Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Solid State Talk all about solid state amplification.

Driving 8 pairs of MJL21194/93
Driving 8 pairs of MJL21194/93
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 30th December 2018, 05:28 PM   #11
rhythmsandy is offline rhythmsandy
diyAudio Member
 
Join Date: Aug 2012
hi bob what if we a class A driver stage like in the pic then there is no chance that the current which is set would go out of class A. Please let me know about one more thing that I have seen some triples with drivers emitter resistors are split and connected to the output . Is there any advantage of it? As far as the stability is concern im playing with compensation caps across the base and collectors of the drivers.
Attached Images
File Type: jpg driver bias.JPG (60.1 KB, 627 views)
  Reply With Quote
Old 30th December 2018, 05:32 PM   #12
rhythmsandy is offline rhythmsandy
diyAudio Member
 
Join Date: Aug 2012
I mean to say about this circuit if you look at emitters of C4793 and C5200 they have resitors of values 120 and 220 1W resistors respectively connected to the output. Now how does that benefit of using such config than the Locanthi triple like where the resistors doesnt connect to the output rather connect between the two complimentary driver transistors by just using one resistor.
  Reply With Quote
Old 30th December 2018, 05:35 PM   #13
rhythmsandy is offline rhythmsandy
diyAudio Member
 
Join Date: Aug 2012
I was checking the other day about Meyer sound amplifiers Class AB/H and found that they are using some IRF140 or some Hexfet mosfet. I was quite interested to know that what if we use 15 pairs IRFp240/9240 instead of Bipolars like 6 pairs of MJL21194/93. One thing which attracts me is the no secondary breakdown in the mosfets i agree the fact about non linearity and other higher bias issues but in pro application like driving dual 18 inch subwoofers wouldnt mosfets are good enough. Now here is the question that what are points to consider if we use 15 pairs of IRFP240/ 9240 like a good TO264 package transistor would drive it? Consdier Im biasing each mosfet at about 20ma as its used more for the subwoofer application.
  Reply With Quote
Old 30th December 2018, 10:37 PM   #14
Bob Cordell is offline Bob Cordell  United States
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by rhythmsandy View Post
hi bob what if we a class A driver stage like in the pic then there is no chance that the current which is set would go out of class A. Please let me know about one more thing that I have seen some triples with drivers emitter resistors are split and connected to the output . Is there any advantage of it? As far as the stability is concern im playing with compensation caps across the base and collectors of the drivers.
The drivers in a Locanthi triple stay in class A unless their bias current is insufficient to turn off the output transistor when there is a very high rate of change of output current (i.e., when there is a lot of need to sweep minority carriers out of the base). This will be the case for most class A driver circuits. Connecting the driver emitter resistors to the output rail is usually not as good, since it allows the drivers to operate in class B.

Cheers,
Bob
  Reply With Quote
Old 2nd January 2019, 11:22 AM   #15
rhythmsandy is offline rhythmsandy
diyAudio Member
 
Join Date: Aug 2012
Hi Bob thank you very much for the reply how about Mosfets? Can we use mosfets in pro application? if so how many IRFp240 / 9240 pairs can be driven with To-264 transistor like 5200 or 1943?
  Reply With Quote
Old 2nd January 2019, 11:25 PM   #16
Bob Cordell is offline Bob Cordell  United States
diyAudio Member
 
Join Date: Sep 2006
Happy New Year! May all of your transistors have beta of 100 and all of your loudspeakers present an 8-ohm resistive load. May you also win the $500,000,000 lottery. Wait, I take that back. I'll take the $500M. Forget about the transistors and the loudspeakers.

Seriously, as responsible designers we are taught to take the worst case seriously, although I will admit that at times I have been over-conservative in making a point. If we want the highest performance in our amplifiers, everything matters.

But the numbers I quoted are not that far-fetched for the worst case. Minimum beta for the MJE15033 driver is only 50, and that minimum is only good up to 1 Amp. The very popular MJE21193/4 has a minimum beta spec of only 25 at 8 Amps and 5V Vce. Beta droop is very real, and Doug Self will also tell you that. The MJE21193/4 get onto the slippery slope of beta droop starting at only about 2A.

A less-experienced designer might assert that one will never require an output pair to deliver 8A in a reasonable design. Do the math. In my book I recommend that one output pair can be used for an amplifier up to about 75W into 8 ohms. At 75W, the peak output voltage will be 34.6V and peak output current will be 4.3A into 8 ohms. That amplifier will usually be expected to perform well into a 4 ohm load as well, at least in a consumer application (we assume good heat sinking and allow that it may not be asked to put sustained high power into a 4 ohm load). With a 4 ohm load, the amplifier will be called upon to deliver 8.6-Amp peaks into 4 ohms. So those MJE21193/4 output transistors may indeed have their beta fall to 25 according to the spec sheet. But not very often.

But let's also look at the case where we are using sustained-beta output transistors that can deliver minimum beta of 45 at reasonably high current, like the MJL3281/1302. If we assume beta of 50 for both the driver and output devices in a 2 EF output stage, we have current gain of 2500. Sounds like a lot? No. With a 4-ohm load, the output stage loads the VAS with a very nonlinear 10k ohms. Do you want your VAS to be really loaded like that?? After all that work you put into the rest of the amplifier? Just to save a dollar or two by not using pre-drivers and implementing a Triple?

Take a look at Chapter 3, Section 3.5 in my book. There I take a decent, plain-vanilla Miller-compensated amplifier design and change the output stage from a Double to a Triple and compare the difference in THD-1 and THD-20 at various load resistances. You need go no further than compare Chart 3.4 to Chart 3.5 to see the obvious benefit of using an output triple. These comparisons are based on simulations that use accurate typical models of the MJL21193/4. Even with just an 8-ohm load, THD-1 is cut by a factor of 8, and THD-20 is cut by a factor of 2.4.

The great Bart Locanthi taught us the value of using triple output stages in the late '60s when I was still in High School, and a great many top designers since followed his advice, including Barney Oliver of HP. Yes, transistors have improved since those days of the 2N3055 and the MJ802, but Bart's advice still rings true.

Cheers,
Bob
  Reply With Quote
Old 4th January 2019, 03:08 AM   #17
rhythmsandy is offline rhythmsandy
diyAudio Member
 
Join Date: Aug 2012
1. Bob I have question with error correction using Mosfets in OPS. What I have heard from few friends is that error correction doesnt sound as good as a simple biased OPS. Can you please tell me that is it really so.

2. How many IRFP240/9240 pairs can be driven using a single 2sc5200 or NJL transistor at 80V?
  Reply With Quote
Old 4th January 2019, 06:19 PM   #18
rhythmsandy is offline rhythmsandy
diyAudio Member
 
Join Date: Aug 2012
I have been going through this link about arcing in the amplifiers in pro application where the dew point giving problems in arcking. Please go through this interesting link about Crest 9001 amplifiers https://splnetwork.com/not-everythin...able-08202014/

How to avoid such dew point scenarios and as well as where do we find these laminated bus bars. I have seen it on few pics but never searched for it.
  Reply With Quote
Old 4th January 2019, 07:02 PM   #19
Bob Cordell is offline Bob Cordell  United States
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by rhythmsandy View Post
1. Bob I have question with error correction using Mosfets in OPS. What I have heard from few friends is that error correction doesnt sound as good as a simple biased OPS. Can you please tell me that is it really so.

2. How many IRFP240/9240 pairs can be driven using a single 2sc5200 or NJL transistor at 80V?
I have not heard anything bad about error correction, but there are not that many commercial amplifiers out there that use it (at least HEC). The Halcros used a version of my HEC implementation and they were well-regarded. What amplifiers were your friends referring to?

It is difficult to answer your second question without some significant analysis and choice of a complementary pair of drivers.

Cheers,
Bob
  Reply With Quote
Old 4th January 2019, 07:27 PM   #20
Bob Cordell is offline Bob Cordell  United States
diyAudio Member
 
Join Date: Sep 2006
Quote:
Originally Posted by rhythmsandy View Post
I have been going through this link about arcing in the amplifiers in pro application where the dew point giving problems in arcking. Please go through this interesting link about Crest 9001 amplifiers https://splnetwork.com/not-everythin...able-08202014/

How to avoid such dew point scenarios and as well as where do we find these laminated bus bars. I have seen it on few pics but never searched for it.
Entertaining article and a nice walk down memory lane. I consulted on the design of the Crest 8001. I recall having to hot-wire a 30-Amp line from my dryer outlet to test it in my basement.

There is probably a lot more detail to why they had "lightning" as a result of some condensation. Even with 150-V rails it seems a little unlikely assuming they had adequate trace spacings and good soldermask. This assumes the problem was on the PC board. But mention of bus bars suggests otherwise. I have not seen laminated bus bars. Maybe they had +/- HV rail bus bars in very close proximity to each other and some condensate got in between. It is easy to imagine that they would have used large bus bars, given the currents involved in delivering that amount of power. Also, once enough condensate caused an arc with ionization, extremely high currents can flow and a lot of light can be created :-). Anyway, it sounds like the guy who did the physical design pushed things a little to close. I'm sure it was a sight to behold!

Cheers,
Bob
  Reply With Quote

Reply


Driving 8 pairs of MJL21194/93Hide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
WTB: MJL21193 and MJL21194 - TO-3 package Thomas Giz Swap Meet 1 27th September 2018 05:12 AM
Mjl21194 & Mj15003 gregjpeters Solid State 6 13th December 2008 05:42 PM
FS: MJL21193 and MJL21194 - new zener_diode Swap Meet 9 27th December 2007 04:36 AM
simple amp, PNP driving PNP pairs wleediy Solid State 0 8th July 2007 11:23 PM
MJL21193 and MJL21194 HuntTheShunt Parts 1 5th December 2004 01:11 PM


New To Site? Need Help?

All times are GMT. The time now is 05:59 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki