I will be using what I KNOW works.
Page 3 - DC protect ... and page 4 - opto overcurrent = (below PDF)
Just a single channel version of each , fed into a diode OR gate to
trigger the SS relays on page 3 of the PDF.
Turn on delay can be either a trigger to the unused dc detect circuit input
or direct to the OR gate (through a 2 transistor multivibrator).
Just 6-8 to-92's , 2 FET's , 2 opto's. TO-92 count will depend on how
I implement the BJT/diode OR gate logic.
OS
I will propose a simplified version of mine with something like:
1 Quad OPA (TL074 ?)
12 resistances
5 caps
2 or 3 adjustables
6 little diodes+1 2N2222 and a relay.
Last edited:
I like your protection. If you can simplified it using TL074 and optional ordinary relay (because mosfet with low RDS-ON is hard to find here), I will make the PCB layout.
And please explain how to adjust it, if there are any adjustment need.
I propose to people interested to move to the dedicated thread,
http://www.diyaudio.com/forums/solid-state/221737-ultimate-amp-protection-circuit.html#post4330719
and stop polluating the OS one.
My purpose, here, was just to offer to OS a nice and original alternative for protecting his nice and original Amplifiers. It is obvious he is not interested.
Last thing, for OS, look at:
http://www.diyaudio.com/forums/solid-state/221737-ultimate-amp-protection-circuit.html#post4467493
if you want to see how simple it can be made.
http://www.diyaudio.com/forums/solid-state/221737-ultimate-amp-protection-circuit.html#post4467493
if you want to see how simple it can be made.
Christophe, the whole thing fits 101.6 x 76.2 mm board, including speaker relays (electromagnetic or SS), providing:
1) Remote (trigger signal) on/off;
2) Soft-start with flexible control over the delays and sensor checks;
3) Protection against:
- DC offset (2 channels);
- Over-current (2 channels);
- Over-temperature (2 zones);
- AC failure.
4) Status indication (single LED);
5) Redundant algorithm - extremely robust in every stage of the working cycle.
Hard to beat performance/simplicity ratio
As OS just mentioned, sensors can be used separately from the controller, although I don't see any reason not to use the controller, unless it's a matter of controller-fobia
Attached layout, designed by jwilhelm, is not the latest one, but already compact enough...
Again - I like your circuit, Op-Amp based sensors are precise, SS relays are fast and powerful, but... sometimes you need an extra control channel, sometimes you need some delay to last a bit longer, etc. That's why I use a micro-controller.
Cheers,
Valery
P.S. Could be a cool project to combine your sensors and my controller / firmware, to have a great combination of precision and flexibility, plus cool indication features
Attachments
Val,As OS just mentioned, sensors can be used separately from the controller, although I don't see any reason not to use the controller, unless it's a matter of controller-fobia
No "controller-phobia" I like the small compact arduino solution.
I just proposed to design a fully BJT/diode solution.
It would be very similar to the uPC1237 (with its 14 internal bjt's).
I'm not going with the flip-flop and latching like the 1237 , so about
8 BJT's are all that is needed.
I know it would be quite simple to integrate the micro and everything
on a amp PCB. But , would that not be overkill for a single channel ?
The BJT's can be any mpsa92/42 -5401/5501 , available in most junkboxes.
OP-amps are nice , but not as "bombproof" as 200V BJT's.
I think it would be fun to go back to the basic transistor multivibrator ,
flip-flop , and diode logic. Would be great learning , too !
The result would be a simple "bombproof" discrete uPC1237 , customized
beyond what the IC is. If it grows to >10 bjts , so what.
For IC lovers .... there would still be the opto's
. For the SMD lovers ,the MOSFETS would remain. A great intro into a few smd solderings.
OS
Slone had a design in one of his books that had a neat led status indicator. It would flash slowly at on delay, light steady in normal operation, and flash quickly on DC detection. It wasn't complex either. I'll see if I can find it this evening. It sounds close to what you are describing.
Valery, with pleasure.P.S. Could be a cool project to combine your sensors and my controller / firmware, to have a great combination of precision and flexibility, plus cool indication features
If you have an idea of the points where a controller could add features, reduction in size and, yes flexibility, go ahead, please.
My idea was not about all those gadgets (soft start etc), but the sensor itself. Whatever we try, any censor based on integration of the output signal will take more than 1/5 S to operate. Not so sensible, not able to detect oscillations from the amp that can burn tweeters. We can burn power devices in the amp very fast, and traditional censors are, often, not fast enough to prevent this to happen in case of short circuits of the outputs. other methods, like detecting currents have their own drawbacks.
Comparing input and output allow the detection to be in µs. And almost any kind of danger in a while without modification of the amp itself.
This said, to jwilhelm, no SMD neither micro-controller fobia from may side. On the contrary. SMD are better for audio, both because they can reduce sizes of tracks and loops, and can help to find better components, nowadays. But both are not so easy for DIYers, because too small, for SMD, and too "complicated" to program for an average DIYer, witch only have a soldering iron, in case of µCs.
Andrej (Lazy Cat) took this nice decision for its VSSA/first one. He decided to propose pre-build circuits with all the SMD parts on it, for the previous reasons. Then he was relegated by the administrators of this forum in the commercial area. Reason why, i believe, he don't want to share anymore any "work in progress".
I would be very happy if we can discuss of all this in my protection thread: I don't like to work alone, and miss cooperative work and sharing of ideas.
Check out Quasi's DC detect circuit. I've been using it with his NMOS (2 pair) power amp for a long time now, pretty reliable.
https://sites.google.com/site/quasisdiyaudiosite/dc-detect-speaker-protection
Hari
https://sites.google.com/site/quasisdiyaudiosite/dc-detect-speaker-protection
Hari
I now have a good idea of the 1237 internals. Just have to convert its
block diagram to discrete (below 1).
(circled) I have ...
-The DC offset circuit - 3-4 bjt's , diodes , caps.
-The overcurrent ... the opto
-I will use a latching flip flop (below 2) - 2 bjt's.
- The 2 bjt multivibrator's (below 3) R/C will reset the flip-flop after 5 seconds -
after it is initially triggered by an inrush surge. This will turn the relay
driver on.
- Power off can just be the quicker timing of the auxiliary supply's collapse
turning the relay off before main rail collapse.
There is my 8 bjt circuit - should work well.
OS
block diagram to discrete (below 1).
(circled) I have ...
-The DC offset circuit - 3-4 bjt's , diodes , caps.
-The overcurrent ... the opto
-I will use a latching flip flop (below 2) - 2 bjt's.
- The 2 bjt multivibrator's (below 3) R/C will reset the flip-flop after 5 seconds -
after it is initially triggered by an inrush surge. This will turn the relay
driver on.
- Power off can just be the quicker timing of the auxiliary supply's collapse
turning the relay off before main rail collapse.
There is my 8 bjt circuit - should work well.
OS
Attachments
And damn slow.
If I insist on the speed, it is for the following reason. If, like I do, you use very high efficiency speakers and overkill amps with rails at 80V or so, imagine that one of your power devices or VAS die in such a way that the full rail voltage is applied to the speaker. And that, murphy is never far, the rail is the one that move your speaker coil in direction of the back field plate. The fragile cylinder (In aluminium or Kapton) that support the moving coil will hit this plate and be deformed. Speaker is ruined, decentered, rubbing in the air gap, long before the moving coil itsel has time to smoke.
Last edited:
off topic
Sorry for this off topic but i want to say Thanks to a good friend ,Gred(coolewater) from Germany . He has sent to me a free package full of parts
I also want to say a thank you to those who offer me free PCB, But i don't know if i have the permission to say their names.
Thanks friends.I appreciate this.
Thimios.
Sorry for this off topic but i want to say Thanks to a good friend ,Gred(coolewater) from Germany . He has sent to me a free package full of parts
I also want to say a thank you to those who offer me free PCB, But i don't know if i have the permission to say their names.
Thanks friends.I appreciate this.
Thimios.
Attachments
Last edited:
