# transformer smoothing and capacitor peak voltage

#### Brian203707

Joined Feb 13, 2010
5
I have built a variable power supply, but in the building process I have found that my 36 volt ac transformer after bridge rectification with the smoothing capacitor equates to 56 volts. Not having a degree in electrical engineering I had to research this effect. (Peak voltage capitor charging). In a nutshell I have to use a 24volt transformer to stay within the voltage levels of my circuit. Is there a way to keep the 36v transformer at 36 volts even after a smoothing capacitor? I have found that when the circuit is at full load it drops down to the correct transformer output voltage, but I assume this is because the cap maintains the load voltage. I do now know about the RMS formula for transformers VOLTS X 1.414 = RMS voltage.

#### The Electrician

Joined Oct 9, 2007
2,761
You need to give some more details.

What is the current rating of your transformer, what is the current drawn by the load, what is the maximum and the minimum DC output voltage your load can tolerate, what is the size (microfarads) of your filter capacitor? Using an ohmmeter, what is the resistance of the primary and the secondary of the transformer (while not energized, of course)?

When you say that "...when the circuit is at full load it drops down to the correct transformer output voltage.", do you mean that the AC voltage measured at the transformer secondary drops to 36 volts, or that the DC output voltage drops?

#### Audioguru

Joined Dec 20, 2007
11,251
A cheap low current transformer has a lot of resistance in its windigs which causes its voltage to drop when the current is high, so they are made with a voltage much too high without a load so that the voltage drops to the correct amount with rated load current.

The capacitor smooths the voltage pulses from the rectifier. If the value of the capacitor is too low then it cannot smooth properly at high currents and causes a high ripple voltage which makes the average output DC lower.

#### eblc1388

Joined Nov 28, 2008
1,542
I have built a variable power supply, but in the building process I have found that my 36 volt ac transformer after bridge rectification with the smoothing capacitor equates to 56 volts.

That's not surprising. Given that the transformer has to provide 36V output at full load, its terminal voltage at no load or low load could be 5~10% higher so we are looking at 37.8~39.6V RMS. Transformer voltage are without exception specified in RMS.

When this RMS voltage is fed to a capacitor smoothing filter, the peak voltage can reach 39.6*1.414 = 56V and that's what you have measured with a meter. Note that this voltage is only reached with NO or LOW load on the capacitor. Once load is applied, the voltage will fall.
Is there a way to keep the 36v transformer at 36 volts even after a smoothing capacitor?

No. That's what happens with a capacitor filter.
I do now know about the RMS formula for transformers VOLTS X 1.414 = RMS voltage.
That is not correct. RMS voltage = VOLTs (peak) / 1.414 so the peak voltage = RMS * 1.414 as shown in the 56V calculation above.

#### russ_hensel

Joined Jan 11, 2009
825
Normally the next part of the circuit is a voltage regulator. The simplest is the 3 terminal linear regulator, the 7805 being a represenative 5 volt part. Issues of voltage, current, and power disapation should then be considered. -- sorry about the spelling

#### Brian203707

Joined Feb 13, 2010
5
The problem with the peak voltage occurred when I flamed out my LM7812
rated at max 35 to volts input. So the RMS took the voltage over the chip limit. So it looks like I just have to stick with my 25 volt 2amp transformer witch only has an RMS peak voltage of 36 to 38 volts.
I Just thought there might be a way to get that peak voltage down to actual voltage without a load. I am using a 50volt 1000uf cap electrolytic cap but I have tried various size caps but the RMS factor happens with all of them.

#### The Electrician

Joined Oct 9, 2007
2,761
I'm confused. In your first post you say that you've built a "variable" power supply, but if you're using a 7812, the output of the supply would seem to be a fixed 12 volts, not variable. Can you clear up the confusion?

#### Brian203707

Joined Feb 13, 2010
5
LM7805, and LM7812 to power cooling fans and power 5 volt panel meter.
LM350TFS variable voltage regulators from 1.5 to 33volts.

I have 4 separate circuit boards in the case for various functions.
One powers the High voltage flyback for HV experiments, one is the variable circuit with 3 LM350TFS in a parallel circuit, and one has the LM7812 and L7805 for cooling fans and digital panel meter.
The Board with the LM7812 flamed out because of overvoltage from the 36volt 3amp transformer. When I was testing an LM317 it blew that out to, The power supply used with the internal 30 volt battery pack works great pushing a max of 5amps, and 4.2 amps with the 25volt 2amp transformer, Yes I know Im pushing the limits 25volt transformer, but it seems to handle the output just gets HOT!!!.
Sorry for the lack of explanation.
My 36v transformer is rated at 3amps that is why I want to use it stays much cooler when pushing the max amp load at 18volts but I need the higher end voltage for some of my experiments.

#### The Electrician

Joined Oct 9, 2007
2,761
How much current are the 7812 and the 7805 supplying?

By running them off of the 36 volts, you are burning up a lot of power unnecessarily. It would probably be better if you got a 12.6 volt transformer to supply those two regulators; then they wouldn't get so hot.

That would then take some load off of the 25 volt transformer and it wouldn't get so hot. Removing the 7812 and 7805 loads would also allow the 36 volts rectified from the 25 volt transformer to be a little higher.

I notice that the max input voltage for the LM350 (actually the input-output differential) is 35 volts. You're running right on the edge there, aren't you?

#### Brian203707

Joined Feb 13, 2010
5
Yes, I could compensate by using another transformer but I dont have the space for it, primarily the 7812 is powering 2 fans and the 7805 that is powering the digital panel meter, between the 2 fans and the meter it only draws 420ma, they are also connected to external optional hook up ports for 12 and 5 volt testing. The 7812 and 7805 are rated at 1.5A.
I have added a switch for the transformer to switch to the 12.6 volt output when doing the love voltage applications, so everything runs much cooler.
Is there a way to run two bridge rectifiers from the 12.6 v transformer pole and the 25v parallel poles on a common ground. I have not found power supply schematic showing this; therefore I could rewire and always run the 7812 and 05 off the low side of the transformer and the LM350's off the high side when the extra voltage is needed.
Yes, it would appear that I am pushing the limits of the LM350s on the high voltage side but they have held up in all of my testing, and they are only initiated on a separate switch when used. They say differential voltage of 35v but they are handling 39v until I apply a load to bring it down. I think this will just probably shorten the rated life span, but I have burned up a few transisters in my learning endevors.

#### MikeML

Joined Oct 2, 2009
5,444
One crude trick is to use only half-wave rectification, and the smallest filter capacitor that will prevent the regulator from dropping out. The reduces both the peak and the RMS value of the voltage going into the regulator, which minimizes regulator dissipation. Another old trick is to install a "bleeder" (minimum load) resistor across the filter capacitor between the rectifier and regulator.

#### MikeML

Joined Oct 2, 2009
5,444
btw- does your 36V transformer winding have a Center Tap?

#### t06afre

Joined May 11, 2009
5,934
You could also use a simple switch mode circuit for the fans. But I think MikeML will help you a lot. Since you are remowing 50% of the energy by using half-wave rectification. Another trick is purchase a small wall wart and modify it slighty to mount it in your chassie. A small wall wart may not take up mose space than your 5 and 12 volt circutry. I have done this some times

#### Audioguru

Joined Dec 20, 2007
11,251
The datasheet for the LM350 shows that its current limit is reduced as the voltage across it increases above 10V. With up to 10V its minimum current limit is 3.0A. With 30V across it its minimum current limit is only 300mA.

#### SgtWookie

Joined Jul 17, 2007
22,210
Brian,
I noticed you mentioned that your cap (caps, maybe?) is rated 1000uF at 50WV.
You should be using caps rated for twice the expected maximum voltage, or else you will have problems with excessive internal leakage, which will cause heating of the capacitor - up to the point of boiling the electrolyte, causing the package to rupture.

You should be using filter caps rated for 100V or better.

#### The Electrician

Joined Oct 9, 2007
2,761
One crude trick is to use only half-wave rectification, and the smallest filter capacitor that will prevent the regulator from dropping out. The reduces both the peak and the RMS value of the voltage going into the regulator, which minimizes regulator dissipation. Another old trick is to install a "bleeder" (minimum load) resistor across the filter capacitor between the rectifier and regulator.
How do you figure the peak is reduced? Even though a half wave rectifier only provides half as many "half-sine" pulses to the filter cap, doesn't the filter cap still charge up to close to the peak of the "half-sine" voltage?

#### MikeML

Joined Oct 2, 2009
5,444
Every transformer I ever tried this on had some internal resistance which limited the available charging current so that the peak voltage in the filter capacitor is about 5% to 10% lower on half-wave...

Now if you operate the rectifier/filter totally unloaded (no load and no bleeder resistor) then it will charge to the same peak value full or half wave.