[SI-LIST] Re: Latching relays in optical switches

  • From: Don Nelson <dhwn@xxxxxxx>
  • To: steve weir <weirsi@xxxxxxxxxx>
  • Date: Tue, 19 May 2009 21:52:13 -0400

Thank you all very much for your comments and suggestions.  I'm currently 
spending some quality time in the lab characterizing the circuit; I'll be back 
soon with what I found.  

Steve: By "protection diodes", do you mean flyback diodes across the coils?  If 
so, the existing circuit does not have them.  The relay driver (an MDC3105) has 
a 6.6V Zener across the collector/emitter for its protection.  The relay two 
coils appear to be wound on the same core, in opposing directions (we ripped 
one apart).  The bistable armature that actuates the optical relay lays 
parallel to the core.  As expected, the applied voltage across the SET coil 
(measured at a rate of 0-to-5V in <10 ns!)  results in opposing spikes on the 
(open) RESET coil of ~2V, below the breakdown of the protection zener in the 
RESET coil driver.  So, unfortunately, it doesn't appear (at first blush) that 
the magnetic flux is being canceled.  I haven't gotten far enough into 
characterizing this yet to know if the 2V observed on the un-driven RESET coil 
is reasonable.

Thanks again,
-don
Netronome Systems
--
Don Nelson

"The whole problem with the world is that fools and fanatics are so sure of 
themselves, and wiser people so full of doubt" --Bertrand Russell

 
On Tuesday, May 19, 2009, at 04:12PM, "steve weir" <weirsi@xxxxxxxxxx> wrote:
>There is a good chance that the problem is the protection diodes for 
>your drive transistors are defeating the magnetic circuit.  This thing 
>probably works as a center tapped transformer.  When one side switches, 
>the other side wants to swing the same voltage about the common terminal 
>voltage, but your protection diodes limit the swing to about a volt.
>
>Using low side drivers, add a diode in parallel to your low side driver 
>to protect against negative spikes when the other side turns off.  For 
>local flyback, connect the cathodes of the flyback diodes to 10V zener 
>in parallel with an RC snubber. Now when one side switches low the other 
>will be able to jump up to 10V without canceling flux in the magnetic 
>circuit.
>
>Before you do all this, you can connect the device to a signal generator 
>to determine the coupling between coils.
>
>The odd comment recommending 5V probably comes from a drive circuit that 
>happened to work, and now 5V has become a witchcraft belief.
>
>Steve.
>
>Don Nelson wrote:
>> Hello,
>>
>>   I know that this question strays from SI, but it involves EM, and that's 
>> something that we in SI seem to know pretty well.  Well, I *thought* I knew 
>> it pretty well, but this has me stumped.  I was brought in to help solve a 
>> problem with an existing product.
>>
>>   We are using an optical switch that utilizes a dual-coil latching relay.  
>> The relay will not reliably switch from one bistable state to the next at 
>> relatively high temperatures (still well within the environmental 
>> specification of the relay).  The company that makes the relay believes that 
>> the problem lies in the method we are using to drive it:
>>
>> We tie one leg of each coil to 5V and the other leg to a relay driver, which 
>> consists of an NPN transistor with an integrated diode to absorb the 
>> back-EMF created when the coil is de-energized.  This transistor switches 
>> the low-side of the coil to GND.
>>
>> The manufacturer recommends the opposite method: tying one leg of each coil 
>> to GND and switching the high side.  They contend that having the 5V 
>> potential on the coil permanently is "interfering with the magnetic field".  
>> I find this hard to believe, but physics was a long time ago.  
>>
>> My working hypothesis is that the coils are being energized and de-energized 
>> too quickly.  Because the relay coils share a core, the dI/dt in the coil 
>> being energized is inducing a current in the opposite coil: I can see a 
>> significant voltage spike generated across the opposite coil when I energize 
>> and de-energize the coil I intend to.  I am concerned that this spike in the 
>> opposite coil is preventing the relay from switching states by opposing the 
>> mechanical force.  So, I am reducing the edge rate at the base of the drive 
>> transistor to lower the dI/dt of the coil.  The spikes on the opposite coil 
>> are now reduced significantly.  I have not, however, gotten permission to 
>> test this modification on our only board that exhibits the problem reliably. 
>>  I, justifiably, need to make my case first...
>>
>> The problem is, the manufacturer disagrees and insists our circuit needs to 
>> be redesigned to permanently tie one leg of the coils to GND and switch the 
>> high side instead.  They will not tell me why, and cannot explain the 
>> physics behind this recommendation.  Since the coils have no reference, I 
>> don't understand why they would care what potential is on either leg--I 
>> thought that only the magnitude and direction of current through them was 
>> relevant.  I did check to see if the 5V rail was moving during the switching 
>> on and off of the coils, but it is stable.
>>
>> I am continuing to perform experiments in an attempt to isolate the root 
>> cause, but I am curious if anyone might have another hypothesis--in 
>> particular, why a coil might care if one leg was permanently tied high while 
>> the low side is switched?  Even if I accidentally try something that seems 
>> to fix the problem in the lab, I don't feel that I am truly understanding 
>> the root cause and am uncomfortable proposing a solution until I DO 
>> understand it.
>>
>> Thank you all kindly in advance for your assistance,
>>
>> regards,
>> Don Nelson
>> Netronome Systems
>> --
>> Don Nelson
>>
>> "The whole problem with the world is that fools and fanatics are so sure of 
>> themselves, and wiser people so full of doubt" --Bertrand Russell
>>
>> ------------------------------------------------------------------
>> To unsubscribe from si-list:
>> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
>>
>> or to administer your membership from a web page, go to:
>> //www.freelists.org/webpage/si-list
>>
>> For help:
>> si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
>>
>>
>> List technical documents are available at:
>>                 http://www.si-list.net
>>
>> List archives are viewable at:     
>>              //www.freelists.org/archives/si-list
>> or at our remote archives:
>>              http://groups.yahoo.com/group/si-list/messages
>> Old (prior to June 6, 2001) list archives are viewable at:
>>              http://www.qsl.net/wb6tpu
>>   
>>
>>
>>   
>
>
>-- 
>Steve Weir
>Teraspeed Consulting Group LLC 
>121 North River Drive 
>Narragansett, RI 02882 
>
>California office
>(866) 675-4630 Business
>(707) 780-1951 Fax
>
>Main office
>(401) 284-1827 Business 
>(401) 284-1840 Fax 
>
>Oregon office
>(503) 430-1065 Business
>(503) 430-1285 Fax
>
>http://www.teraspeed.com
>This e-mail contains proprietary and confidential intellectual property of 
>Teraspeed Consulting Group LLC
>------------------------------------------------------------------------------------------------------
>Teraspeed(R) is the registered service mark of Teraspeed Consulting Group LLC
>
>
>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


List technical documents are available at:
                http://www.si-list.net

List archives are viewable at:     
                //www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu
  

Other related posts: