EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: Carabus on November 01, 2021, 05:36:19 pm
-
Hi I made 555 timer flyback driver, but my 555 died. It happened after I figured out that somehow if I feed mosfet gate higher than 4v the arcs become hotter and bigger.
What caused the 555 to fail, how would I protect it? I later want to add audio input to the 555 timer so arcs play music.
https://ibb.co/nz9Ky3h
I know that not putting diode across primary can kill mosfet due to high voltage spikes, but I had no problem and all my mosfets worked fine and they never died.
but 555 did. Can voltage spikes go from source to gate?
-
Also Flyback and 555 timer has separate supplies. 555 has 8v and flyback 19v
-
The circuit doesn't make any sense. Why are both sides of the coil connected to ground? The coil will be short circuited and nothing will happen.
Try adding 100R in series with the MOSFET's gate. It will reduce the current surges through the 555's output.
(https://www.eevblog.com/forum/beginners/why-did-my-555-timer-die/?action=dlattach;attach=1313033;image)
-
You're right. I messed up drawing the schematic, the grounding you drew was the one I used.
I will try the resistor.
Another thing that may have occurred is that maybe the transformer arced to the output of 555 as they were pretty close to each other. Later I replaced 555 and it didn't die anymore.
I have another question. Why did the current through the mosfet increase, when I increased gate voltage from 4 to 8v. The arcs became bigger and hotter.
-
I have another question. Why did the current through the mosfet increase, when I increased gate voltage from 4 to 8v. The arcs became bigger and hotter.
4V isn't enough to turn most MOSFETs hard on, especially given the 555's output stage drops just over a volt.
-
Also what I noticed just now that when I increase voltage to 555 the frequency drops significantly from 32Khz at 4v to less than 20kHz at 7v and less at higher voltages.
That is why when I increase voltage at some point arcs are gone. Any idea what may cause this issue?
-
Also what I noticed just now that when I increase voltage to 555 the frequency drops significantly from 32Khz at 4v to less than 20kHz at 7v and less at higher voltages.
That is why when I increase voltage at some point arcs are gone. Any idea what may cause this issue?
Could be ceramic timing capacitor dielectric DC bias dependency. Try C0G or film type capacitor if you want stability.
Also for your circuit, you should add a snubber network for your transformer primary. Maybe zenner diode for gate protection. Generally it would be good to have lower gate driving impedance, because gate might be ringing together with your inductive load. Consider complimentary emitter follower or a proper gate driver.
As for 555 getting damaged, I would think that it is caused by your flyback action and bad layout. Where high voltage, fast transients are present it is easy to get all kinds of interference, sometimes enough to damage things. Good grounding, decoupling capacitors and proper placement of parts is essential.
-
Yes, I don't see any supply decoupling capacitors. You want one close on the package supply pin to ground. Add a 100nF from the control voltage pin (5) to ground too. It's another possible transient path into the chip.
-
Thanks, I am quite green in this field and your input is very valuable for me. I will try adding and probing around to learn more about what's going on.
-
Oh, yes, missing / badly placed bypass capacitors could also affect frequency stability. By the way, is it TTL or CMOS version?
-
Yes, I don't see any supply decoupling capacitors. You want one close on the package supply pin to ground. Add a 100nF from the control voltage pin (5) to ground too. It's another possible transient path into the chip.
Good point about supply decoupling. 100nF is fine for pin 5, but much more is required across the power supply. I'd go for 1µF minimum, 10µF recommended.
Also what I noticed just now that when I increase voltage to 555 the frequency drops significantly from 32Khz at 4v to less than 20kHz at 7v and less at higher voltages.
That is why when I increase voltage at some point arcs are gone. Any idea what may cause this issue?
The voltage characteristic of the timing capacitor would explain that, as mentioned above. To expand. Some ceramic capacitors' dielectric saturates at higher voltages. The dipoles in the molecules all line up with the external field, resulting in a reduction in capacitance. Another option is a film capacitor.
https://article.murata.com/en-eu/article/voltage-characteristics-of-electrostatic-capacitance (https://article.murata.com/en-eu/article/voltage-characteristics-of-electrostatic-capacitance)
By the way, the 555 timer is not specified to work under 4.5V, which might also help to explain why it behaved erratically, with a 4V supply.
https://www.ti.com/lit/ds/symlink/lm555.pdf (https://www.ti.com/lit/ds/symlink/lm555.pdf)
-
It's CMOS
-
A mosfet will get shorted 9 out 10 times when they fail, so the power from the drain will go directly to the 555.
Considering this will likely happen during a coil spike (due not using protection), it can easily damage the 555 regardless of the gate resistor.
Driving a flyback transformer with a diode is not good either, you need a RC snubber that takes part of the back emf spike, but not all. Otherwise you're wasting energy. You'll see how the diode literally burns down in seconds.
This is not a problem for a relay, where bemf is a secondary product, not the main target.
That's why, ex. TVs used 900-1200V transistors to switch the HV transformer, to let the coil "flow" in a controlled manner, allowing part of the bemf, which the transistor would see un series with the supply voltage.
Consider using a mosfet driver and 12V for the gate...
-
Yes, I don't see any supply decoupling capacitors. You want one close on the package supply pin to ground. Add a 100nF from the control voltage pin (5) to ground too. It's another possible transient path into the chip.
Good point about supply decoupling. 100nF is fine for pin 5, but much more is required across the power supply. I'd go for 1µF minimum, 10µF recommended.
Also what I noticed just now that when I increase voltage to 555 the frequency drops significantly from 32Khz at 4v to less than 20kHz at 7v and less at higher voltages.
That is why when I increase voltage at some point arcs are gone. Any idea what may cause this issue?
The voltage characteristic of the timing capacitor would explain that, as mentioned above. To expand. Some ceramic capacitors' dielectric saturates at higher voltages. The dipoles in the molecules all line up with the external field, resulting in a reduction in capacitance. Another option is a film capacitor.
https://article.murata.com/en-eu/article/voltage-characteristics-of-electrostatic-capacitance (https://article.murata.com/en-eu/article/voltage-characteristics-of-electrostatic-capacitance)
By the way, the 555 timer is not specified to work under 4.5V, which might also help to explain why it behaved erratically, with a 4V supply.
https://www.ti.com/lit/ds/symlink/lm555.pdf (https://www.ti.com/lit/ds/symlink/lm555.pdf)
I will try adding decoupling caps.
Also shouldn't reduction in capacitance mean that frequency would become higher? In my case frequency drops :-//
-
Yes, I don't see any supply decoupling capacitors. You want one close on the package supply pin to ground. Add a 100nF from the control voltage pin (5) to ground too. It's another possible transient path into the chip.
Good point about supply decoupling. 100nF is fine for pin 5, but much more is required across the power supply. I'd go for 1µF minimum, 10µF recommended.
Also what I noticed just now that when I increase voltage to 555 the frequency drops significantly from 32Khz at 4v to less than 20kHz at 7v and less at higher voltages.
That is why when I increase voltage at some point arcs are gone. Any idea what may cause this issue?
The voltage characteristic of the timing capacitor would explain that, as mentioned above. To expand. Some ceramic capacitors' dielectric saturates at higher voltages. The dipoles in the molecules all line up with the external field, resulting in a reduction in capacitance. Another option is a film capacitor.
https://article.murata.com/en-eu/article/voltage-characteristics-of-electrostatic-capacitance (https://article.murata.com/en-eu/article/voltage-characteristics-of-electrostatic-capacitance)
By the way, the 555 timer is not specified to work under 4.5V, which might also help to explain why it behaved erratically, with a 4V supply.
https://www.ti.com/lit/ds/symlink/lm555.pdf (https://www.ti.com/lit/ds/symlink/lm555.pdf)
I will try adding decoupling caps.
Also shouldn't reduction in capacitance mean that frequency would become higher? In my case frequency drops :-//
Good point. The frequency should drop, so it can't be that. The only other thing I can think of is poor supply decoupling. Perhaps the noise generated by the coil, as the voltage increases, is interfering with the 555 timer. Have you tried probing at the supply voltage rail, with an oscilloscope?
-
I haven't i will do that. Maybe putting 555 in a metal enclosure help with noise?
-
I haven't i will do that. Maybe putting 555 in a metal enclosure help with noise?
You could try, but the noise is most likely to be conducted through the power lines, so it's unlikely to help. Better supply decoupling will do more good.
-
I will add 10uF to the power rails of the 555 as suggested. Ultimately I would like to play music with those arcs, but I need good cooling solution for the mosfet and I want arcs to be bigger which means raising gate voltage (the problem I'm having) and primary coil voltage and adding protection for mosfet from the reverse spikes of the coil.
-
What fet are you using?
-
Also would powering 555 from a battery help?
-
Irf840
-
Not good for 5Vgs. Go for 15-18V.
Doesn't have a too big gate capacitance, but the output of the 555 is pretty weak, causing slow switching times.
Get a mosfet driver... TC4451/4452 are very fast, but you also need good decoupling caps to deal with the fast current spikes.
Two 100nF ceramic caps were ok-ish, but what made a really huge difference was adding a 10uF tantalum cap.
Last time I played with them, I was able to switch a a bigger mosfet at 200KHz with pretty fast edges.
-
Today I tried the decoupling caps. I would say capacitor on the powerlines did nothing significant, but capacitor between pin 5 and ground helped to stabilize the frequency a bit. Now its 20khzat 7v, but still not great.
-
Ok so I might have figured out why the frequency drops significantly. I think that is because the circuit is wired on a breadboard and jumper wires collect interference, because I found that placing my hand on the jumper change frequency aswell (makes it higher).
-
Yes, for something like this (few components and high noise levels) it is worth abandoning breadboard construction as quickly as possible. Try dead bug / Manhattan style construction on a piece of copperclad. That will give you a nice big common ground plane.
-
Ok so I might have figured out why the frequency drops significantly. I think that is because the circuit is wired on a breadboard and jumper wires collect interference, because I found that placing my hand on the jumper change frequency aswell (makes it higher).
That's no surprise. The general rule is, make it as small as possible.
-
Also today I bought NE555 chips (I had TLC555) and surprisingly the frequency did not shift with these at all. All the wiring is the same and on the breadboard.
-
Not sure if this situation applies some what here . I had similar problems with CMOS TLC555 timers frying with a Tesla coil project . It turned out to be the EMR coming from the coil . Moved the driver board farther away from the coil and the 555 timers stopped failing. Perhaps shielding may be in order .