kanwar,

I think your design of full Nch+folded cascode could end up in a good DIY amp. Please keep up the good work 😀
Click to expand...

Thanks for the supportiveness and thanks for further encouragement.🙂

darkfenriz said:
Sorry Kanwar
I wanted to ask you:
why don't you join pnp driver's COLLECTOR with mosfet's source?
this could change voltage drop on source resistor at HF.
Click to expand...


Suggestion noted!
Thanks for the comments

New update schematic.
 

Attachments

  • gladiator.jpg
    gladiator.jpg
    33.4 KB · Views: 1,693
#44
hi DarkFenriZ,

I think this is what you want to say, its a good approach during HF switching especially at Turn-off. But If i had to use parallel mosfets then is it ok for the pnp driver to stay connected with first mosfet .

regards and gratitude,
Kanwar
 

Attachments

  • ppd.jpg
    ppd.jpg
    5.1 KB · Views: 1,475
#46
Using source resistors goes against rail to rail swings, don't you think? Of course, you have no choice if you go paralelling output devices.
I would not connect the C of the driver to the source of just the 1st MOSFET, as it worsens current sharing at switching. Not much of a problem for resistive loads, but for reactive...
 
#47
darkfenriz said:
Yes, that's it.
wondering if you still need gate resistor now... as slew should be fast and pretty symetric compared to other types of driving.
best regards
Click to expand...


Hi DarkfenriZ,

In reality one must use gate resistor or Ms Oscillation would welcome unintendedly...............😉
Thanks for the information.....
and thanks again for your effort

What are your views regarding the new biasing regulator and push-pull drivers .....

best regards,
Kanwar
 
#48
Hi darkfenriZ,

I simulated the design but, its slew rate is nod good and again it has very low open loop gain , thats why i am going to implement second differential + cascode loading in the place of folded cascodes, thereby facilitating the increase in gain and elimination of miller effect to some extent as well.

what do you think about this.

regards,
Kanwar
 
#49
Hm... strange
everything seemed OK.
push-pull drivers were good idea,
bias regulator did not provide thermal stability, but it shouldn't affect slew
maybe you've overcompensated fequency?
I'll try to simulate it too and maybe change some bits and pieces.

regards
 
#50
darkfenriz said:
Hm... strange
everything seemed OK.
push-pull drivers were good idea,
bias regulator did not provide thermal stability, but it shouldn't affect slew
maybe you've overcompensated fequency?
I'll try to simulate it too and maybe change some bits and pieces.

regards
Click to expand...



Yeah , the PP drivers were a good idea😀
The bias regulator has nothing to do with slew or gain.
I donot even use the compensation cap at the feedback till now.

Waiting for your sim results................

cheers,
kanwar
 
#52
Hi darkfenriZ,

Have you come up with your simulation results, just waiting for them...........
The folded cascodes are just good for less phase shift at high frequencies , but the open loop gain still suffers alot, i need to check it throughly... but very busy, but might give some precious bits of time to this DIY amp and it will be a project in reality very soon.

reply soon..... waiting🙂
 
#54
back again.......

just going to test the very much appriciated design of mine.......😀 😉

input -> differentials with cascode loading....
VAS -> again differential with something extra....++
Driver stage -> DarkfenriZ Current Mirror style or Kanwar custom driver ++ pushpull effects...
Output -> all of you know that will would be our N-Channel mosfets

regards..
Kanwar 😉
 
#58
hi dark fenriz,

Your current mirror decreases the openloop bandwidth and also causes stability problems in high frequency region....

rolling back to push-pull drivers and folded cascode increases the performance and using regulated bootstraped driver supply eliminates the seperate supply issue and solved the prblem alot....

regards,
K a n w a r
 
Status
Not open for further replies.
Top Bottom