Rectifying twin secondaries

[multime]

I have a toroidal with 2 x 12v secondary's which i need to connect in series 24vac output. If do the series connection and bridge rectify it can i also bridge rectify one of the 12vac secondary's as well as I need a 12v output to power a relay, the coil of which needs 100/120mA.

[Zero999]

Connect the two windings in series and a bridge rectifier to either end. The centre tap then becomes 12V with respect to the common.


Note any smoothing capacitors (not shown on the above schematic) will change up to the peak voltage, which is equal to the peak * √2, so for two 12V secondary windings you'll get 17V and 34V out.

[jonpaul]

beware of ploarity of sec wdgs, (dot = start) see mfg spec sheet for color code

j

[Ian.M]

It depends.  I think Zero999 has misunderstood you as you want 24V AC out.

If you use a bridge rectifier on one half of the total secondary, the resulting DC circuit *MUST* have no other common connection to the 24V AC circuit as its negative side swaps between the end of the winding feeding the bridge and the center tap with each half cycle.

Its probably better to make a full wave rectifier with two diodes, one from each end of the 24V winding, which gives you DC with respect to the center tap.

See Hammond Manufacturing Ltd's Design Guide For Rectifier Use, the "FULL WAVE Capacitor Input Load" diagram. 

N.B. for a light load, the DC voltage will be closer to 18V^* than 12V, so you will probably need to either regulate it down to 12V or put a dropper resistor in series with the relay coil.

Regarding Jonpaul's concerns: As you are connecting the windings in series getting the relative phasing wrong is non-hazardous.  If you ge tit wrong, you will simply get a very low AC voltage where you expect 24V RMS, but each end to the center tap will still measure near 12V RMS.  To fix this, swap the connections to ONE and only one of the secondaries.

If you were connecting them in parallel mis-phasing would cause a very large current to flow which could damage the transformer and would trip any reasonably rated circuit protection in the mains supply circuit.
*DONT* *DO* *THAT* !!!

* Q: Why 18V instead of the 17V Zero999 calculated?
A: Transformer nominal secondary voltage is specified at the secondary's full rated load current, and rises when less heavily loaded.  The percentage the no load voltage is greater than the full load voltage by is known as the regulation factor, which   is commonly in the 5% to 10% range for medium sized transformers.

[multime]

Zero999, so the 12v would need no further rectification.

jonpaul, thanks for the info i was wondering what the dot indicated.

 Ian.M, the 24vac will be bridge rectified then fed to a 7824 vreg this will be used to supply part of the control board for a power supply. For the 12v supply for the relay  i had considered using a 7812 vreg  connected to the same output from the bridge as the 7824 vreg but the large amount of header voltage present after capacitor smoothing of the 24v may have been to much for it, I need to look at the specs for the 7812 as i may be worrying unnecessarily. The design guide you pointed me to is very helpful.
 

[mikerj]

It depends.  I think Zero999 has misunderstood you as you want 24V AC out.

I believe the OP ultimately wants a DC voltage "If do the series connection and bridge rectify it can i also bridge rectify one of the 12vac secondary's as well?"

[Ian.M]

OK, in that case Zero999's advice above is good.

[multime]

This is a power supply i built many years ago it was sold by Velleman as a kit, i never bought the kit but used the circuit for the parts i could get hold of. I have attached a pdf showing the circuit, at the top and top left it shows the switching arrangement for the two windings ( i never used that part of the circuit ,used an alternative.) The principal was you would have one of the twin secondary's engaged up to 12v-ish above this the relay would kick in and join the secondary's in series so you now had up to 30v at the output. But you can see that there is double rectification of one secondary, the bridge and a half wave rectifiers. Am i right in saying that.

[p.larner]

have you the full article for that psu,or a link,looks like a usefull psu to build.

[Ian.M]

Your Google-fu is lacking.
Google: october 1992 maplin magazine

When you find it, its pages 40-47 and is a 30V 10A 'Laboratory' PSU.   N.B. its worth looking for the next couple of issues in case there were any errata.

[multime]

p.larner, have you now got the article if not i have it. Here are the pcbs plans anyway 

[Ian.M]

p.larner, have you now got the article if not i have it. Here are the pcbs plans anyway 

That's got good enough PCB artwork for the PSU itself for home etching. 
There's nothing on the parts list that is too difficult to source, so its still buildable over thirty years later. 

ICL7107 3 1/2 digit voltmeter/ammeter kits are readily available from Amazon, eBay and the usual far east ebazars. so the lack of the display PCB artwork isn't a show stopper.

[multime]

It was a good power supply mostly trouble free, it had a couple of faults sometimes there was no voltage output regardless of the voltage pots position, it didnt happen often. May have been a fault during construction. Its other fault was that the output never went to zero volts, i did cure that eventually cant remember how though. It had three digital displays, volts, current, and temperature (monitored heatsink and bridge rectifier) all built using ICL7107,s. I also did not use the nichrome resistance wires instead using an 0.01 ohm resistor which fitted in with the requirements of the 7107 current display.

Home etching, great stuff