• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Group Buy - Very Simple Almost Universal Speaker Protection

Jump to Latest
#64
We recommend that you use a 10M resistor or smaller value for Rrs.
10M will give the longest reset time while still sufficient to compensate for any leakage current of OP1.

But you can also use anything between 1M and 10M.
You will still get reasonable reset time, sufficient for you to switch off the amplifier for fault investigation.


Patrick
 
Last edited:
#65
EUVL said:
A large DC will speed up the charging time for Ci through R1,2.
But not that of R5/C1.

If you want fast cutoff, the easiest way is to reduce C1.
You don't want to change Ci because you need it for AC filtering.


Patrick
Click to expand...
OK.
If a high DC voltage presents, Ci charging speeds up and OP1 turns on quicker. But it still needs 2.5s to cut off speaker because of R5/C1.
Is it right?
 
#66
Not exactly.

The time delay is shared between R1,2/Ci and R5/C1.
High DC will speed up the former, not the latter.
You can if course change all the default values to make one dominant over another.
Best is to simulate first in Spice.

The values we gave are the values we used ourselves, in multiple projects.
So they are good enough for us.


Patrick
 
#67
I meant :
You can of course change all the default values to make one dominant over another.

If you reduce C1 by 10x, and increase R1,2 by 2~3x, the delay by R5/C1 is reduced by 10x, and the triggering time will be almost entirely dependent on the input DC voltage.

The minimum trigger DC voltage is approximately :
1.2V + (0.05mA) X (R1 + R2 + R11 + R12)


Patrick
 
Last edited:
#69
How would you get high DC ?
If you lose a rail, the protection cuts in immediately.

There is still time needed to charge up Ci, be it much shorter for a high DC value (few tenths of a second).
If you want immediate triggering (ms) with a fixed threshold, easiest is to use MCU based solutions.
But I never want to use software solutions for safety.

Happy to give you full refund if this is not what you need.
Just send them back.



Patrick
 
Last edited:
#71
How could be a high voltage? Just imagine that one of output devices (power BJT/Mosfet) breaks and shorts to + or - PW.

I can tell you an another example. I have a composite LM3886 amp. It has 7815/7915 +/-15VDC for error-correction OP in front of LM3886.

It happened once. The major power supply +/-30VDC for LM3886 was ok. But the -15VDC regulator for the OP failed. +13VDC showed on speaker terminals in no time. Luckly it was still on the bench and no speaker was on it.

That's why the alarm sounded right away when I saw this 2.5s trigger time on the instruction.

I am not asking miliseconds. That's not my points here. I like you guys to improve the circuit.

Also I bought it and will keep it. But I am not going to use it until all my concerns are addressed.
 
#73
wwwtttwww said:
I am not asking miliseconds.
Click to expand...

Quote :
I like that it triggers instantly if DC at output reaches a certain value.

Please define "instantly" and "a certain value".

I have already explained in post #67 how to reduce the response time of R5/C1, and rely only on the input R1,2/Ci filter for the delay.
You can even leave out Ci altogether, but it is there for a purpose.
Even then, there will always be a delay, until Ci is charged up to the minimum trigger voltage as described above, however high the DC.
It is explained in detail in the link in post #70.
And Elliot's recommended RC time constant is in line with that suggested in post #67.

All speaker protection circuits I have see rely on an RC filter to separate AC signal from DC offset.
I don't see any need to change the circuit.


Patrick
 
Last edited:
#74
I like that it triggers instantly if DC at output reaches a certain value.

I don't think it is wrong.

For example I would like it shut off instantly when DC reaches 10V.

But it takes time to detect reaching 10V on Ci. I am not asking your circuit to detect 10V instantly.

What I don't like is that OP1 turns on, you put another delay afterwards.

The major purpose is to protect Speaker, not for latching the circuit for debugging.

This is purely for technical discussion, not for undermining your efforts.

Of course it is pure your right to or not to change the circuit.

Don't be mad.
 
#76
For the sake of technical discussion :

Let's say you have a BTL or balanced amplifier with+/-50V rails.
And you have every now and then a big drum at maximum output, at say 30Hz for 10 seconds.
An input low pass filter at 0.5Hz will reduce this input AC by 60x, or +/-1.7V.
The protection will trigger in and out, as the filtered signal rise and fall beyond 1.2V, even in the absence of any DC, unless you have C1.

So let's filter at 0.2Hz to be safe.
By doing that, you now increase the delay time of the input low pass by a factor of 2.5


Patrick
 
Last edited:
#77
To further add to the above, here are some combinations as simulated with LTSpice :

(R1+R2) / Ci / C1

6k / 220u / 10u
3V DC --> 2s
10V DC --> 1.6s

18k / 220u / 1u
3V DC --> 1.6s
10V DC --> 0.5s

18k / 220u / 100n
3V DC --> 1.4s
10V DC --> 0.4s

6k / 220u / 100n
3V DC --> 0.5s
10V DC --> 0.15s

So you can choose what suits your application best.

As said, all protection circuits I have seen use the same RC input filter, and hence follows the same trend.

BTW, 10DC into 8R is 12.5W.


Patrick
 
#78
OK. I am fine now. These are things you need to tell people and let them select their own risk.

I would pick up the last R, C combination. I don't mind a lower DC voltage for longer time. But for a high voltage DC on speakers, I would like that the circuit can cut off it off asap.

Another friendly suggestion.

If you would like to do a revision in the future, you could make the two speaker in/out connectors bigger (6.35mm or 7.62mm pitch).

Thanks for the discussion.
 
You must log in or register to reply here.
Top Bottom