LC voltage spike app. note

Technical discussions about electronics, sensors, circuit design, and testing.

LC voltage spike app. note

Postby jan » Mon Jun 30, 2008 6:40 pm

Hello,

We just posted a new article about LC-induced, power-up voltage spikes. In our tests, applying 12 V from a bench power supply to a ceramic capacitor caused 40 V spikes, which could easily destroy products with 15 V or 24 V maximums.

Please share any feedback or related experiences in this thread.

- Jan
User avatar
jan
Pololu Employee
 
Posts: 2176
Joined: Sun Dec 28, 2003 7:12 pm
Location: Las Vegas, NV

Re: LC voltage spike app. note

Postby Mike » Fri Jul 18, 2008 11:48 am

Hi Jan,

Great article!
I have burned 2 pololu dual serial motor controllers! I have 8 more and don't wan't to burn another one!!!!

My power supply is a couple of 12v 700mAh batteries and I have 2 DC motors of 12v 100mA (600mA max).

Should I add some resistors to limit LC spikes? Where? Between batteries and pololu power input? Between pololu power output and motors?

Regards,
Mike
Mike
 
Posts: 5
Joined: Fri Jun 27, 2008 12:46 am

Re: LC voltage spike app. note

Postby jan » Fri Jul 18, 2008 11:56 am

Mike,

I don't think your problem is related to the LC spike discussed in the article. The dual serial motor controller doesn't have any SMT ceramic caps on it, and it can generally handle voltages quite a bit higher than 12V. Since your motor driver chip is burning out, I think that is just an indication that the load is too big. Given the experience I've had with the motor drivers, I suspect that the current is actually higher than you think it is. How do you know what the current is?

- Jan
User avatar
jan
Pololu Employee
 
Posts: 2176
Joined: Sun Dec 28, 2003 7:12 pm
Location: Las Vegas, NV

Re: LC voltage spike app. note

Postby dkgeo » Sun Mar 27, 2011 3:52 am

Hello Jan !

I am planning to unbrick a seagate hard drive and my search through relevant formus led me to a guide offering a solution using the Pololu 23201a Serial adapter and thus your site.

I did read your work on destructive LC voltage spike. Quite enligthing I must say and congratulations on your will to offer this informartion to the public.

I would like to ask for your kind advise/guidance on the following issues: (I am a Mechanical Engineer, I have no practical experience on electronics but as I had taken a few general courses on electrical engineering I hope I will be able to try to follow any reply)

1) with reference to your article page 17 of 18, some experiment is required to determine the size of the external capacitor to be used. I understand that some voltmeters can measure capacitance also. Considering that I am a total begginer, if I buy such a voltmeter, is there a practical way to determine/estimate the PSU (with wires) LC figures in order to select the size of the external capacitor? I can also open the cover of the PSU to try to read its components particulars, if this helps, at least for capacitors, for inductance I do not sure at the moment how to read them.

2) I have the option to power the 23201a adapter either with 3.3V or 5V from the PSU . Apart from the volume of the capacitor for which I have no problem, Can I use as much larger capacitor I can find? any negative implication?

3)In case I use a battery (3.3V or 4.8V) and select a much larger capacitor (2 or 3 times then the 330uF) do you see any potential negative implication on this? (is it possible the battery energy be drained quiclky by the capacitor?)

The PSU I intent to use has the following particulars:
Power output : 130W (max 230W)
Length of wires about 50cm each made of 20AWG (a few 22AWG also)

I am sorry for this long post, I just try to take into consideration all available parameters before implementing the procedure.

Thank you very much for your time!! Best regards!!
dkgeo
 
Posts: 2
Joined: Sun Mar 27, 2011 3:05 am

Re: LC voltage spike app. note

Postby jan » Mon Mar 28, 2011 12:13 pm

I don't recommend that you dig around in your power supply. It's also mostly irrelevant to this LC problem, which comes from the inductance of your wires and the capacitance on your target board. Anything from a hundred uF to a few thousand uF should be fine for either of your scenarios; just make sure to put the cap close to your target board.

- Jan
User avatar
jan
Pololu Employee
 
Posts: 2176
Joined: Sun Dec 28, 2003 7:12 pm
Location: Las Vegas, NV

Re: LC voltage spike app. note

Postby dkgeo » Tue Mar 29, 2011 10:16 am

Ok. All noted and I will act accordingly. Thank you very much for the support Jan !!
dkgeo
 
Posts: 2
Joined: Sun Mar 27, 2011 3:05 am

Re: LC voltage spike app. note

Postby Zoom » Tue Apr 03, 2012 5:32 am

That is a really good tutorial on a little-understood but important topic.

Thank you,
Colin
Zoom
 
Posts: 44
Joined: Wed Feb 29, 2012 7:16 am

Re: LC voltage spike app. note

Postby tsteele93 » Thu Jun 07, 2012 5:48 pm

I have a question about how to configure the oscilloscope to capture the startup transients. How do you do that? I don't know how to see only the first moments of operation and not continue to overwrite it with new info... sorry for the basic question.
tsteele93
 
Posts: 2
Joined: Thu Jun 07, 2012 5:46 pm

Re: LC voltage spike app. note

Postby ryantm » Fri Jun 08, 2012 11:25 am

Hello, tsteele93.

If you have a digital storage oscilloscope, which most modern ones are, you just have to use the single-shot or single-sweep trigger mode. You should look at the documentation for your particular oscilloscope to see how to set that up.

- Ryan
User avatar
ryantm
Pololu Employee
 
Posts: 1267
Joined: Mon May 19, 2008 1:00 pm
Location: Las Vegas, NV

Re: LC voltage spike app. note

Postby tsteele93 » Mon Jun 11, 2012 11:51 am

Thank you Ryan, I found it in the trigger menu - single. Worked great!
tsteele93
 
Posts: 2
Joined: Thu Jun 07, 2012 5:46 pm

Re: LC voltage spike app. note

Postby razvan » Thu Mar 14, 2013 11:32 am

I got into a problem by following the advice of adding a capacitor (electrolytic 44microF) between the ground and VIN of a
Pololu Adjustable Step-Up/Step-Down Voltage Regulator S8V3A
to protect against LC spikes.

The voltage regulator was powered from a 9V battery connected through a relay (Normally Opened):
Meder SIL05-1A72-71L
The relay switch support up to 200V at 0.5A (20W) so I thought will be no problem.
The relay coil was connected on and off to 5V.

But after some tests (with load or not on the output of the regulator S8V3A, doesn't make any difference) the relay switch began to hung in 'Closed' position.
When relay coil wasn't powered the relay switch didn't opened anymore, it remained closed.

After some readings on the internet I learned that capacitor can cause high inrush current from the battery through the relay switch.
It will protect the regulator ok, but the inrush current still exist between the battery and the capacitor thus affecting the relay switch.
If I remove the capacitor or put a resistor in series with the relay (100ohm) the relay switch do not hung anymore in closed position when the relay coil isn't powered.

So in this case the capacitor maybe solve one problem (protect the regulator) but create another (damage the relay).
Attachments
schematic.jpg
schematic.jpg (13.74 KiB) Viewed 4804 times
razvan
 
Posts: 4
Joined: Thu Mar 14, 2013 11:10 am

Re: LC voltage spike app. note

Postby jan » Thu Mar 14, 2013 4:42 pm

Hi.

It looks like what you used was a reed relay, so I suspect that rather than "creating" another problem, what you did with the capacitor exposed another problem: that relay was never going to be a reliable part for switching power. Can you confirm that it's a reed relay, and if so, why were you using one for power switching? (I think of them as for signal switching, but I have not ever used them.)

- Jan
User avatar
jan
Pololu Employee
 
Posts: 2176
Joined: Sun Dec 28, 2003 7:12 pm
Location: Las Vegas, NV

Re: LC voltage spike app. note

Postby Jim Remington » Thu Mar 14, 2013 5:30 pm

With a relay, you cannot multiply the voltage and current ratings together to come up with some sort of "safe amount of power", as you did. In fact, you are not allowed to exceed either the voltage or current rating, at any value for the other parameter.

You are probably correct in that the inrush current exceeded the rated 0.5 amps, which is very damaging to the contacts. As Jan suggests, that relay was never intended to be a power switch. In general, relays are very unreliable for switching DC power.
Jim Remington
 
Posts: 514
Joined: Mon Aug 21, 2006 11:32 pm
Location: Eugene,OR

Re: LC voltage spike app. note

Postby razvan » Fri Mar 15, 2013 1:38 am

yes, it's a reed relay.

the power rating of the relay was specified in its datasheet, I assumed I'll never exceed either the voltage or the current.
The consumer connected at the output of the pololu regulator will never draw more then 20mA, but I got the same situation even without any consumer connected.

Any suggestion what should I use for this power switching?
In the complete circuit the relay coil receive 5V/10mA from a BC337 when a low current is received in the transistor base.
Should I eliminate completely the relay/diode + battery and connect the pololu regulator GND to BC337 collector and pololu regulator VIN to VIN +5V?
In this case do I still need the capacitor between GND and VIN at the pololu?

Thanks,
Razvan
Attachments
tx.rx - Copy.jpg
tx.rx - Copy.jpg (24.88 KiB) Viewed 4764 times
razvan
 
Posts: 4
Joined: Thu Mar 14, 2013 11:10 am

Re: LC voltage spike app. note

Postby Jim Remington » Fri Mar 15, 2013 8:12 am

There are many ways to switch power by electronics, and a relay is generally a bad choice. It would help if you tell us what you want to do and specified your circuit requirements. Are you aiming for the lowest possible power consumption while the rest of the circuit idles? Does part of the circuit have to be electrically isolated from another part?
Jim Remington
 
Posts: 514
Joined: Mon Aug 21, 2006 11:32 pm
Location: Eugene,OR

Next

Return to Electronics

Who is online

Users browsing this forum: No registered users and 1 guest