Since I am working on a project to try to make use of my gigantic toroids, I wonder if anyone has thought of the idea to use thyristors to start a toroid at half duty cycle.
I had one of them connected to a 1.75 amp trip current circuit breaker, testing it out. When I plugged it in, all the lights went dark and then the service panel breaker blew but not the little one going directly to the transformer! I knew a startup mode was mandatory, then.
I find that trying to start it with thyristors using simple circuitry which I want to use causes timing imbalances due to inductive current lag. I have noticed that as the duty cycle is increased toward half, this problem becomes less and less. My thought is that half duty is the asymptote. Wow! Did I remember that math term and use it properly?
Well, I guess if once the duty cycle is started at slightly less than half, the problem will be virtually impossible to occur. Has anyone else encountered or looked at this idea? Edit: I mean slightly more than half duty cycle. It would be slightly less than half of the total time in triggering delay.
I had one of them connected to a 1.75 amp trip current circuit breaker, testing it out. When I plugged it in, all the lights went dark and then the service panel breaker blew but not the little one going directly to the transformer! I knew a startup mode was mandatory, then.
I find that trying to start it with thyristors using simple circuitry which I want to use causes timing imbalances due to inductive current lag. I have noticed that as the duty cycle is increased toward half, this problem becomes less and less. My thought is that half duty is the asymptote. Wow! Did I remember that math term and use it properly?
Well, I guess if once the duty cycle is started at slightly less than half, the problem will be virtually impossible to occur. Has anyone else encountered or looked at this idea? Edit: I mean slightly more than half duty cycle. It would be slightly less than half of the total time in triggering delay.
Indeed, any DC component with associated flux imbalance does cause severe problems. I wanted to use the thyristors because i need to regulate power to an electric heat system I need to be isolated from the mains.
I am using SCRs and rectifiers to allow the chance for balance. I use the same timing and trigger circuit for both of them. They both get trigger pulsed each half cycle even though one of them is reversed biased.
I am using SCRs and rectifiers to allow the chance for balance. I use the same timing and trigger circuit for both of them. They both get trigger pulsed each half cycle even though one of them is reversed biased.
I hope to get away with it on the primaries because of the damping of the resistive load and because it is important for me to try to leave out a separate soft start circuit.
I read here quite a while ago that the problem with transformers starting is when the residual flux is in the same direction as the start-up flux and in that case when the plug is inserted during zero crossing. Double whammy.
I hope to avoid having to try my dual bidirectional MOSFET circuit (Thanks to LTspice.) because it is not quite as simple.
I read here quite a while ago that the problem with transformers starting is when the residual flux is in the same direction as the start-up flux and in that case when the plug is inserted during zero crossing. Double whammy.
I hope to avoid having to try my dual bidirectional MOSFET circuit (Thanks to LTspice.) because it is not quite as simple.
Thanks for your thoughts, especially since I have been seriously slaving on this project. I just remembered that if I cannot get it reliable the way it is, I can resort to placing the primary winding in series with the secondary rated at the same voltage, increasing the impedance. Then I will have to deal with needing twice the turns in my own secondary to get the same voltage.
I have four of these things I can use all at once if I get desperate. I can place all the primaries of the four in series if necessary and place the secondaries in parallel. Then I can even avoid winding my own secondaries.
I have four of these things I can use all at once if I get desperate. I can place all the primaries of the four in series if necessary and place the secondaries in parallel. Then I can even avoid winding my own secondaries.
I agree with aandy. The use of TRIAC it is the best solution in 110 or 220 Vac mains control. There are for exampe the ST BTA40-400B (for USA) of 40A. To control the duty cycle of TRIAC you need : 1) A ramp generator 2) A comparator 3) A zero crossing detector of the mains cycle such an optocoupler 4) A controlling pot. in one of the inputs of comparator or a soft start circuit in this place (R-C) 5) An optotriac to isolate the low voltage part of circuit from the mains voltage around TRIAC.
The SCR it is offered for controling of DC current or of the half wave of the AC current. Thus for full wave you need two Thyristors (SCR). Instead TRIAC it is offered for controlling of the full swing of mains voltage. If you want i can post ready and tested circuits. The only that i don't understand it is: you are of controlling DC part or AC part in your arrangement?
Fotios
The SCR it is offered for controling of DC current or of the half wave of the AC current. Thus for full wave you need two Thyristors (SCR). Instead TRIAC it is offered for controlling of the full swing of mains voltage. If you want i can post ready and tested circuits. The only that i don't understand it is: you are of controlling DC part or AC part in your arrangement?
Fotios
I just need AC for my project. I devised a circuit using my favorite IC, the 74C14 which does zero crossing detection, but then I realized that it needed a separate power supply for its power. It wanted a lot of power to trigger the SCRs.
The SCRs take the place of the two positive-going legs of a bridge rectifier. The AC input and the transformer primary are in the AC portion of the bridge. The negative connection of the bridge is grounded and the cathode terminal of the SCRs are also grounded.
I found using the classic diac for triggering to be simple like I am hoping for. I am using a constant current source to give a smooth ramp for the timing of its trigger capacitor. It helped a lot, apparently. I am using an optocoupler, so I need one power supply for it. A single transistor shunts current from the optocoupler diode when the temperature reaches the set point. I'd show a diagram, but I am a messy diagrammer to other people.
Edit: The transistor base-emitter junction itself is the thermostat.
The SCRs take the place of the two positive-going legs of a bridge rectifier. The AC input and the transformer primary are in the AC portion of the bridge. The negative connection of the bridge is grounded and the cathode terminal of the SCRs are also grounded.
I found using the classic diac for triggering to be simple like I am hoping for. I am using a constant current source to give a smooth ramp for the timing of its trigger capacitor. It helped a lot, apparently. I am using an optocoupler, so I need one power supply for it. A single transistor shunts current from the optocoupler diode when the temperature reaches the set point. I'd show a diagram, but I am a messy diagrammer to other people.
Edit: The transistor base-emitter junction itself is the thermostat.
As i understand Electrone you have made a complex circuit operated exclusively from the mains electric supply. IMHO it is a difficult approach (at least to us in Europe because the 220Vac mains) and some sensitive in damage. Why you are not using a seperate control circuit operate with low supply to control the heavy mains load? There are many inexpensive parts (such temp. sensors) and the cost for a small dc supply and few op. amp. and one or two opto couplers and one Triac it not exceeds as i think the 40usd.
Fotios
Fotios
I think it sounds more complicated than it really is. It presently uses only two transistors, an optocoupler, a diac, and two SCRs. That reminder about sensitivity to mains power surges is good to have. I want to be sure to include a 125VAC MOV. Our 120VAC mains helps here because of the lower voltage although the timing capacitor charge transistor (constant current source) is rated 350v C-E. The timing capacitor is also the trigger current supply capacitor.
You still did well. That stuff pertaining to your question is on the secondary side. The primary side circuitry we are dealing with isn't affected by them much since it is a resistive load still. One thing which particularly caught my attention is your mention of a timing ramp. I noticed that the usual timing arc in thyristor type circuits makes short, low power pulses very hard to obtain stably.
I plan to apply a foil heating strip in an up and down, back and forth fashion to a 4 foot section of copper pipe. The electrical insulation between the strips and the pipe is thin for ease of maximum heat transfer. If the film layer becomes breached, there will not be a short to ground.
I will prefer to step down the voltage because then any shorts in the foil will burn away easier, somewhat like how a film capacitor does it. I hope that the shorts present between adjacent tracks will also burn away. The aspect ratio of the width to the thinness of the foil is crucial in that self-healing process.
I plan to lay a wide strip on a 4 foot long piece of two inch wide tape and slice through it in an up and back pattern, forming a longer more resistive path than the old way which i used--one wide pass down the pipe. That way required 200 amps at 5 volts of power. I may use either three or five tracks. Three is about optimal, I think, since the resistance is increased 9 times over just one pass, yet the aspect ratio is hopefully still wide enough for self-healing. The odd number of tracks allows access points at each end.
I plan to apply a foil heating strip in an up and down, back and forth fashion to a 4 foot section of copper pipe. The electrical insulation between the strips and the pipe is thin for ease of maximum heat transfer. If the film layer becomes breached, there will not be a short to ground.
I will prefer to step down the voltage because then any shorts in the foil will burn away easier, somewhat like how a film capacitor does it. I hope that the shorts present between adjacent tracks will also burn away. The aspect ratio of the width to the thinness of the foil is crucial in that self-healing process.
I plan to lay a wide strip on a 4 foot long piece of two inch wide tape and slice through it in an up and back pattern, forming a longer more resistive path than the old way which i used--one wide pass down the pipe. That way required 200 amps at 5 volts of power. I may use either three or five tracks. Three is about optimal, I think, since the resistance is increased 9 times over just one pass, yet the aspect ratio is hopefully still wide enough for self-healing. The odd number of tracks allows access points at each end.
Thanks aandy and fotios. I think it looks better now when I plug the transformer in without a load. Then I plan to switch the power to the control and load. But I may still be able to switch the input power to the transformer instead, depending on if I only draw 900VA from it. So the load would be about 12 ohms if I decide to use the secondary there already at 120v. Or maybe it is that I improved the circuit last night so much it works all the time even with maximum load. It only has two transistors now and no optocoupler and no soft start components. When something doesn't work for me, simplification often is contained in the solution. Maybe it works now, and it was doing better so far.
I just had time to begin to test that latest control circuit. I was able to plug in the power transformer with that 12ohm load connected without overload several times so far. Tomorrow, more tests.
So, for general soft start I thought of an idea which takes this information into account. Use a circuit in which you close a switch to send power to only the transformer to power it first. Once it is on, it can power up a relay after a couple seconds delay, completing the circuit to the rectifiers and filter capacitors. No resistors to burn here if the relay circuit fails.
I just had time to begin to test that latest control circuit. I was able to plug in the power transformer with that 12ohm load connected without overload several times so far. Tomorrow, more tests.
So, for general soft start I thought of an idea which takes this information into account. Use a circuit in which you close a switch to send power to only the transformer to power it first. Once it is on, it can power up a relay after a couple seconds delay, completing the circuit to the rectifiers and filter capacitors. No resistors to burn here if the relay circuit fails.
LTspice Simulation of Circuit Available
I just finished the latest back-to-the-simulator step in the back and forth circuit refining process. Sometimes it helps to go to the real circuit to test out ideas developed using simulation. Later, at times, it helps to take the changes added to the real circuit that made it work due to what was not checked for in the simulator and then go back and change the simulated circuit to that way. This is a strong learning tool and a great boost in circuit development speed. I have added the latest SCR regulator to the files section of my group.
The simulated version contains snubbing capacitors and the 33kohm rectified-supply pull-down resistor which I can try adding to the real circuit next time. The 33k resistor was something I realize now may help the real circuit most especially.
I just finished the latest back-to-the-simulator step in the back and forth circuit refining process. Sometimes it helps to go to the real circuit to test out ideas developed using simulation. Later, at times, it helps to take the changes added to the real circuit that made it work due to what was not checked for in the simulator and then go back and change the simulated circuit to that way. This is a strong learning tool and a great boost in circuit development speed. I have added the latest SCR regulator to the files section of my group.
The simulated version contains snubbing capacitors and the 33kohm rectified-supply pull-down resistor which I can try adding to the real circuit next time. The 33k resistor was something I realize now may help the real circuit most especially.
Be careful about using a triac for soft start. Simple zero crossing of the voltage will not do it because the current is 90 degrees out of phase with the voltage - its the zero crossing current point that you have to switch. One other issue is to make sure the triac switches cleanly in all quadrants otherwise you get some funny mains waveforms applied to the transformer, satuaration and blown fusues.
Don't get me wrong, you can do it, but its not quite as simple as it sounds.
I use 4 x 220 Ohm 7 watts in parallel via a relay in series with the transformer for soft start. They are in circuit for about 2 seconds (transformer in rush and smoothing cap pre-charge time). After this, I switch a second relay in which shorts the pre-charge relay and caps out. I've had no problems.
Transformer is a 15 Kg 2kW beast wound for low noise (i.e. audio grade). I asked the supplier what the in-rush current was and they told me 500-600 Amps.
Don't get me wrong, you can do it, but its not quite as simple as it sounds.
I use 4 x 220 Ohm 7 watts in parallel via a relay in series with the transformer for soft start. They are in circuit for about 2 seconds (transformer in rush and smoothing cap pre-charge time). After this, I switch a second relay in which shorts the pre-charge relay and caps out. I've had no problems.
Transformer is a 15 Kg 2kW beast wound for low noise (i.e. audio grade). I asked the supplier what the in-rush current was and they told me 500-600 Amps.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.