Amplifier DC protection - Page 2 - 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 3rd April 2012, 05:35 PM   #11
diyAudio Member
 
Nico Ras's Avatar
 
Join Date: Nov 2005
Location: East Coast of South Africa
Fdw, the circuit you propose works well. It is in some commercial applications. Some adds a small SCR to latch the relay so that only power down will reset it.
__________________
Kindest regards
Nico
  Reply With Quote
Old 4th April 2012, 09:51 AM   #12
FdW is offline FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Default First 'final' design

This is what I have now, it is reasonably fast it can be made to work with very specific operating parameters (see table). The first set of parameters will allow an amplifier running at 5 Hrz and full power while still detecting 190 mV output offset. The second set of measurements will allow the amplifier to run at 10 Hrz and detect (even) a 110 mV offset.

Reaction times vary, from more than 1 Sec for small offsets and down to 100 mS for large offset. Real amplifiers would switch even faster (when failing totally) due to their higher rail voltage.

The output protection relay stay's on for the duration of the DC offset event, and after that it will stay on for an extra 5 Seconds.

In the attached picture you see,
Green = Offset event
Blue = Protection relay
Gray = Servo output

I really would like to see some comments on the schema and timing results

; While these measures where take the aplifier was running at 28 Vpp

;.param Rref 76
;.param Fv4 5Hrz; Amplifier opperating frequency
;.param Vv7 19mV; Reaction time 8381 mS @ offset 190mV; Min detect
;.param Vv7 25mV; Reaction time 4771 mS @ offset 250mV
;.param Vv7 50mV; Reaction time 1962 mS @ offset 500mV
;.param Vv7 100mV; Reaction time 959 mS @ offset 1V
;.param Vv7 250mV; Reaction time 540 mS @ offset 2.5V
;.param Vv7 500mV; Reaction time 326 mS @ offset 5V
;.param Vv7 1V; Reaction time 261 mS @ offset 10V
;.param Vv7 2V; Reaction time 171 mS @ offset Vrail; Total failure

.param Rref 38
.param Fv4 10Hrz; Amplifier opperating frequency
.param Vv7 11mV; Reaction time 6685 mS @ offset 110mV; Min detect
;.param Vv7 25mV; Reaction time 2056 mS @ offset 250mV
;.param Vv7 50mV; Reaction time 990 mS @ offset 500mV
;.param Vv7 100mV; Reaction time 492 mS @ offset 1V
;.param Vv7 250mV; Reaction time 283 mS @ offset 2.5V
;.param Vv7 500mV; Reaction time 176 mS @ offset 5V
;.param Vv7 1V; Reaction time 109 mS @ offset 10V
;.param Vv7 2V; Reaction time 99 mS @ offset Vrail; Total failure


The shown solution uses a DC servo that is an existing part of the amplifier.
Attached Images
File Type: png DC protect (ATL style).PNG (101.6 KB, 230 views)

Last edited by FdW; 4th April 2012 at 09:55 AM.
  Reply With Quote
Old 4th April 2012, 10:33 AM   #13
AndrewT is online now AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
The reaction times of 99ms and 171ms for rail voltage offset seem a bit long.

Is it the 220k for R6 that gives this long time delay?

Could you explain the operation of the network between u7 & u6?
__________________
regards Andrew T.
  Reply With Quote
Old 4th April 2012, 11:00 AM   #14
FdW is offline FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Quote:
Originally Posted by AndrewT View Post
The reaction times of 99ms and 171ms for rail voltage offset seem a bit long.
But much shorter and it would not allow the amplifier to run at 10 Hrz. You can very easily experiment with these parameters (if using a simulator program). As stated, in practical situation these values will be lower, the sample amplifier only has a rail of 15 V and the reaction time gets shorter with rail voltage (but you will need to use diodes with an higher Vak and the lower detection limit will be compromised).

Quote:
Originally Posted by AndrewT View Post
Is it the 220k for R6 that gives this long time delay?
Yes, this one controls the integration time of the DC servo, it could be made a bit faster but then it starts to influence the amplifier at low frequencies. The other parameter that controls reaction time is the value of Rref.

Quote:
Originally Posted by AndrewT View Post
Could you explain the operation of the network between u7 & u6?
This is an low pass filter with voltage limiter and is used to protect the DC servo from the amplifiers output voltage (normally 100 Vpp or more). The diode Vak sets the detection response, it limits the amount of AC (also DC when above 250 mV) and (at low values) it makes that the DC value is large compared to the AC value. To explain: when the amplifier runs at 100 Vpp and the offset is 100 mV the offset voltage is 0.1% of the servo input, if the signal is clipped by the diodes then the AC part will be 250 mV and the offset will be 40% of this.

P.s. U7 acts as the power amplifier and U6 is its DC servo circuit, I do not consider them part of the DC protection circuit!

Last edited by FdW; 4th April 2012 at 11:03 AM.
  Reply With Quote
Old 4th April 2012, 01:20 PM   #15
AndrewT is online now AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Now I understand.
Power Amp = u7
Why inverted?
__________________
regards Andrew T.
  Reply With Quote
Old 4th April 2012, 01:45 PM   #16
FdW is offline FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Quote:
Originally Posted by AndrewT View Post
Now I understand.
Power Amp = u7
Why inverted?
No reason, just like it happened
  Reply With Quote
Old 5th April 2012, 09:12 AM   #17
FdW is offline FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Default DC protection (after tuning/tweaking for speed)

First of all, I really would like to get some comments, about timing and threshold’s as in this circuit.

Attached is a measurement that shows all involved signals.

First let’s identify them all:
• Azure: Input signal 1400mV sinus (after 10 seconds) and superimposed 100mV of DC (resulting in one volt output DC) (after 20 seconds).
• Green: The resulting DC-servo output signal.
• Blue: The input DC signal (voltage source V7).
• Purple: The upper detection threshold (set by Rref).
• Gray: The lower detection threshold (set by Rref).
• Red: The protection relay (protect1/protect2).

As you can see, the minimum signal to be detected is set by the voltage that you see at 10 seconds, if the detection limit is set lower then ‘false’ detections will happen. It is possible to lower this voltage by using a larger capacitor (C4). Using a larger capacitor will lower the low frequency gain and thus the signal that is generated by the DC servo (at the time of the ‘step’-voltage, when the AC signal switches on).

Detection of DC takes place when the DC-servo output signal crosses the detection threshold (Rref). The detection threshold cannot be set lower due to the fact that then ‘false’ detects are generated. It is possible to do the detection a bit earlier by lowering the ‘step’-voltage when the AC signal is switched on (at the 10 seconds point) by selecting a larger capacitor (C4).

Whatever you do, you cannot (ever ) detect any DC in a time shorter than ‘(1/2) * (1/Hrz)’, this is due to the ‘step’-response of the DC-servo (and possibly the 'Nyquist'-limit).

In addition to all this, you may think that using a sufficiently large capacitor will lower the ‘step’-response of the DC-servo (for practical frequencies) far enough to get nearer to the limit (as described above), this is not so in the sample given Vref is at 75mV and larger capacitors will force you to lower the reference voltage, at some point Vref will be lost into the circuit noise. When you get near the noise limit of the circuit, the circuit will start to trigger on spurious signals (that are always there, maybe it starts to detect alpha decay’s ).

P.s. The Azure (tst) signal looks a bit ragged-edged but that is due to the fact that I connected it to the Opamp's input instead of the 'real' input, but it still serves it's purpose
Attached Images
File Type: png DC protect (ATL style).PNG (122.0 KB, 163 views)

Last edited by FdW; 5th April 2012 at 09:35 AM.
  Reply With Quote
Old 5th April 2012, 10:24 AM   #18
FdW is offline FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Next graph is made using an 20u integration capacitor (C4). It shows the limits of detectability. In theory, the situation shown, could detect in about ‘(1/2) * (1/Hrz)’ seconds, in this case 100mS but you need to create a low-low-low noise and ultra-clean environment to do that. The difference voltage between 'off' and 'on' is about 100uV.

It also shows that if you are not able to detect the peek at ‘(1/2) * (1/Hrz)’ seconds, then the next possible peak to detect is at ‘(1 + (1/2)) * (1/Hrz)’ seconds (it will be detectable just before it is at its peak).

The first diagram shows the DC-servo output signal.

The second diagram shows the critical parts, of the first, that determine detectability.
Attached Images
File Type: png DC protect (ATL style).PNG (117.9 KB, 151 views)
File Type: png DC protect (ATL style) at the limit.PNG (99.6 KB, 137 views)

Last edited by FdW; 5th April 2012 at 10:29 AM.
  Reply With Quote
Old 5th April 2012, 10:31 AM   #19
diyAudio Member
 
jan.didden's Avatar
 
Join Date: May 2002
Location: Great City of Turnhout, Belgium
Blog Entries: 7
Have you thought about putting some hysteresis around your comparators to help it to make up its mind and switch cleanly?

jan
__________________
If you don't change your beliefs, your life will be like this forever. Is that good news? - W S Maugham
Check out Linear Audio!
  Reply With Quote
Old 5th April 2012, 10:45 AM   #20
FdW is offline FdW  Netherlands
diyAudio Member
 
FdW's Avatar
 
Join Date: Feb 2009
Location: Both feet on the ground (near Gouda)
Quote:
Originally Posted by janneman View Post
Have you thought about putting some hysteresis around your comparators to help it to make up its mind and switch cleanly?

jan
Jan, it has R7/R8 and that works nicely (simulated).
  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
Making car amplifier SMPS with tl494 + DC Protection Risto80 Car Audio 750 18th May 2014 07:59 AM


New To Site? Need Help?

All times are GMT. The time now is 08:59 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