Cycloconverter

OMG: Another project with nil research nor ultimate parameters by the OP.

This looks dangerous, make sure you get competent help if you build it.

For each semi you need to do a "back of envelope" calculation to find what is the maximum current it will carry and then what will the maximum voltage across it be (remembering lamps can draw higher current when they are cold).

Once you've got a Current & Voltage value, double them and look up a suitable part in a catalogue. Your supervisor should be able to guide you in the design process.

Interesting circuit.

What happens when the filament in the lamp breaks?

Okay.. I think i screwed this one up, I'm pretty bad at electronics. My primary project was in Matlab, where this circuit works, wont work in real life i guess.. In the actual circuit i'm making, i've used SCR's in the bridge not diode, I just thought that turning on and off two switches would've made things simpler. Thanks for the help anyways.

The circuit does need looking at again. Everything is just fine while either one or the other transistor is turned on. Turn both on at the same time and the two power sources will be "looking at each other". The rectifiers will turn to ashes pretty quickly....

How about some fuses?

Equivalent circuit, from where I got the idea: Image 1: Now, this is the original circuit from the text book, i wish to make. The author has used SCR's here which are triggered two-at-a-time to work as a bridge. I, however, modified the circuit because triggering each SCR would be complicated. Also I've not shown in the circuit diagram, but the ac supply is provided using a Dimmer-stat. The idea of using fuses is good and I'll try that. Keeping these requirements in mind can you please suggest me the components that I should use. 1. Load is a light bulb of 230V,15W(I can use upto 100W if needed). So a load current less than 0.5A can be expected. 2. Using a dimmer-stat, the supply voltage can be varied between 0-280V. The input frequency will remain constant at 50Hz. By the switching, the output frequency is to be stepped down to 25Hz. 3. If the transistors are used, the base will be supplied with a DC voltage via an IC555 timer and a not gate(IC7404) so, that both of them aren't conducting at a time.

Rajankin,

Welcome to power electronics. There are very few of us...

It is surprising to see that you are considering this 30+ year old technology. I have built one in my younger days but quickly moved to bipolar transistors inverters for the following reasons:

1-While the SCRs are (were) less expensive than the equivalent transistors, the higher switching frequency produced a higher quality result.

2-Protections are easier to implement with switches that can be turned off within uS (transistors). SCRs relied on semiconductor fuses that must be replaced each time a short is produced. This is very common in most product. You want your product to protect itself and re-start without opening the box and changing the fuses.

3-Your circuit will need a relatively large balance chokes between the bridges to allow the SCR conduction around the zero crossing. There will be a large distortions if you don't allow both bridges to conduct simultaneously (circulation current by the balance choke).

4-The load filter is much larger when the voltage ripple is 240Hz (cycloconverter) than when it is 10-20KHz (transistors).

5-SCR technology and their gate drivers are a dying breed. It is an interesting academic subject but you will be better served investing your time developing a nice inverter. You can find various applications notes and development boards from the semiconductors manufacturers. This will kick start your project.

6-Cycloconverters never worked very well on single phase applications. You need three phases to have a decent result. The only exceptions were for electric trains but are not really used anymore.