How to design a bootstrapped VAS? - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 21st August 2013, 08:37 PM   #1
danny92 is offline danny92  Portugal
diyAudio Member
 
Join Date: Nov 2012
Location: Lisbon
Default How to design a bootstrapped VAS?

Hello everyone,

I've noticed that some designers use bootstrapped VAS, I wanted to know why they opted for this VAS topology, some designers say that bootstrapping is a way to increase the output voltage swing, other designers say that a CCS VAS or a differential VAS are more linear and have lower THD.

I've started and I'm still envolved with some circuits using chip amps, and I've noticed that some known chip amps have this VAS topology, namely TEA2025, TDA7293/7294, STK series like STK4231 and STK461, TBA820M, LA4450, just to name a few.

Why those known manufacturers (On Semiconductor, Sanyo, ST) opted by using bootstrapped VAS instead of CCS or differential, like some LMs, like LM4766 (National Semiconductor/ Texas Instruments).

I have discussed this topic in a similar post, but this time I wanted to simulate, and perhaps build an amplifier based on this topology, the problem is that I don't have much information about this VAS topology in books.

I'm trying to recreate an amplifier similar to LA4450 for single supplies, but I have lots of THD in the simulation.

Best regards,
Daniel Almeida
Attached Images
File Type: jpg LA4450.JPG (193.7 KB, 315 views)
  Reply With Quote
Old 23rd August 2013, 07:58 PM   #2
brig001 is offline brig001  United Kingdom
diyAudio Member
 
Join Date: Jan 2009
I must admit, I don't fully understand the diagram you posted, anyway...
T44 (? hard to read) is a constant Current source into T17. This current is mirrored in T40, T17 and T40 then have the bootstrap applied to them. I can only think that in this case, the bootstrap is to get the base of T25 above the supply rail, while still keeping the linearity of the CCS and the mirror.

Does that make sense?

On a THD, front, T29 and T30 don't make sense to me, does it get better if you remove them? If not, what about T24? Not sure about its purpose either.

Just looked at it again, the input doesn't look right. T20 is open loop, and how is it biased? Does VG2 have some DC in it?

Brian
  Reply With Quote
Old 23rd August 2013, 08:22 PM   #3
danny92 is offline danny92  Portugal
diyAudio Member
 
Join Date: Nov 2012
Location: Lisbon
Hello and Thank you very much for your help,

I know that this design must have some mistakes, the input must have a 30kohm resistor to gnd for correct dc biasing, but that input is gnd referenced and the reference of the source is gnd too, I don't know if that current mirror is right it's based on TEA2025B and TBA820M by ST but they use the same current also for the input differential stage. T29 and T30 are setting the supply middle point at the output for single supply operation like it's made in the most gnd referenced input single supply chipamps. T24 limits the current in T23 in the case of the current limiter PNP is activated.

Best regards,
Daniel Almeida

Last edited by danny92; 23rd August 2013 at 08:38 PM.
  Reply With Quote
Old 23rd August 2013, 09:51 PM   #4
danny92 is offline danny92  Portugal
diyAudio Member
 
Join Date: Nov 2012
Location: Lisbon
Hi everyone,

I reduced the current at the VAS and now the THD is a little lower (<0.2 % driving 8 ohm), I also have a better output swing, but it still far from the desired. I added another output pair, I don't know if I should use another model of output transistor.

Best regards,
Daniel Almeida
Attached Images
File Type: jpg LA4450_2.JPG (197.4 KB, 227 views)
  Reply With Quote
Old 24th August 2013, 09:56 PM   #5
brig001 is offline brig001  United Kingdom
diyAudio Member
 
Join Date: Jan 2009
Hi Daniel,

Looked at it again, and I have a better idea what is going on now, however, there's a few things I'm not sure about:

How much gain are you going for? R28 is zero, making the gain 52dB (400 times). This seems too high - you would be clipping with 100mV input. I would add a 400 ohm resistor and go for 36dB (lowest the datasheet allows), or try different values to suit. Please check the resistor calc though - it's been a long day. Reducing the gain is increasing the feedback, so should reduce distortion too.

The other things are related to the input transistor and using a current mirror T29 and 30 to give a constant current through R24 to bias the output at half supply. These will both give you distortion, but I'm not sure if they are optimised on the chip to cancel. Personally, I would get rid of T20, T29 and T30 and figure out a different biasing scheme for the input LTP. For this, I would use a stable voltage at half the supply, and connect this through a suitable resistor (10k) to the base of T18 (the new input), and add an input capacitor. This is just my opinion, but looking at the datasheet, the distortion quoted is 0.07%, 0.4% and 0.1% at various points, so your 0.2% looks bang on for that circuit design.

Brian
  Reply With Quote
Old 24th August 2013, 11:04 PM   #6
danny92 is offline danny92  Portugal
diyAudio Member
 
Join Date: Nov 2012
Location: Lisbon
Hello brig001,

There's another ways to set the half supply output voltage without using the voltage divider at the non inverting input?

Best regards,
Daniel Almeida
  Reply With Quote
Old 25th August 2013, 02:07 AM   #7
expert in tautology
diyAudio Member
 
bear's Avatar
 
Join Date: Apr 2002
Location: New York State USA
Why a single rail & ground and not a +/- rail + ground, permitting complimentary output devices and getting rid of the output cap?
__________________
_-_-bear
http://www.bearlabs.com -- Btw, I don't actually know anything, FYI -- [...2SJ74 Toshiba bogus asian parts - beware! ]
  Reply With Quote
Old 31st August 2013, 04:31 PM   #8
danny92 is offline danny92  Portugal
diyAudio Member
 
Join Date: Nov 2012
Location: Lisbon
Hi everyone and thanks for your help,

I've changed the current at the VAS and IPS to be the same, in this case 1mA, and I noticed that the distortion characteristics are better (but with 0.1% THD1), so I've changed the output cap and the capacitor from V- of the differential stage to higher values and now I've distortions of 0.05-0.06 % that is closer to the 0.07% referred in the datasheet.
I've also made a buffer circuit to replace the circuit that sets the half supply voltage at the output and a CFP to improve the output voltage swing, but I don't know if it can work.

I've simulated the THD20 and it gives me aprox. 1% I think that's too much THD20 but the signal on the transient response seams clean, without clipping or crossover distortion.

Circuits attached.

Best regards,
Daniel Almeida
Attached Images
File Type: jpg LA4450_1.JPG (167.0 KB, 128 views)
File Type: jpg LA4450_2_2.JPG (166.7 KB, 120 views)
  Reply With Quote
Old 1st September 2013, 10:00 PM   #9
diyAudio Member
 
Join Date: Nov 2004
Location: Germany
I think that is about how well you can expect a circuit like that to work. These ICs were typically used in portable radios and such after all. You could still give the quasicomp output stage a Baxandall diode, might work better than the CFP.

52.5 dB of gain is quite a lot to ask for, besides I bet that with the level shifting, effective input noise will be quite high (at least that of the 30k resistor R16 for obvious reasons), meaning the amp will be quite hissy. ICs amps like that often gave ~1 mVrms worth of output noise. '80s equipment using them was quite notorious for hiss. Actually I get 1.25 mV @ 20 kHz BW for R16 alone. Ugh.

The level shifter was included to eliminate an external coupling cap and biasing. A neat feature for an IC amp at the time, but not without its problems. Sure you could bridge R16 with a cap, but that would defeat the purpose. I suggest biting the bullet and just using a coupling cap, with a clean input bias voltage from C12 via a high-value R.

That being said, a buffer ahead of the differential pair could still have its uses. Use a very small, high-beta (and constant beta) transistor, with just enough Ic to keep noise reasonable. Should keep input impedance nonlinearity in check when looking at higher source impedances. Then again, if you start adding extra circuitry in the first place, splitting gain across two stages would seem like a brilliant idea. Offload the first 20 dB to a 2-or 3-transistor opamp circuit.

How much quiescent current does the simulation give? Note that TIP31/32 are very old and slow. Ever looked at GBW(Ic)? More modern parts are likely to allow for less compensation at same output current requirements, hence more loop gain and less distortion.

Have you replaced your bootstrap current source with an ideal one and looked at the influence? The bootstrap does give you some extra voltage swing to drive the output transistors (and higher impedance), but you are giving up some output loading indepedence in return.

Last edited by sgrossklass; 1st September 2013 at 10:02 PM.
  Reply With Quote
Old 1st September 2013, 10:26 PM   #10
borys is offline borys  Poland
diyAudio Member
 
borys's Avatar
 
Join Date: Feb 2005
Location: Krakow
danny92
Hello, from my experience boostrapped VAS is increasing slightly THD values but You wouldnt notice it anyway by listening of the amp. In my opinion better is to use double boostrap and instead high power resistors just use diode (or zener diode) to separate whole input+VAS stage from output stage. Boostrapped voltage separated with diode do not need so low values and high power resistors in boostrap ciruit. How does it work - just take a look bellow, boostrap caps are 47uF and resistor 1k, output is swinging nearly rail to rail.
Attached Images
File Type: png boot21.png (136.6 KB, 75 views)
File Type: png boot23.png (127.1 KB, 9 views)
File Type: png boot25.png (111.9 KB, 8 views)
File Type: png boot24.png (125.4 KB, 8 views)
File Type: png boot26.png (117.0 KB, 8 views)
File Type: png boot28.png (139.9 KB, 9 views)
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
A superior VAS clamp and VAS current limit for the Blameless. GK Solid State 128 19th October 2011 03:03 PM
Cascode input and VAS amplifier design, will it work :) BachAudioDK Solid State 8 23rd January 2011 03:44 PM
@steven: VAS stage in your zero-feedback design? capslock Solid State 6 14th November 2009 02:18 AM
ported box design question: is exceeding the Vas of the driver bad? BrianGT Multi-Way 13 15th May 2006 05:35 AM


New To Site? Need Help?

All times are GMT. The time now is 12:51 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2