Question for all mighty ILIMZN and others... - Page 2 - diyAudio
 Question for all mighty ILIMZN and others...
 User Name Stay logged in? Password
 Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read Search

 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
diyAudio Member

Join Date: Nov 2005
Location: serbia, zajecar
ok, this is the update...

I limited the current to ~47mA, and connected the diodes between the output and vas stage(I hope I understood it well)?

1. THe firs thing I didnt understand, is where to put resistors for stable shunt impedance? collectors of Q9 and Q10 to the ground? What should be the value of these resistors?
Attached Images
 update01.jpg (66.0 KB, 374 views)

 7th February 2006, 06:41 PM #12 diyAudio Member     Join Date: Dec 2005 Location: the north The resistors to reduce Open Loop gain is from Q9 collector to ground and other from Q10 to ground. Value can be something like 10k - 22k. If we look at Q7, the emitter resistor is 100 Ohm. The gain in Q7 then is reduced to 10k / 100 = 100 or 22k / 100 = 220. As the gain is mostly determined by 10k ( or 22k ) resistors, this gives better linearity. ( Resistor is very linear load compared to transistor, MOSFET Gates ) =================================== The gain in input stage transistors is: 680 / 47 Ohm ~ 15 The total open loop gain will be = gain in input stage x gain in VAS stage. The feedback will be = Open Loop Gain / closed loop gain. You have closed loop gain (22k/1k) + 1= 23 Say open loop gain is 2300, this gives a feeback factor of: 2300 / 23 = 100 So the amount of feedback will be 100 = 40dB 40 dB is a normal feedback. =================================== You can put in compensation caps, as told between Q7 base and Q9 collector ( and Q8 base to Q10 collector ) Maybe a good value to start with could be 47 pF caps. =================================== As mentioned, you will probably not have much problem with DC-offset. Especially if your input stage have transistor pairs that are closely matched. You could probably removed C9 and C14. I do not think you will get much DC-offset, if you do. __________________ lineup
diyAudio Member

Join Date: Feb 2005
Location: Zagreb
Quote:
 Originally posted by bogdan_borko ok, this is the update... I limited the current to ~47mA, and connected the diodes between the output and vas stage(I hope I understood it well)?
Yes... but you need two large capacitors on the other sides of the diodes (the driver side). These will hold the local rail voltage high when the actual rail droops under load.
Current limit of the driver stage was ok even at 63mA, you are still (although barely) under the BC550/560 SOA, but this sort of current will only be demanded from the amp under extraordinary conditions, not in normal operation.

You still miss MOSFET gate protection diodes. As it is now, even a momentary short on the output, normally easily survived by the MOSFETs if fuses are included in the drains, will kill them instantly, even with the fuses. This is because Vgs will momentairly go to almost the full rail voltage, which is higher than the maximum +-20V. Current limit is secondary here - any break of NFB operation for any reason (oscillation, very reactive load, short), will result in the driver stage attempting to drive the MOSFET gates as far as they will go - which is essentially rail voltage minus 4.7V. Omitting these diodes is a classical beginner's mistake which tends to fry the MOSFETs for 'unknown reasons' (trust me, this is from bitter experience ). Do not confuse VMOS with LMOS, most if not all LMOS have built in gate protection, VMOS do NOT!

Quote:
 1. THe firs thing I didnt understand, is where to put resistors for stable shunt impedance? collectors of Q9 and Q10 to the ground? What should be the value of these resistors?
Lineup has already provided a good explanation...

BTW I would not recomend a 4 ohm load unless your power supply was deliberately made 'soft' but that has other unwanted repercussions. You may get away with it if you use two isolated heatsinks (so no mica is needed between MOSFET and heatsink), and also remove the anodization where the transistor is in contact with the heatsink - which needs to be VERY big. Alternatively, you could solder the transistors onto a copper heat spreader, then mount that isolated onto a heatsink. In both cases you need premium thermal greese. Neither of these options is trivial, the idea is to have minimum thermal resistance. One pair will, however, work fine with an 8 ohm nominal (4 ohm minimum) load.
My sincerest recomendation would be to use a lower voltage power supply for the output MOSFETs. Your driver stage plus MOSFET Vgs at maximum current already loses about 8-10V off each rail, on the MOSFETs, and this is simplu just heat - not something you want here. Using 35V rails for the output would probably make very little difference in performance, but would quite signifficantly reduce dissipation on the MOSFETs, i.e. alow you the use of lower load impedance with a given heatsink.
There is always the possibility of paralleling output devices, but I have found that the elimination of the source resistors that is possible when a single pair is used, has some sonic advantages - even if it means a lower output power.

Oh, and BTW, I am HARDLY 'almighty'

diyAudio Member

Join Date: Aug 2005
Location: K-town
Quote:
 Originally posted by bogdan_borko is there eny other bc transistor with lower noise then bc550/560?
Although it's not BC, how about the KST5089 from Fairchild? This one is only 25V instead of 30V, but also has a fairly linear Hfe of around 600!

http://www.datasheetarchive.com/semi...asheet=1078145

I suppose it could be complemented with KST5087.
__________________
All the trouble I've ever been in started out as fun......

 8th February 2006, 03:33 AM #15 diyAudio Member     Join Date: Oct 2003 Location: Near the sea These seem like SMD versions of the classic 2N5087 and 2N5089 that come in TO-92 package.
diyAudio Member

Join Date: Oct 2004
Location: At the sea front, Rotterdam or Curaçao
Quote:
 Originally posted by ilimzn I am HARDLY 'almighty'
No, but you are good.
Attached Images
 lange 20 watts.jpg (39.1 KB, 312 views)
__________________
A veces la Moral y la Ley no están de Acuerdo. Si No, se puede esperar a mi llegada, el placer es todo mío.

diyAudio Member

Join Date: Dec 2005
Location: the north
Quote:
 Originally posted by jacco vermeulen No, but you are good.

In jacco v attached symmetrical input MOSFET output amplifier
68pF compensation caps is put in input stage.
This might be an alternative to try.

This amplifier is very similiar to bogdan_borko amp.
Uses fully symmertical BC550/560 input biased with CCS, Constant Current Sources.
Cascoded VAS and MOSFET push-pull output.

To use separate supplies, for input/VAS and output,
in this case +24 Volt and +20 Volt, is a very good option.
__________________
lineup

 8th February 2006, 08:25 AM #18 diyAudio Member     Join Date: Apr 2004 Location: Gütersloh Hi bogdan, i have also some comments about your amp... Maybe you should reduce powersupply a bit (36v ?), bc550/560 has max rating of 45v, including swings from inputsignal and variations from powersupply you risk blowing the inputstage. You should also observe dissipation in Q3/5, 7ma*45v=315mw, they might be too hot to touch. An inputfilter is always helpful, maybe a 100pf in paralell to R2 ? You should use basestoppers for Q9/10, it's very likely they enter parasitic oscillation without them. (~47-150ohms). 15ma in vas might not be enough to drive the mosfets for higher frequencies ? R7/10/12/14 will need to be 2watt versions, you have 225mw dissipation here. The same for r25. Mike
diyAudio Member

Join Date: Feb 2005
Location: Zagreb
Quote:
 Originally posted by MikeB Hi bogdan, i have also some comments about your amp... Maybe you should reduce powersupply a bit (36v ?), bc550/560 has max rating of 45v, including swings from inputsignal and variations from powersupply you risk blowing the inputstage.
Well spotted, MikeB, see, you are far more mighty than I am
At one time I did have a similar problem, I just solved it by including zener clamps to the LTP common emitter points, rarther than going to cascode, or using different LP transistors. However, I also had a regulated power supply for the front end... more below.

Quote:
 You should also observe dissipation in Q3/5, 7ma*45v=315mw, they might be too hot to touch.
This is one reason why I mentioned a higher reference voltage for the CCS, which means higher Re's for the CCS transistors, and also lower Vce for them, therefore lower dissipation. Also, the CCS thermal dependence may be lower. Since the reference is bypassed heavily, normal zener diodes would work here, 24V say.
Alternatively, you can use a series resistor between the CCS and the LTP tail, to drop voltage and take over some power dissipation. My favorite way of doing this is actually providing two separate resistors twice the value, one to each LTP emitter. Then, the emitters are connected with a single resistor. This alowes the use of lower power resistors as there are two to dissipate the heat, and also adjusts LTP degeneration using a single resistor - while keeping the number of components the same. The circuit is actually a star-to-triangle transformation from the classical 'separate emitter resistors + voltage drop resistor' topology. For the voltage drop resistor >> emitter resistors, this is easy to approximate by using the E-to-E resistor as 2x previous Re, and splitting the voltage drop resistor into equal parallel resistors twice the value.

The 24V CCS reference version has an advantage if you ever want to cascode the LTPs (to be able to use low Vce low noise parts safely), since you can use the opposite references as the cascode transistor base voltage.

Quote:
 15ma in vas might not be enough to drive the mosfets for higher frequencies ?
It should be just fine for a single pair, remember that you are essentially only driving a bit more than the Cgd of both MOSFETs (Cgs are bootstrapped by MOSFET follower action).

 8th February 2006, 04:28 PM #20 diyAudio Member     Join Date: Apr 2004 Location: Gütersloh Yes, the series resistor in the ccs to ltp is the easiest way out... In this case i would still prefer the 2bjt-ccs, it's thermal completely stable. (the sensing bjt sees practically no dissipation (Vce=1.2v), and will reduce current when heating up). About the ltp-devices, 2sc2240/sa970 might be a way out ? Mike

 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 OffTrackbacks are Off Pingbacks are Off Refbacks are Off Forum Rules

 Similar Threads Thread Thread Starter Forum Replies Last Post wrenchone Tubes / Valves 13 20th October 2008 04:57 AM imo Solid State 0 18th January 2006 10:10 PM zox2003 Solid State 1 9th April 2005 10:38 PM breguetphile Chip Amps 2 30th May 2003 01:29 PM wallijonn Multi-Way 10 5th November 2002 06:03 AM

 New To Site? Need Help?

All times are GMT. The time now is 01:14 PM.