Electronic fuse

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
and where is pot for current regulation?
There isn't any. The circuit was intended as a replacement for the thyristor circuit.
Adding one is certainly not impossible, but it would be difficult for a simple, two terminal fuse substitute like this.

for the 100-th time... R8 should be less as possible and fixed value

you dont want a fuse of 1kohm
You have to remember one thing: the voltage drop will remain the same, be the resistor 1Kiloohm or 1milliohm.
Take also into account the fact that a real 1mA fuse, like the ones used for patient protection in medical appliances will have a resistance of that order too.

Av trouvaille:
Please stay on the thread. Most members (including me) find private messages a nuisance, but when they are used without a good reason, they become a PIA.
 
Member
Joined 2003
Paid Member
Excuse me for posting unwanted pm’s. It was sent out of modesty.

I have tested the second circuit as presented by Elvee in post 38, which uses an additional transistor to sense the output current and block the restarting if an overload condition is still present. Test conditions were 22V at the input with a variable load of 25W. It works as it should. Elvee, thanks again. What a relief, no more fuses to change in experiments.

I have played somewhat with different values of the sense resistor, which sets the maximum current at which the electronic fuse stops passing current. With R-sense = 1 ohm, the maximum current is some over 1A at a load of 25 ohm. I have made two switchable options in my jig: 1.35 ohm with a maximum of 800mA, and 3.5 ohm for about 200mA. This is what I actually measured as shut off current. Elvee, please tell me if I talk non-sense here.

Some points that may require attention:
- Before turn on, or after shut off there is still a small voltage present at the output. At 75 ohm this results in 2,9V at 39mA (In the first circuit this was 1,5V at 19mA). I guess this is of no harm, unless you want to test at very low current. Elvee, what could be the cause of this?
- A nice visual feature to add might be a led when the fuse is ‘refusing’ to pass current, and for the case when an overload condition is still present.
- In post 37 is mentioned it is a two terminal circuit, and the transistors could all have their polarity reversed. Does this mean it is useable in a positive as well as a negative supply without further changes?
 
I have played somewhat with different values of the sense resistor, which sets the maximum current at which the electronic fuse stops passing current. With R-sense = 1 ohm, the maximum current is some over 1A at a load of 25 ohm. I have made two switchable options in my jig: 1.35 ohm with a maximum of 800mA, and 3.5 ohm for about 200mA. This is what I actually measured as shut off current. Elvee, please tell me if I talk non-sense here.
Values look consistent. Values could certainly be optimized, as the calculations were in fact quick estimations.
Some points that may require attention:
- Before turn on, or after shut off there is still a small voltage present at the output. At 75 ohm this results in 2,9V at 39mA (In the first circuit this was 1,5V at 19mA). I guess this is of no harm, unless you want to test at very low current. Elvee, what could be the cause of this?
This is normal: it is caused by the 1K base resistor of the shutdown transistor.
This resistor could be increased, but at the expense of the tripping threshold (it will increase).
- A nice visual feature to add might be a led when the fuse is ‘refusing’ to pass current,
It is possible, by wiring an LED in series with a resistor between input and output, but it will also contribute to the off current.
and for the case when an overload condition is still present.
This is extremely difficult: in order for the circuit to "know" there is an overload, it has to sense it by sending the power.
- In post 37 is mentioned it is a two terminal circuit, and the transistors could all have their polarity reversed. Does this mean it is useable in a positive as well as a negative supply without further changes?
Indeed.
You can build a P version or an N version, and use any of them on either polarity.
The circuit only sees the difference between input and output, and if the polarity of this difference is compatible with the transistors used, it will work.
It is probably preferable to opt for a N version (main transistor): a 2N3055 would do in all cases.
The protection transistor could be a superbeta, ZTX869 or similar: the 1K could be increased to 4K7, or something like that.
If you put the circuit inside a rectifier bridge, it will also work in AC. Some adaptations would be required to cope with the zero-crossings.
 
This one is the adjustable, self-resettable, low off current, N-version.
It remains two-terminal though.
 

Attachments

  • Itrip2.gif
    Itrip2.gif
    30 KB · Views: 106
Elvee,this looks interesting!

How about version with additional reset switch and N mosfet (BUZ11 for example) instead 3055?
A reset button could be added in the same ways as in previous circuits: either a short of BE/Q2, CE/Q1, or one of the more sophisticated options.
Using a MOS isn't really possible: the voltage drop would increase from ~1V in this circuit, to several volts due to the threshold voltage of the MOS.
 
Member
Joined 2003
Paid Member
Another use for the electronic fuse would be in the B+ of a tube amplifier. Here some issues may rise, such as:
- In tubed equipment the voltage is high: usually in the range from 150V up to 500 volts. Above 600V is really another league. As the electronic fuse is a two terminal device without reference to ground, the level of the voltage may not be an issue. Elvee, what do you think?
- As pointed out by Elvee in post 44 some ‘leakage’ of current of the fuse in shut off state is normal, as it is caused by the 1K base resistor of the shutdown transistor. However, in a supply for tube amplifiers, the currents are usually small. So a leakage of 20 or 40 mA might be higher than the total current draw of a tubed circuit. It would certainly be worthwhile to think this through further. The application of, for instance, a relais might extend the trip time considerably though.
 
Another use for the electronic fuse would be in the B+ of a tube amplifier. Here some issues may rise, such as:
- In tubed equipment the voltage is high: usually in the range from 150V up to 500 volts. Above 600V is really another league. As the electronic fuse is a two terminal device without reference to ground, the level of the voltage may not be an issue. Elvee, what do you think?
The voltage to ground may not be a problem, but the fuse still has to be able to block the full voltage.
It is doable however, as bipolars can operate at 1KV and beyond, provided they are used as on/off switches, which is the case here.
- As pointed out by Elvee in post 44 some ‘leakage’ of current of the fuse in shut off state is normal, as it is caused by the 1K base resistor of the shutdown transistor. However, in a supply for tube amplifiers, the currents are usually small. So a leakage of 20 or 40 mA might be higher than the total current draw of a tubed circuit. It would certainly be worthwhile to think this through further. The application of, for instance, a relais might extend the trip time considerably though.
The leakage current is related to the nominal current of the fuse.
As a tube circuit is unlikely to require amperes of supply current, the general level of impedance can be raised.
Using a superbeta transistor for Q2 also helps.
 
Member
Joined 2003
Paid Member
Quote:
"The voltage to ground may not be a problem, but the fuse still has to be able to block the full voltage. It is doable however, as bipolars can operate at 1KV and beyond, provided they are used as on/off switches, which is the case here."

So to be sure, I learn that the transistors should meet the maximum voltage requirements of a tube supply.

A question on fool-proving the electronic fuse:
When used in the ground connection of a supply - which is for instance handy in case of a full wave rectifier with center tap - the connections of the fuse have to be reversed. Is this correct?
 
A question on fool-proving the electronic fuse:
When used in the ground connection of a supply - which is for instance handy in case of a full wave rectifier with center tap - the connections of the fuse have to be reversed. Is this correct?
Of course: if you build the N version, the emitter of the main transistor has to be connected to the most negative point of the circuit, which is the center-tap if you adopt the usual configuration of common cathodes for the rectifier diodes.
 
Member
Joined 2003
Paid Member
On further testing with a scope I notice an oscillation of 20 mV at about 4 MHz when the fuse is idle. After turn on, with a current of 0,3 A at 22V at the output it rises to about 600 mV. Hmm, how to tame this oscillation while keeping it a two terminal device? Any suggestions on a solution, while I will be exploring with a snubber?
 
Member
Joined 2003
Paid Member
The electronic fuse appears to be as sound as can be, in all three variations I have made so far. The scope shows the fuse does not add anything to a straight line, just as the first version of post 34 already did.

It was the stand alone power supply that was about to go south, and caused the oscillation. Excuse me, Elvee. Oh well, you think to start a next project. Looking at the almost empty fuse box, the idea of an electronic fuse is born, and suddenly you are reworking that good old power supply. Now that is fixed, the nest project may really begin.

Just for inspiration, I attached a picture of the simple version of the electronic fuse (upper schematic of post 38). The layout measures 6 x 8 cm, and includes a switch to toggle between two sense resistors. Indeed, somewhat bigger that a real fuse, but more versatile and green. In smd the foot print can be much smaller of course. The layout is straight forward. In order to prevent mistakes in direction I added a diode of 3A at the input, and at the output. To the upper right is the input banana chassis part, the left is the output. Disregard the two banana inputs on the lower right, as they are only there for convenience to connect two ground wires and not connected to the fuse.
 

Attachments

  • Electronic Fuse Elvee post 38 first schem.jpg
    Electronic Fuse Elvee post 38 first schem.jpg
    43.5 KB · Views: 85
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.