I have built this circuit from an EPE magazine. I could not get the 2 x HUF75337P3 and replaced them with 4 x IRF530. I tried to trigger the circuit with a 9v PM3 battery (voltage is 8.3v) but just striking the positive pole, the battery negative is connected to the circuit negative, it would not work.

When the circuit is switched on, I can measure 50 mv at the coil.

Is their any suggestions, or alterations to the circuit to get it working?

Virus

Attached Images

Coil driver.png (38.0 KB, 43 views)

 

The circuit is designed to work on 21V that can supply a 10A surge so it's not surprizing it won't work on a 9V battery.

Please post a link to the orgional circuit.

IF all you want to do is switch a small relay then use this circuit.
http://www.mech.uwa.edu.au/NWS/How_t..._course/power/

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez
And http://www.silicontronics.com, same screen name as here.

 

Sorry, my fault. I have a 240v primary to 15v 10 Amp secondary transformer connected as indicated, circuit is built as per diagram, except for the HUF’s swapped for IRF’s (power transistors).

Can I use a 9v PM3 to simulate the TTL (trigger) signal?

The idea is to later connect this circuit to a signal generator and drive the coil at different frequencies.

Thanks for helping.

Virus

 

Virus, you said Can you explain what you mean? Do you mean that the power is on, but there is only 50mV across the coil, even with the 9V battery connected to the input?

__________________
Ron (aka Rube)

 

Fliping heck, the circuit I posted needs a diode in reverse parallel with the relay to protect the transistor from the back EMF generated in the coil.

__________________
I also post at the following sites:
http://www.stop-microsoft.org http://www.heated-debates.com
Screen name: Aloone_Jonez
And http://www.silicontronics.com, same screen name as here.

 

@Roff

When I switch the mains on to the circuit, (the 21v 10 amp circuit) I read 5.6 mv across the coil (not 50 mv just checked it again), If I then connect the 9v+ to the TTL connection and the negative to the ground of the circuit, it drops to 1.6mv across the coil.

The TTL should drive the 2N3904’s and they should drive the IRF 530 allowing the 21v 10 amps through the coil.

I don’t think that a relay will cope at ±1000Hz and even more.

Thanx

Virus

 

Have you tested your MOSFETs? I think your IRF530s are toast. If the zener will conduct in the forward direction, as most (all?) will, it's just a diode when TR2 is off. It provides negative feedback, which is very bad in this case. I ran a DC operating point sim on this circuit (a simplified version, with TR2 off). Below are the results.
The dual IRF530 power dissipation is V(d)*Id(M1)=39.3 watts. Unluss you have a big heatsink, one of them will burn out, and then the other, soon after you connect the 9V battery to the input. If you turned it on before you installed TR2, they would have burned out then.
I don't know why they used a zener there. The normal way to snub the flyback is with a reverse biased diode across the coil. It would have to be a fairly big one, capable of handing 10A pulses. Maybe that's the purpose of the zener. It won't draw significant current. You can add a diode in series with the zener to eliminate the negative feedback, but the zener is a very bad idea. I can show you why in a later post. I would get a 10A diode and put it across the coil.
Below is the schematic I simulated. Note that I used a diode instead of a zener, because it would only normally break down due to the inductor flyback voltage when TR2 turns on. I didn't have a model for a 20v zener (although I could easily make one).

Attached Images

MOSFET toaster.PNG (8.0 KB, 14 views)

__________________
Ron (aka Rube)

 

Here is a sim of the original circuit, with a diode added in series with the zener to solve the problem of negative feedback. Unfortunately, the dissipation is still high during turn-off, as you can see. Not shown is the fact that the current through the zener during turn-off goes through TR2 (Q1 in my schematic), limited only by the beta of Q1, so that current could be on the order of 100mA while the coil is discharging, stressing Q1, the zener, and the diode.
I removed the zener and replaced it with a diode across the coil in the second schematic. Notice the drastically reduced power dissipation in the MOSFETs. I also got rid of a couple of unnecessary resistors.

What is your point about the relay running at 1000Hz?

Attached Images

	MOSFET toaster fix sch.PNG (27.0 KB, 11 views)
	MOSFET toaster fix waves.PNG (33.0 KB, 4 views)

__________________
Ron (aka Rube)

 

My theoretical knowledge of electronics are zip, have built a couple of kits/projects, thanks for your help I really appreciate it.

Can see in your circuit the switch off time is also faster than the original circuit

This is the original text from the document regarding the circuit I built:
For anyone interested in doing further experiments, the circuit diagram of a second, very simple prototype of my pulser is shown in (Fig.5, posted above). The input is a simple TTL square wave at the frequency of interest. The circuit consists of two simple buffer/inverter stages, around transistors TR1 and TR2, to translate the TTL signal into a larger voltage swing to drive the two parallel MOSFETs in the output stage, shown combined as TR3, into hard conduction. The MOSFETs are wired in parallel to increase their current capability and mounted on a big heat sink, although they have a very low on-resistance and so the dissipation of the circuit is low. The 20V Zener diode, D1, wired from the drain to gate of the MOSFETs prevents inductive voltage overshoot and gives the MOSFETs a very hard switch off edge, which is supposed to be ideal for bioactive effect. The power supply is a simple 21 volt unregulated supply (a rectified 15V a.c. transformer with a 4700µF smoothing capacitor) but which needs to have a high surge capacity as the transients generated by the coil can overload the transformer and rectifier. The second prototype uses a 200VA transformer and 10A rectifier which works well. My first prototype used only a 100VA transformer and died suddenly in a cloud of smoke!

This is the description of the coil: The coil, L1, consists of approx 333 turns of 22swg enameled wire wound on a plastic former of about 13cm diameter (a standard plastic waste pipe coupler) in eight layers, across a span of about 25mm. The turns of the coil need to be tightly wound and well varnished into place to prevent oscillation and heating in operation. The prototype coil gets mildly warm in normal use, about 40°C.

Hope it does make sense to you?

If you have a beter designed circuit that I can build, I will not think twice doing it; should I go with your circuit (posted)? and give you feed back.

Thanx

Virus

 

Reference to the relay was for Hero999 who suggested to use one in his circuit as posted, the site quoted have some nice stuff I can learn from and will go back to that soon.

Virus

 

I put your coil design into this calculator, and it came up with the data below. I'll rerun the sim with the new values for the coil. What is the maximum pulse rate you want to run at?

EDIT: I tried another calculator, and got very similar results. This doesn't mean either is correct.

Attached Images

coil char.png (15.0 KB, 7 views)

__________________
Ron (aka Rube)

 

Ron

I take a guess about 0.5 - ±1500Hz.

Would the coils design (size) determine the frequencies that can be used? One can then get different coils for different frequencies, right.

Virus

 

With your present coil, the maximum frequency you can run is about 30Hz if you want the coil current to reach its maximum value before you turn it off. To run at higher frequencies, you would need fewer turns. Fewer turns would result in higher peak current unless you used finer wire, or put a resistor in series.

__________________
Ron (aka Rube)

 

Ron

I take it at lower frequencies it is not a problem then because of the longer on times and the coil get saturated. With higher frequencies, the circuit switches of before the coil is saturated. Right ?

I understand the fewer turns, but finer wire would mean higher resistance? Higher resistance means less power ?
Please explain ?

Is so could one then switch different resistors in series with the coil to get to higher frequencies ?

Sorry for all the questions.

Virus

 

The time constant of a series RL (resistor-inductor) circuit is L/R. It takes about 5 time constants for the current to reach 99% of its final value. For your coil, the calculators came up with about L=23milliHenries, and R=7.5 ohms. So, 5*L/R=15.3 milliseconds. If this is 1/2 cycle of a square wave, the period is T=2*15.3ms=30.6ms. The frequency is F=1/T=32.6 Hertz. If you make the coil with fewer turns, the inductance will go down and so will the resistance. However, inductance is (ideally) proportional to the square of the number of turns, while resistance is linearly proportional (assuming a single layer, and the same wire diameter). So, if you reduce the number of turns of your coil by half, your inductance will be 23/4=5.75mH, and your resistance will be 3.75 ohms (again, this is ideal). So your time constant will now be 1.53mS, and the max frequency will be 65.2Hz - two times the original value. However, the peak current will double, because the resistance is half. (The peak current is just Vsupply/Rcoil). So, if you make the wire diameter smaller, the resistance will go down, the current will go down, and the max frequency will go up.
So yes, you can add series resistance instead of changing the wire size, and the peak current will drop and the max frequency will go up. However, as the peak current drops, the maximum magnetic field strength will also drop.
Did any of that make sense?

__________________
Ron (aka Rube)