After the resistor I did not put the condenser, did not assume such effect... It was necessarily necessary? What for stabilizer fluctuations? Yes also top mosfet has burnt down. Whether there can be still any reason of an event?
Attachments
Hi Vogor,
Is correct use +18V (not from power-vcc) for 7812.
big cap before 7812is ok, after (7812) use small max 10uF+100nF near to pin-out.
Is correct use +18V (not from power-vcc) for 7812.
big cap before 7812is ok, after (7812) use small max 10uF+100nF near to pin-out.
Did you started your amp without any heatsinks??? 😱 At +/-65V rails even at idle the output mosfets will overheat very quickly!
Also, at startup IR2110 consumes very low current, and there was not enough voltage drop on 4.7K, so 7812 was still powered with about 60V at least, giving about 30-40V at output 🙄 So it was not very good idea to use 4.7K resistor to GND 🙁 It was very simplier to disconnect the fans temporarily, to test the functionality 😱
How you represented the driver Vcc is derived from -65V rail via 7812 stabilizer except as through the resistor? Without amplifier connection your scheme of start/deenergizing worked.
I did not recommend to do such connection at all! I was thinking previously, that it is connected in such way, but I was wrong. 🙄
My scheme cannot blow any IR2110, so the reason is either you started your amp without heatsinks or 7812 gave too much output voltage from unknown reason (because of oscillation, for example)...
Last edited:
The amplifier worked stably and without heatsinks, the unique reason which is possible this formation of parasitic communication through a PSU, explosion has shown that for your scheme of deenergizing the obligatory external stabilizer for 7812 is necessary. You wrote that probably fluctuation 7812 it that for effect, you already saw such?
Yes, I have seen oscillating 78XX very often. They can even heat from such reason, when there is nothing connected to the output! 😀
For my startup/shutdown schematic is important, that the driver vcc will fall down AFTER the main power supply. Disconnecting the fans from your 18V supply would provide this effect easily.
are you sure, that the amp was started in this case? you wrote, that it was needed to give some small input signal, before it starts to work 🙄
Then why you needed the startup/shutdown circuit at all? Only because of small PUP at shutdown?? 🙂
Yes, I wish to eliminate this PUP at shutdown, then still protection and a limiter will need to be added 🙄
Also, did you used the Zener diode parallel to C13? With 100K resistor at driver idle there can be produced too high voltage also!
Also, I think I understand the reason of this failure... In my startup/shutdown schematic I have seen in simulations, that the 100uf cap will charge to 15V in 2-3 seconds via 100K resistor (providing SD delay) with 65V rails, and if you connected my startup/shutdown circuit, then it has produced the startup delay, and your 47uf C13 cap have charged via 100k resistor in 2-3 seconds to about 30 Volts!!! (at idle the driver consumes very small current) Sure it has killed your upper mosfet after startup! And I have recommended you to use the Zener diode parallel to C13 already in previous posts! 😉
LOL! I have seen in IR2110 datasheet now, that after active SD, the high and low side will be NOT reenabled, if the HIN or LIN is active 😀 So delaying the SD signal at startup will not help to start amp oscillation! 😀 (have AP2 said about it already?)
Also, connecting the 100K resistor between +65V and the amp OUTPUT will help to start oscillation in the same way, as connecting this resistor to Vb! 🙄 It works because 100K resistor produces POSITIVE DC offset at output during startup (via charging C13 or directly) and the input integrator needs to initiate the oscillation with NEGATIVE pulse 😀
Shame on me, that I did not realised it previously 😀
Last edited:
Yes, similar top mosfet has burnt down for this reason, zener in parallel с13 could rescue the scheme... So what your general result taking into account result of my last experience? your scheme of deenergizing will work if to use the separate rectifier for 7812?
yes, this will work! 🙂 The startup delay seems to be useless, but since you have implemented this already, it will not interfere. 😉
One more moment, what elements in your scheme are responsible for operation and occurrence SD border, at amplifier "PUP" deenergizing appeared almost at once, I think it is necessary to raise a threshold to 55V, what it is necessary to change for this purpose?
- Status
- Not open for further replies.