Modelling an L C R power supply

[Kindhornman]

Joshua,
What abraxalito is trying to tell you is that you need to add a four diode bridged rectifier input into your schematic instead of only the two discrete diodes you are showing on your schematic. You need to make a full wave rectifier to have real results and a dc current into the circuit.

Steven

ps. Richard, I hope I got that correct!

 

[abraxalito]

No, not really It will work with two diodes but there's no point in having two, as AndrewT said - two diodes in series is totally pointless. A half-wave is fine for now.

The primary problem here - what's giving rise to zero voltage, zero current results is the lack of relevant sim being set up.

 

[Kindhornman]

Richard,
I did pick up on what you were saying about there being no time component to the sim, that is why I left that to what you said. I thought that part of what you said and how to fix that was very clear.

 

[abraxalito]

Thanks - but Joshua_G hasn't found that part at all clear so far.

Once a transient sim has been set up and starts running the problem of the discontinuous inductor current will rear its head - the sim might take an age to run, or LTSpice might come up with some error message about inability to converge on a solution. Fixing that problem would call for an RC snubber before the chokes.

 

[Joshua_G]

I have looked at the schematic, seems you have not digested what I'm saying about needing to run a transient analysis, not an operating point. Right click on the text '.op' which opens the window where you select the kind of sim, the left most tab is what you want. Fill in details like how long you want it to run and how precise you want the results (min timestep). Bear in mind that a diode feeding an inductor directly will give simulation problems when the diode switches off as the inductor current has nowhere to go....

Okay, thank you. Now the simulation is running. What would you recommend for start time, stop time and number of points?

 

[Joshua_G]

Thanks - but Joshua_G hasn't found that part at all clear so far.

Once a transient sim has been set up and starts running the problem of the discontinuous inductor current will rear its head - the sim might take an age to run, or LTSpice might come up with some error message about inability to converge on a solution. Fixing that problem would call for an RC snubber before the chokes.

Indeed, it take ages. I'll put a snubber before the choke.

 

[abraxalito]

What would you recommend for start time, stop time and number of points?

I'd begin with something like 20 cycles of mains, so about 0.4s then eyeball it to see if its settling down to a steady state. Set the max timestep around 100uS or so, having a bigger number here will speed things up as you're not concerned about very fine details just the ripple amplitude.

 

[Joshua_G]

I made it .tran 0 10 1 10.
With a snubber it give results in an instance. However, neither the DC voltage value nor the ripple voltage value doesn't make any sense. From 380VAC sine I get 146.4V DC with some 0.2V-0.3V ripple.

 

[Joshua_G]

That is ;tran 0 10 1 10.

 

[abraxalito]

Are you sure its reached a steady state? Try changing the inductor value temporarily to 10mH to see if you get more DC voltage.

Also right-click on the sine source to see you've got it set up correctly for the .tran. You'll need the third bullet point down in that menu, 'SINE...'.

 

[Joshua_G]

The sine source is defined correctly. With 10mH chokes I get about 319VDC, which is reasonable, but 1.1VAC ripple, which isn't reasonable.

 

[Joshua_G]

And yes, it reached a steady state. With 10mH almost instantly (I start after 1 second), with 10H after about 3 seconds.

 

[abraxalito]

If the sine source is set to 380V peak, then 319VDC isn't reasonable as far as I can see - that's a huge loss happening somewhere.

 

[Joshua_G]

With choke input, the DC value is about 0.9 of the source RMS value.

With sine source R series 100R (instead of 1R before), I get a ripple voltage closer to what it should be, yet too low. So I wonder how to set LTSpice to give me approximately correct values.

 

[gootee]

Did you right-click on each diode and set them to a real type of diode? (You should also right-click on the other components and set their parasitics.)

 

[godfrey]

With choke input, the DC value is about 0.9 of the source RMS value.

That is correct if you use a bridge rectifier. However the circuit you showed earlier used half-wave rectification, which will give a lower DC output voltage with higher ripple.

 

[Joshua_G]

Did you right-click on each diode and set them to a real type of diode? (You should also right-click on the other components and set their parasitics.)

I just did it, there no change in the simulation results.

 

[Joshua_G]

That is correct if you use a bridge rectifier. However the circuit you showed earlier used half-wave rectification, which will give a lower DC output voltage with higher ripple.

Since I doubled the mains frequency, the 2 diodes should give the same results as diodes bridge with the correct mains frequency.

 

[godfrey]

Since I doubled the mains frequency, the 2 diodes should give the same results as diodes bridge with the correct mains frequency.

No, the results are not the same with a choke-input supply. If you try the simulation with a bridge, you will see the difference.

 

[Joshua_G]

No, the results are not the same with a choke-input supply. If you try the simulation with a bridge, you will see the difference.

Thank you.
Indeed, with bridge rectifier, correct mains frequency and 10 Ohm series R of the sine source, the DC voltage is very close to the expected voltage. However I wonder how accurate is the ripple voltage result.