Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Filter to smooth 30 amp PWM controller input

Status
Not open for further replies.

Gromit

New Member
Hello, I am trying to setup an unusual battery charging system for my RV.

I have a DC/DC battery charger connected to the alternator of my vehicle. The charger takes nominal 12v dc in and then controls its boosted output to provide “two stage” charging. The charger adjusts the output voltage to ensure that it will not draw more than 30 amps on the input (this is stage one), then as the battery charges and the current draw goes down it will limit the output voltage to approx. 13.4v (this is stage two).

I would like to have this charger feed into a Xantrex C60 PWM charge controller which then feeds the battery. The reason is that the batteries never really get fully charged if the max output voltage of the DC/DC charger is set to basically a float voltage, also it takes a very long time to complete the charge since the voltage is relatively low. My intention is to bump up the output voltage of the DC/DC charge controller to say 16v and then let the C60 PWM controller ensure that the batteries are fully charged without overcharging them.

I have confirmed with the manufacturers that the C60 is fine with virtually any DC source up to 60 amps, and the DC/DC charger can be set to output up to 17v.

I have bench tested this setup and it works reasonably well except when the battery gets near the end of its charge and the current get less than about five amps. At this point the DC/DC converter starts to buzz due to the PWM of the C60. Once the C60 decides that charging is complete it shifts to “float” mode where the voltage is kept around 13.2 volts and the current can be as little as 50ma. At this point the output voltage of the DC/DC charger starts to swing wildly and buzzes quite loudly.

I need a filter between the DC/DC charger and the C60 PWM charge controller to smooth out the PWM. I have placed a 4700 uF 25v capacitor across the inputs of the C60 and this reduced the buzzing somewhat as well as fully stabilized the voltage at all currents. I also tried up to six of these capacitors in parallel, but performance was exactly the same. This is an OK solution but I’m wondering if using an inductor as well might make a better filter. I really don’t know much about filter design. I don’t have a scope so I don’t know what frequency the PWM is operating at.

It’s a strange setup I know, but any assistance would be appreciated.
 
An inductor between the DC/DC charger and the caps may help but it would have to be a rather large, high-current inductor.

But it could also lead to strange behavior since the inductor and capacitor will form a resonant circuit that may interact adversely with the DC/DC or the C60 charger. So if you try it, be ready to remove the power quickly if it seems to be acting strangely.
 
Thanks for the reply Carl.

Do you think I would be better off using just an inductor? Can an inductor do anything on its own in a case like this?

I’m also wondering if I would be better off using a much larger capacitor by itself. I could get one of those car stereo caps. I saw one that’s 5 farads 24 volts.
 
use the inductor and the capacitors together. Keep the capacitors on the input of the C60 with the inductor between them and the DC/DC battery charger. You may not see any improvement in the performance by adding more capacitors but it will help de-stress the capacitors themselves. Those caps are supplying AC ripple current to the C60. If the ACrms ripple current rating of the caps is exceeded, they can fail - so more is better. Without knowing the max ripple current its best just to throw more cap at it as the effective ripple rating will increase as the number of caps increases. I wouldn't suggest a single cap.

As for the inductor - I would guess that any value between 1 and 10uH would be fine. You could do say bi- or tri-filar winding of 14AWG or 12AWG (if you can wind it) on a large toroid - that amount of copper would easily handle 30A no problem with only minimal I2R losses.
 
I can get an AIRD-03-8R2K inductor. Its specs are listed in this **broken link removed**. Would that be suitable?

So, I would run the positive output of the DC/DC converter into the inductor, then from the output of the inductor run into the pos. input of the C60. I would keep let’s say two of the 4700 uF caps across the pos/neg inputs of the C60. Is this correct?

Do you think the resonant properties of this circuit pose a high risk of damage to either of my components? I’m just looking for an opinion, I won’t hold you responsible :)
 
I submitted the above post on Friday June 19th and it didn't show up until Monday the 22nd. This is just a "bump" post so that the thread isn't lost in time :).
 
There was no datasheet for an inductor included in that link - none that I could find, anyway.

The double pole of the filter occurs at around 600Hz. I would suspect that any switching that is happening is set over 20kHz. I don't see a problem with it.
 
Thanks for having a look. When I click the link I get a spec sheet PDF (I must have posted it wrong or something, sorry). Anyway, here is the info from the sheet:

Thru-Hole Drum Core Power Inductors, (Ferrite core with UL tube, Wire wound, High current, low DCR)
L (uH) +/-10%: 8.2
Rdc Max (ohm): 0.004
Idc Max (A): 42
Iac Max (A): 21
Operating temp: -40 - 125 degC
Lp measure at 1KHz 0.1 VRMS
IDC: Lp drops 10% typical from its initial value
IAC: DT of 40 degC temp. rise max

I think I will try to find out from Xantrex what the PWM switching frequency of the C60 is.
 
Be careful about using an inductor in this case. It's going to give a pretty big inductive kick when the charge controller stops drawing power, basically creating a boost converter. This could easily cause the input voltage of the charger to be exceeded.

I would drop the inductor and just use caps.
 
It would depend on the switching frequency, current drawn, and inductance of the coil. Let me do some simulations to see how bad it would be.

Do you know the switching frequency of the charger? It's probably <10Hz or so.
 
Using a 10uH inductor and two 4700uF caps, the max voltage spike is ~18.5V. 20uH makes it ~19.5V. The switching speed is too slow for it to ramp up, so that's good. Tried it up to 200hz without problems. (this is all based on 16V input)

Anything above 20uH and the circuit starts oscillating on it's own, which doesn't help things at all.

One thing that might help is using a ramp of capacitors. 1x4700uF, 1x47uF 1x.47uF. The smaller caps will have a quicker response and help with the low loads better.

I still think the inductor is not necessary in this case. I've never seen one used to filter DC power rails. Only caps.
 
Last edited:
Wow! Thanks for running the simulation. I definitely need to learn to use one of those programs some day (so much to learn, so little time...). I have asked Xantrex what the PWM frequency is of the C60 and if it’s constant or variable. I’m sure it will take some time to get a response. I think the ramp of capacitors is a good idea (and easy to implement). I would still like to try the inductor as well, but I will wait until I get more information.
 
I just realized I forgot to mention the simulations were based on a 16V input. FYI.
 
inductors are used on the input of switching regulators quite a bit. they keep the input rail quiet because it forces all the ac switching current to be supplied by the capacitors.
 
Yes, they are used on the mains as an EMI filter for the AC input, but I've never seen them after the rectifier. If you have specific examples (schematics) I'd love to see them so I can understand better.
 
Easy as opening your PC and looking at the core regulator. You will likely see an input inductor to the multiphase regulator - used to keep the switching noise off the 12V rail. Any rail that you don't want to see the noise from a switching regulator on can use an input LC filter. Just have to remember that when you do this ALL the ACrms current will be supplied by the local input capacitors. If you don't have the inductor, the distributed capacitance of the rail can usually provide some the ac currents which can help to lower the number of caps used on the input. Of course, putting all the local caps required for the rms current on the input - even if you don't have the inductor - would be a more conservative design approach.
 
Looking at a computer PSU now. I do see a toroid fitted between the PFC and the main buck switch. The only thing it could be used for is filtering, as it's not really attached to anything else to behave differently.

Of all the coils/transformers in this one, that's the only one that is not used for switching, well, besides the EMI filter on the AC input.

Thanks for pointing that out.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top