Power supply for DC compressor circuit

[dan_thompson87]

I joined the forum last year asking for help understanding the basics of my final year project. Luckily I managed to make what I had in mind and successfully completed the project. I eventually got an A for it, so thanks for helping guys!

It's now a year on, and with me having 100% understanding of my circuit and a working prototype (plus some spare time and a job to pay for it!) I've decided to make a rack mounted version.

The project which I submitted used a +15V and -15V dc power supply regulated down to 12V where necessary. This was from one of the uni's standard lab power supply's, however with my new version being rack mounted I would like to make it mains powered.

My problem is that I have no idea on power electronics. How do I achieve +/- 15V from 240V ac mains? I had a play around with a transformer which stepped the ac signal down to 20V dc and some switch mode power supplies to invert it, but I just spent a lot of money blowing up surface mount chips!

The ideal solution would be to buy an all in one power supply that will transform, step down and invert the mains supply, however I think this may be a bit much to ask for!

Any help will be appreciated, thanks in advance!

[highfieldrebel]

Hi Dan

lots of transformers have two secondary windings, use one for +ve rail, the other for -ve rail. No inverting needed.

S

[dan_thompson87]

Thanks for that, put so simply it suddenly makes a lot of sense. Now I feel silly

A step-down transformer with 2 secondary windings and a +v and -ve rectifier should do the trick right?!

[highfieldrebel]

Either dual secondary windings, or one secondary winding with a centre tap. You don't even need to use two rectifiers, type "dual voltage power supply" into google and hit the Images button, you'll see your options. You will need two voltage regulators though, like the 7812 and the 7912.

[dan_thompson87]

That's exactly what I was hoping to hear. It looks as though it'll be much simpler than I first thought! Thanks again.

[dan_thompson87]

Next question!

I've designed the circuit, but I could do with a bit of help on the maths when choosing the transformer. Basically I'm having trouble choosing the power rating which I require.

The circuit uses very little current as it is made up of passive components and op-amps. There are 30 LEDs being used as a VU meter though, which is where most of the current is drawn.

So onto the maths. Each LED uses 10mA when lit, so 10mA*30 = 300mA. (Assuming they're all lit at once, which is very unlikely!)

Another noob question, but if I buy a transformer with 2 18V secondary windings and wire them in series for centre tap configuration, do I multiply the current by 18V (from ground to +ve or -ve) or 36V (from +ve to -ve)?

Allowing 500mA for the whole circuit should be more than enough, so I've concluded that VA must be 18*0.5 = 9VA OR 36*0.5 = 18VA, depending on which potential difference I use. Can anybody tell me which is correct?

[dan_thompson87]

I've just found this information on a website:

The maximum current that can be pulled in centre tap configuration is half that rated by the transformer (i.e. the rated current for each winding) as you are effectively getting two times the voltage.

I take it this means that for independent wirings 18VAC*0.5A = 9VA and for centre tapped 36VAC*0.25A = 9VA also.

Seeing as I need +18V, -18V and 0.5A I will in fact need a transformer rated at least 18VA (36VAC*0.5A). Can somebody please confirm this for me?

Thanks.

[highfieldrebel]

This is not an easy one to answer usually, but in your case I have even more doubts. How will you construct your VU meter, can you post a sketch? The reason is that I'm not sure you have to use a different 10 mA for each LED, maybe you can make a scheme where the same 10 mA will pass through a number of LEDs.

[dan_thompson87]

The whole circuit is already constructed and working, including the VU meter. It's actually 2 separate VU meters, one for input and one for output. Both are driven by LM3915 chips.

Unfortunately I never drew schematic as the circuit was constructed straight from the data sheet. The best I can do is attach a picture of the PCB to aid in visualising the circuit. It's a simple circuit though consisting of a peak detecting rectifier, 2 cascaded drivers per meter and the necessary reference components.






I do know however that the output pins on the chip have current regulation which I set to 10mA. I'm fairly positive that each LED draws a further 10mA as the compressor is currently powered by a bench supply with current limiting. As more LED's turn on the current limiter needs to be increased otherwise they dim. This is also where I drew my 500mA estimation from!

[highfieldrebel]

So if I understand, the bench supply is +/- 15 V, and the current indicator on the bench supply reads 500 mA for both polarities when the VU reads max?