An ultimate amp protection circuit ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
add a serial impedance to the amp, reducing damping, where you want-it.

The serial impedance or resistance is much less than usually added resistor to get that cool slow tube bass, so not a big issue :D


If you chose a too big value; the charge will suffer from the charging current in them.

Yes, but I don't know the effect. Must be that thump at turn off.

They are expensive, reduce their value with time and use, and you have to change them after 10 years...

I have plenty of MKP to be utilized :cool:

They will not have an impact on the stability of an amp, here, the added serial resistance and inductance will reduce the effects of the capacitive loads.

Cool. Then it will "improve" stability. I will use cap output then for SSA-BJT-CFP (instead of using opamp servo). And use it also temporarily during tweaking component values and updating transistor models.

Thanks.
 
Jay;3210824Yes said:
Yes, thump, can destoy a Hp if the cap is too high value and rails high voltages too.
Please never // MKPs with electrolitics.You will not decrease impedance at HF, but create a resonant circuit between the inductance of the big cap and le low value of the little one. With, as a result a pic and a higher impedance at HF.
The good method is to // big caps of little value: Each have less inductance and you divide-it by they numbers. To, you are safer against failure and get less tolerance on the resulting value..
 
The most sophisticated protection circuit I've ever come across was that of the MC2 Audio MC series amps.
It is digitally controlled and never in the audio circuit but probably beyond most diy efforts. Essentially it fades the supply rail voltage up or down as necessary.

Just sayin'…
 
You can zoom on the web page to get dimensions ? They are standard, and will allow 2 contact relays a well.

You can zoom in, there are spacings (27mm) in one direction, but there are no spacings orthogonal to these. If you know of a standard perhaps you would be good enough to share it with me?

We need Alex here soon

What am I, chopped tomato? You don't like my layout? I can always keep it to myself...
 
You can zoom in, there are spacings (27mm) in one direction, but there are no spacings orthogonal to these. If you know of a standard perhaps you would be good enough to share it with me?
What am I, chopped tomato? You don't like my layout? I can always keep it to myself...
No way to measure-it on the screen ? I will try to find a better data-sheet, be patient.
I like your layout and I am very thankful to you. I was just waiting to final realization to congratulate-you. Now, i feel guilty...
Anyway, you know, both are better than one, dicrete components VS surface ones etc... And, as you see, i'm not good anymore in finding parts ;-)

I suppose this will be a longer work with modifications, other DUYers suggestions and returns etc...

The clipping circuit i provided, by example, i do not know if this one (From Elvee) would be better ? ...
http://www.diyaudio.com/forums/atta...10964-my-little-posh-tringlophone-clipper.png
 
This is your circuit.

I put in NE5532s, these are very cheap, will work OK in this application.

If you want thru-hole parts, you need to ask for them. 0.6W resistors, 40 thou hole spacing, e.g. or I will design the board for the minimum size/lowest cost, components one side.

Which transistor? Which opto's, which LED footprint positioned where, 5mm, 10mm?

I absolutely want the MOSFET relays included, part no. Farnell, Newark, Mouser, RS, Digikey / the MOSFET relays are a nonessential option.

I would like the board to fit in:- 4000x4000 thou, 2000x8000 thou if possible.

Plus we need critical input on the circuit functionality from a 3rd. party. I already showed some suggestions with regard to the connection of signal to ground.

Get the idea?
 
I put in NE5532s, these are very cheap, will work OK in this application.
Some parts need hight impedance: Better use Tl072 or TL074 instead.
If you want thru-hole parts, you need to ask for them. 0.6W resistors, 40 thou hole spacing, e.g. or I will design the board for the minimum size/lowest cost, components one side.

Which transistor? Which opto's, which LED footprint positioned where, 5mm, 10mm?

I absolutely want the MOSFET relays included, part no. Farnell, Newark, Mouser, RS, Digikey / the MOSFET relays are a nonessential option.

I would like the board to fit in:- 4000x4000 thou, 2000x8000 thou if possible.

Plus we need critical input on the circuit functionality from a 3rd. party. I already showed some suggestions with regard to the connection of signal to ground.

Get the idea?
It is your PCB design :)
Transistor can be the good old 2N2222. Leds not necessary on the board, they are useful to sit in the amp facade to monitor power and protection. We can change our idea and chose a multicolor instead ? Chose what you feel and the diameter you like...

What exact parts did-you miss, not included here:
http://www.diyaudio.com/forums/solid-state/221737-ultimate-amp-protection-circuit-2.html#post3206964
I will look at your ground suggestion.
Hp line input need to go both in the front opamp comparators and across the relays, of course. Not separated.
Input lines have to go in and out to the amp. It would be good to add a clipping circuit. Let me little time to make simulation between the two possible to see each one will be better for no distortion. Input switch is necessary to protect the output relay in case of short circuit. It will cut the input signal before the relay has to cut huge amps.

But, again, it is your version, do as you feel for you...
 
Last edited:
Power is coming into the circuit at the left-hand side, marked +12V, close to lable 'To Amp-In R'. Current flows through 2 solid state mosfet relays (2 forward-biassed diodes [LEDs?]) through 2 Opto-Isolators to a point marked 15/20mA, through a 330R resistor with no number, to the junction of D24 and Q2 where one of the relay coils is also connected, up through D24, back down through R35, R7, through the red LED to ground. Because D24 shorts the relay coil, the relay never closes, even when Q2 is turned on.

Another 25R resistor with no number connects the junction of R35 & R7 to the TOP of the other relay. Surely this resistor should connect to the TOP of R35?

You need to write a description of how all aspects of the circuit are intended to operate.

There is too much circuitry to expect anyone to look at it and be able to figure out whether it will work as intended. There are too many components without designators to allow meaningful discussion of the circuit.

Genius, they say, is 10% inspiration, 90% perspiration. You need to sweat this one a bit.
 
Last edited:
AC Power (left back) is provided to the trasfo 0.1 by two optional ways.that you can chose with the strap A1 or A2. The first one is direct, across a mechanical power button., with the strap A1. The second is a remoted from a preamp, by example. 12v comes from it by the pw-on and light the opto relayacross the 750ohm resistance. After the trasfo, a 7812/7912 regulated supply provide the +- 12v for the all circuitry.

Two other diodes, directly from the secondary of this trasfo gives a +17v signal to the 1k/1K bridge. Limited to ~2V by the LED. and flattened by the // 0.1 cap right in the+input of the K3 OP amp. As the -input of this OP amp is at 0.5V across the 10K and 400 bridge his output will turn to +, no current across the D diode.
This has no interest for the moment. It will take sense when you power off.
At this time, 12v still exists because the big caps of the power supply, so is the 0.5V at the -input. But no more 17v. quasi instant (the 0.1µ has few inertia), the +input will go to less than the 0;5V very fast, as the 0.1µ discharge to ground across the //1K.
IC3 out will turn to -12V, Forcing the V base of Q1 Q2 to less than V emitter across the D diode. They cannot conduct anymore, relays are cut, whatever their state.
This disconnect your speakers at power off or losse of power, before your amp power supply has time to discharge for no noise at power off.

to follow... please ask for comment if anything not clear.
 
Let's analyses K1.It is for soft start.
When power on we have 17V to the yellow indicating LED.then across the 2K to the optical relay in // with R 19. This provide enough current to light the diodes if the other side (out of K1 is at -V.. it is not yet, because +input of K1 is at ground, and 33µf is not charged.
(There is a mistake on the ground connection of the 110k, it has to be to the -V.)
-input is at -12v. optical relay is cut for the moment.
AC Power is feed to the main amp supply across R19.
This resistance, depending of your amp power, witch is not on the board, a big enough wired one with radiator, will limit the current flow in the AMP power supply.
The 33µf cap will slowly charge and, because the 100K/110K bridge, will go upper to 0V after some seconds (~3). At this time, the K1 output will turn to -V, the yellow led and the internal led of the rely will light, The optical relay will turn on, short circuit the R199: the amp PSU will receive full power.
 
Now, K2 (delay for speaker at power on)
When we start, K1 out is at +12V. The -input of K2 is at +12V across both 100K and 110k.
Out will be near -V rail, so across the diode D, bases of Q1, Q2 are at less potential than their emiter. Again, they will not conduct: the relay are cut.
When K1 out turn to -V after his delay (soft start is finished) , 47µf will slowly charge across the 110k. So the - input voltage will slowly decrease.When it will reach less than 0V, the K2 out will turn to +, liberating the bases of Q1, Q2: the relay can close if any V at his base, monitored by the green LED.
 
K4 (not referenced) is for triggering Q1 & Q2.
When no error signal +input is at +5V.-input at ground, base of Q1 Q2 are at +5V across 4.7K.
if the amplified error signal across the diode Do goes to an upper V level than 5V, out of the OP amp will turn to -V, and Q1, Q2, doe not conduct any more: relay is cut. The 0.22µf assure a delay to maintain protection when protection fire, Quickly charged across the low impedance of U3, it will discharge slowly across 750K. This assure that, if any very short error fire the system, it will reconnect immediately.
 
+12v is provided to the leds of input SSR,in serial, and Opto isolators of OI in serial too. then it goes to the collector of Q1, Q2.
When the collectors will go near 0V, current will light the diodes: all those relay will conduct.
When protection is on, those collectors are at +13V across the two 25 ohms resistances connected to +17V.
At equal potential on the two lines, no current, no light in the LEDs, no one of the opto isolators conducts.

Any more comment ?
 
Last edited:
The two SSR are here to protect the relay in case of short circuit of the speaker's line. When they conduct, they short circuit the input line before they return to the amp input. they are faster than the mechanical relay. With no more signal (shorted) in its input, the output of the main amp will be at 0V before the relay cut the lines, so those relay will not have to deal with the huge current across the short circuit.

The C8 &C9 caps are here to protect a direct input amp from DC in the input. As they remove dc from the signal entering in Us U1 to be compared, if no DC, no difference. if any DC, as the amp will amplify-it, and the same Dc will not be present at the+ input of the comparator, we will see an error signal. If your amp is not direct coupled at its inputs, better to replace C8 & C9 with a strap.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.