Go Back   Home > Forums > >

Solid State Talk all about solid state amplification.

Understanding SOA
Understanding SOA
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
Old 18th November 2020, 04:23 AM   #1
profdc9 is offline profdc9
diyAudio Member
 
Join Date: Aug 2020
Default Understanding SOA

There is an aspect of rating transistors SOA I do not understand for class AB amplifiers.


On the SOA graph,



Ic = Output Voltage / Load Impedance
Vce = Rail Voltage - Output Voltage



(for Load Impedance approximately real and constant). Plotting Ic vs. Vce is called the "load line" I believe. When Vce is small, Ic is high (near the rails), and when Vce is large (near zero output) Ic is low.



For a transistor with a constant collector current, this load line is plotted on the SOA graph. It should remain under the SOA graph, properly derated for temperature at constant power dissipation.



For sinusoidal rather than constant Vce, high Vce only occurs near zero output. Therefore for a purely resistive load, the part of the SOA near second breakdown occurs only for the brief period near the zero crossing. So one would expect that the useful SOA would be somewhat better than the DC prediction.


However, with reactive loads, it is possible to have high current near the zero crossing because of the lead/lag of Ic relative to Vce. How does one account for this? Is using the DC prediction for SOA typically considered safe for this, or is there some other criterion?
  Reply With Quote
Old 18th November 2020, 05:00 AM   #2
BesPav is offline BesPav  Russian Federation
diyAudio Member
 
Join Date: May 2011
Location: Russia, Moscow
Quote:
Originally Posted by profdc9 View Post
Is using the DC prediction for SOA typically considered safe for this, or is there some other criterion?
Keep in mind SOA as a thermal dissipation ability and just simulate dissipation at the reactive load.
  Reply With Quote
Old 18th November 2020, 06:24 AM   #3
indianajo is offline indianajo  United States
diyAudio Member
 
indianajo's Avatar
 
Join Date: Jan 2010
Location: Jeffersonville, Indiana USA
Andrew T told us that sum of DC soa current rating of all parallel output transistors at the rail voltage they are subject to predicts capability to perform class ab amplification at a certain wattage (current) if the heat sink is near the 25 C tests are done at.
This was an experience based rule of thumb having nothing to do with the details of the soa test. Actual speaker currents are much higher, he said, than the resistive load amps are rated for wattage at.
Looking at class AB amps that are successful in the marketplace, ie don't blow up during the warrenty period, those follow this rule pretty closely. Amp repairmen have asserted on PA forum that 4 brands of PA amp were pretty reliable at rated wattage 24/7, at least the ones from the nineties: Crown, QSC, Peavey, Yamaha. If you look at their products from that era, they follow this rule for wattage rating.
__________________
Dynakit ST70, ST120, PAS2,Hammond H182(2 ea),H112,A100,10-82TC,Peavey PV-1.3K, MMA-81502, Herald RA88a mixer, Steinway console, Wurlitzer 4500, 4300

Last edited by indianajo; 18th November 2020 at 06:29 AM.
  Reply With Quote
Old 18th November 2020, 08:32 AM   #4
BesPav is offline BesPav  Russian Federation
diyAudio Member
 
Join Date: May 2011
Location: Russia, Moscow
Ok, let's draw some graphs.
Reactive load, +-72 V supply.


Now, compare current/voltage/time figures with datasheet SOA.
Attached Images
File Type: png OPS SOA Estimation.PNG (63.5 KB, 141 views)
  Reply With Quote
Old 21st November 2020, 03:03 PM   #5
profdc9 is offline profdc9
diyAudio Member
 
Join Date: Aug 2020
Based on your simulation, it seems like three pairs of transistors would be needed to operate safely at +/-72 V. At peak Vce (72V), the Ic is 6 A. Looking at the data sheet, the SOA current is 2.5 A. So 3 pairs would be barely adequate. Is this about right?
  Reply With Quote
Old 21st November 2020, 03:20 PM   #6
BesPav is offline BesPav  Russian Federation
diyAudio Member
 
Join Date: May 2011
Location: Russia, Moscow
Quote:
Originally Posted by profdc9 View Post
Is this about right?
Yes, highly likely, but doesn't forget safety margins for worst-case reactive load.
For hard driven amp i'ld put 4-5 pairs or design in G/H-class OPS with switched/tracking supply rails.
In case of non-diy professional use there are no way to override TD- or D-class in low-freq operation while for midband and high-freqs it can be built in classical AB-class.

Quote:
Originally Posted by profdc9 View Post
At peak Vce (72V), the Ic is 6 A.
~8A in the given not a really bad case.
So keep your fingers crossed.

Last edited by BesPav; 21st November 2020 at 03:22 PM.
  Reply With Quote
Old 21st November 2020, 06:58 PM   #7
john_ellis is offline john_ellis  United Kingdom
diyAudio Member
 
Join Date: Sep 2006
For an 8 ohm load? 72V gives 9A in the limit. Not that the amplifier would get to 72V peak output perhaps, but limiting conditions are precautionary.
For a load line I sometimes consider a 45 degree phase shift. That generates an ellipse instead of a straight line on the voltage:current (load line) graph. It's where the maximum of the current which arises which needs to be protected.
For reliability I would check that the ellipse is handled with the DC power level. Obviously, a reactive load line is frequency dependent but the SOA limits don't often allow for much safety above the DC level- and usually, the SOA graphs are specified for single cycles. Although the peak current at zero volts can be calculated (easily) without constructing a load line.

Last edited by john_ellis; 21st November 2020 at 07:02 PM.
  Reply With Quote
Old 21st November 2020, 08:24 PM   #8
john_ellis is offline john_ellis  United Kingdom
diyAudio Member
 
Join Date: Sep 2006
and a graph (as I only had to mod an existing one).
I'd say from the red curve you need 4 pairs (at least) of MJL3281A's compared with a single one (blue line). But you can consider what frequency the load line is at compared with SOA data. It's just that repetitive SOA data is not available, usually. I think the "DC" data is actually only for single 1s pulse.
Green curve is 45 degree load line.
Peak current is 9A/root 2 at the zero crossing (72V). Only half the cycle plotted.
Attached Images
File Type: png LoadLine 72.png (47.1 KB, 50 views)

Last edited by john_ellis; 21st November 2020 at 08:32 PM.
  Reply to this post

Reply


Understanding SOAHide this!Advertise here!
Thread Tools

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
Please teach me - triple slope SOA protection R.G. Solid State 41 28th February 2013 04:05 PM
Understanding output stage SOA?? Zero Cool Solid State 8 22nd November 2007 10:57 AM
SOA and sagging PS JohnSz Solid State 12 3rd March 2004 05:37 AM
SOA - What is true and what is wrong ? djdan Solid State 1 13th December 2001 01:42 PM


New To Site? Need Help?

All times are GMT. The time now is 07:16 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2020 DragonByte Technologies Ltd.
Copyright ©1999-2020 diyAudio
Wiki