Matronics Email Lists :: View topic - Next generation Crowbar OVM

Matronics Email Lists
Web Forum Interface to the Matronics Email Lists

Get Email Distribution Too! FAQ Search Memberlist Usergroups Register

Profile Log in to check your private messages Log in

Next generation Crowbar OVM

Matronics Email Lists Forum Index -> AeroElectric-List

View previous topic :: View next topic

Author Message

nuckolls.bob(at)aeroelect
Guest

Posted: Mon Aug 14, 2023 4:29 pm Post subject: Next generation Crowbar OVM

 	  Quote: 	
		  Looks promising. Do you have a schematic with part numbers? I'm not familiar with the zener diode symbol with a third pin (input) which appear to stabilize the 10V reference. 	


   That's an LM431 'adjustable zener' . . . a

   big player in many of my contemporary projects.



   I'm reluctant to publish values and part numbers

   until the concept is, at least, bench tested

   for conformance to design goals.



   The motivation for publishing the 'cartoon' was

   to plant seeds of an alternative idea based on

   a couple decades of experience, allegiance to

   the legacy design rules and just perhaps, a

   bit of irritation at decisions driven more

   by economics and bureaucratic intransigence

   than a quest for the best we know how to

   do.



   The two ic's are lm311 style comparators.

   You can probably dope out a workable

   constellation of values based on the

   stated design goals.



   While the FET may indeed be suited to the

   'crowbaring' of a circuit breaker, how

   do we insure that it maintains a saturated,

   on condition from the time it switches ON

   until the breaker opens?



   The gate voltage cannot fall below the minimum

   value for necessary conduction while the

   breaker is deciding to toss in the towel . . .

   probably 15 to 50 milliseconds.



   Thid critical performance question demands

   more prayer over the drawing board.



   Thank you so much for your interest . . .



 

   Bob . . .



                    //// 

                   (o o) 

    ===========o00o=(_)=o00o========= 

    < Go ahead, make my day . . .   > 

    < show me where I'm wrong.      > 

    =================================

  

    In the interest of creative evolution

    of the-best-we-know-how-to-do based

    on physics and good practice.

Eric Page

Joined: 15 Feb 2017
Posts: 245

Posted: Mon Aug 14, 2023 9:11 pm Post subject: Re: Next generation Crowbar OVM

Thanks for including all of the history on your OVP efforts in your first reply, Bob.  It's fascinating to hear how these things developed over time, and to see how the regulatory and corporate bureaucracies stand in the way of progress.



Quoting from your replies in my thread and this one, in no particular order...



 	  Quote: 	
		  DO-160/Mil-STD-704 dictates even back then called for a certifiable appliance to withstand a 20V excursion for 1 Second; a 40V excursion for 100 milliseconds. 	


So, if I understand this correctly, and read between the lines a little bit, the ~3mS delay in your legacy OVP circuit...



http://www.aeroelectric.com/DIY/DIY_Crowbar_OVP_F.pdf



...is shorter than ideal, and the relevant standards allow a much longer response time.  If we adopt a longer delay, we get a system that's significantly more tolerant of bus transients but still protects our avionics when things truly go sideways.  Does that sound right?



 	  Quote: 	
		  While the FET may indeed be suited to the 'crowbarring' of a circuit breaker, how do we insure that it maintains a saturated, on condition from the time it switches ON until the breaker opens?



The gate voltage cannot fall below the minimum value for necessary conduction while the breaker is deciding to toss in the towel... probably 15 to 50 milliseconds. 	


Yeah, good point.  When the FET turns on, the supply is pulled HARD toward 0V, which kills the FET's gate drive.



Perhaps the following might work (see "Crowbar FET.pdf" below): Isolate the pullup resistor from the rest of the 10V circuit using a diode with low reverse leakage (i.e. BAS33: 1nA(at)15V), use a modestly low value pullup resistor (1k?) and a logic-level FET so the minimum Rds-on occurs at a low drive voltage.



Something like the IRLZ44 (TO-220 through-hole) or BUK962 (TO-263-3 surface mount) might work.  Both are limited to a Vgs of +/-10V, but that's protected by the forward voltage of the isolating diode.  I couldn't find any high current, logic level FETs with higher Vgs.



FET datasheets:

IRLZ44: https://www.vishay.com/docs/91328/irlz44.pdf

BUK962: https://assets.nexperia.com/documents/data-sheet/BUK962R5-60E.pdf



 	  Quote: 	
		  You can probably dope out a workable constellation of values based on the stated design goals. 	


Indeed, I took a stab at it with an eye toward using it as a relay driver for a PM stator (see "2-Comp OVP Relay Driver.pdf" below).

- Used an LM393 (vs the LM311, the LM393 has two comparators in one package and it's cheaper).

- P-FET on the output to drive relay coil(s).

- N-FET to latch timer comparator's reference input to ground.

- Separate reference divider for timer comparator so latching FET can't pull 431's REF pin to ground.

- 7.5V Zener on both comparators to protect non-inverting inputs from exceeding Vsupply minus 2V (LM393's input voltage range).

- N-FET to ensure timing capacitor is fully discharged during reset so comparator can't restart in latched state.

- Timing: 10V charging 10uF through 10k should take ~69mS to reach the 5V reference.

What do you think?  Is a 69mS delay OK, or would you aim for something shorter or longer?



-Eric

Crowbar FET.pdf

Description:

Download

Filename:

Crowbar FET.pdf

Filesize:

7.7 KB

Downloaded:

83 Time(s)

2-Comp OVP Relay Driver.pdf

Description:

Download

Filename:

2-Comp OVP Relay Driver.pdf

Filesize:

21.97 KB

Downloaded:

77 Time(s)

finn.lassen(at)verizon.ne
Guest

Posted: Tue Aug 15, 2023 12:49 pm Post subject: Next generation Crowbar OVM

On 8/14/2023 8:28 PM, Robert L.       Nuckolls, III wrote:

     

      	  Quote: 	
		                 	  Quote: 	
		  Looks           promising. Do           you have a schematic with part numbers? I'm not familiar with           the zener           diode symbol with a third pin (input) which appear to           stabilize the 10V           reference. 	
       

       � That's an LM431 'adjustable zener' . . . a

         � big player in many of my contemporary projects.

         

         � I'm reluctant to publish values and part numbers

         � until the concept is, at least, bench tested

         � for conformance to design goals.

         

         � The motivation for publishing the 'cartoon' was

         � to plant seeds of an alternative idea based on

         � a couple decades of experience, allegiance to

         � the legacy design rules and just perhaps, a

         � bit of irritation at decisions driven more

         � by economics and bureaucratic intransigence

         � than a quest for the best we know how to

         � do.

         

         � The two ic's are lm311 style comparators.

         � You can probably dope out a workable

         � constellation of values based on the

         � stated design goals.

         

         � While the FET may indeed be suited to the

         � 'crowbaring' of a circuit breaker, how

         � do we insure that it maintains a saturated,

         � on condition from the time it switches ON

         � until the breaker opens?

         

         � The gate voltage cannot fall below the minimum

         � value for necessary conduction while the

         � breaker is deciding to toss in the towel . . .

         � probably 15 to 50 milliseconds.

         

         � Thid critical performance question demands

         � more prayer over the drawing board.

         

         � Thank you so much for your interest . . .

         

                

           � Bob . . .

           

           .

                     	
     

Thanks Bob, that's a handy device if one doesn't happen to have a       zener on hand with desired voltage, even if it doee require two       resistors.     

I'd forgotten that LM311's can have open collector outputs (when       tying output transistor emitter to ground). Still have several in       my old parts bins but more than four decades since last used one.

          

I'm thinking the SCR is still the better choice for a crowbar       circuit. But if we're changing the OVP circuit to a relay       disconnect between alternator and battery/bus a FET or other       device that can drive the relay may be OK. Assuming the OVP       monitors alternator voltage and not bus voltage.     

Finn

nuckolls.bob(at)aeroelect
Guest

Posted: Tue Aug 15, 2023 4:16 pm Post subject: Next generation Crowbar OVM

 	  Quote: 	
		  So, if I understand this correctly, and read between the lines a little bit, the ~3mS delay in your legacy OVP circuit...



  http://www.aeroelectric.com/DIY/DIY_Crowbar_OVP_F.pdf 	


    Good eye my friend! That's a typo. Just to be sure, I ran the

    traps on circuit values and came up with approx 60mS to trip

    for a more realistic rate of rise for a faulted alternator

    driving a flightworthy battery. Still pretty fast having

    sorta caved to legacy certified designs. Gonna quit that

    and go for any event exceeding 16.0+/-0.5Volts AND 500+/-0.50Seconds; well

    inside design goals cited by DO160/MIL-Std-704.



    I'll go fix that posting . . . or maybe just take

    it down. The very first CBOVM-14s I delivered

    exploited a relatively precise 10.V trigger

    diode that were 'calibrated' as-assembled.



    First the diode went obsolete and now attractive

    SCRs are more difficult to find. Time for the 

    BROAD BRUSH upgrade.



 

  	  Quote: 	
		  ...is shorter than ideal, and the relevant standards allow a much longer response time.  If we adopt a longer delay, we get a system that's significantly more tolerant of bus transients but still protects our avionics when things truly go sideways.  Does that sound right?



 

 > While the FET may indeed be suited to the 'crowbarring' of a circuit breaker, how do we insure that it maintains a saturated, on condition from the time it switches ON until the breaker opens?

 > 

 > The gate voltage cannot fall below the minimum value for necessary conduction while the breaker is deciding to toss in the towel... probably 15 to 50 milliseconds.



 Yeah, good point.  When the FET turns on, the supply is pulled HARD toward 0V, which kills the FET's gate drive. 	


    Also good eye. Vcc for the electronics will need to be

    diode isolated from the effects of the crowbar pull down.

    The capacitor across Vcc will be large enough to hold

    the crowbar action ACTIVE long enough to trip the breaker.



    I'll have to check again but I think the circuit demands on

    Vcc will be on the order of 1mA. Keeping Vcc at or above

    5v for 200mS or so shouldn't be hard.



  	  Quote: 	
		  Perhaps the following might work (see "Crowbar FET.pdf" below): Isolate the pullup resistor from the rest of the 10V circuit using a diode with low reverse leakage (i.e. BAS33: 1nA(at)15V), use a modestly low value pullup resistor (1k?) and a logic-level FET so the minimum Rds-on occurs at a low drive voltage.



 Something like the IRLZ44 (TO-220 through-hole) or BUK962 (TO-263-3 surface mount) might work.  Both are limited to a Vgs of +/-10V, but that's protected by the forward voltage of the isolating diode.  I couldn't find any high current, logic level FETs with higher Vgs.



 FET datasheets:

 IRLZ44:  https://www.vishay.com/docs/91328/irlz44.pdf

 BUK962:  https://assets.nexperia.com/documents/data-sheet/BUK962R5-60E.pdf

 

 

 > You can probably dope out a workable constellation of values based on the stated design goals.



 Indeed, I took a stab at it with an eye toward using it as a relay driver for a PM stator (see "2-Comp OVP Relay Driver.pdf" below).

 - Used an LM393 (vs the LM311, the LM393 has two comparators in one package and it's cheaper).

 - P-FET on the output to drive relay coil(s).

 - N-FET to latch timer comparator's reference input to ground.

 - Separate reference divider for timer comparator so latching FET can't pull 431's REF pin to ground.

 - 7.5V Zener on both comparators to protect non-inverting inputs from exceeding Vsupply minus 2V (LM393's input voltage range).

 - N-FET to ensure timing capacitor is fully discharged during reset so comparator can't restart in latched state.

 - Timing: 10V charging 10uF through 10k should take ~69mS to reach the 5V reference.

 What do you think?  Is a 69mS delay OK, or would you aim for something shorter or longer? 	


    I'd go for the whole 500 MSec enchilada.



    Yeah, need to resolve the issues with trip

    dynamics to make sure the FET stays on

    for at least 200mS to open the breaker.



    I've toyed with the notion of a 3-wire

    configuration where Vcc is derived from

    say the battery bus (A draw of about 1mA

    . . . insignificant) leaving only the

    crowbar breaker on the main bus. 



    At the moment, I'm rather enamored of the

    NEC NP110N03PUG FET. This little guy will

    grunt over 200A at low temperatures with

    a Vgs of only 5V.



    Let's continue to continue . . .



 

   Bob . . .



                    //// 

                   (o o) 

    ===========o00o=(_)=o00o========= 

    < Go ahead, make my day . . .   > 

    < show me where I'm wrong.      > 

    =================================

  

    In the interest of creative evolution

    of the-best-we-know-how-to-do based

    on physics and good practice.

NP110N03PUG_Transfer_Characteristics.jpg

Description:

Filesize:

200.11 KB

Viewed:

1514 Time(s)

Eric Page

Joined: 15 Feb 2017
Posts: 245

Posted: Wed Aug 16, 2023 8:05 am Post subject: Re: Next generation Crowbar OVM

 	  nuckolls.bob(at)aeroelect wrote: 	
		  ...go for any event exceeding 16.0+/-0.5Volts AND 500+/-0.50Seconds; well inside design goals cited by DO160/MIL-Std-704.

[...]

I'd go for the whole 500 MSec enchilada. 	


Copy all; thanks.  I'll update my schematic accordingly.  Hitting that target will be a lot easier than it was to hit exactly 3mS!



 	  Quote: 	
		  At the moment, I'm rather enamored of the NEC NP110N03PUG FET. This little guy will grunt over 200A at low temperatures with a Vgs of only 5V. 	


Looks good, but...  does the 30V max Vds present a risk of avalanche breakdown during a short-duration (i.e. not long enough to trip the timer) OV transient?  The BUK962 that I mentioned above has a max Vds of 60V and the transfer characteristic chart looks even better.



But...  depending upon what the crowbar pulse looks like, it might fall outside the bounds of the Safe Operating Area chart for either of those FETs.



What is it that this task demands that isn't met by an available SCR?

nuckolls.bob(at)aeroelect
Guest

Posted: Wed Aug 16, 2023 2:06 pm Post subject: Next generation Crowbar OVM

 	  Quote: 	
		  Thanks Bob, that's a handy device if one doesn't happen to have a zener on hand with desired voltage, even if it doee require two resistors.



 I'd forgotten that LM311's can have open collector outputs (when tying output transistor emitter to ground). Still have several in my old parts bins but more than four decades since last used one.



 I'm thinking the SCR is still the better choice for a crowbar circuit. But if we're changing the OVP circuit to a relay disconnect between alternator and battery/bus a FET or other device that can drive the relay may be OK. Assuming the OVP monitors alternator voltage and not bus voltage. 	


  Those two voltages are generally the same as

  long as we're talking about the main bus.



  Yeah, the SCR configuration has some really

  desirable features, namely it's an AC device

  that inherently 'latches' in DC service. Don't

  need to worry about gate drive latency.



  So for the moment, I'm trying to stuff a more

  modern NFET into that slot. But if I could find

  a suitable SCR not likely to disappear in the

  next few years . . . yeah, let us lament the

  evolution of selenium rectifiers.



  The 'crowbar' approach that opens the control

  power source yields a kind of universal configuration

  in the spirit and intent of the original CBOVM-14

  products. Yet it may well be that 'progress'

  in component evolution makes that a

  goal-too-far.



  Good input sir!



 

 

   Bob . . .



                    //// 

                   (o o) 

    ===========o00o=(_)=o00o========= 

    < Go ahead, make my day . . .   > 

    < show me where I'm wrong.      > 

    =================================

  

    In the interest of creative evolution

    of the-best-we-know-how-to-do based

    on physics and good practice.

nuckolls.bob(at)aeroelect
Guest

Posted: Thu Aug 17, 2023 11:12 am Post subject: Next generation Crowbar OVM

 	  Quote: 	
		    	  Quote: 	
		  Thanks Bob, that's a handy device if one doesn't happen to have a zener on hand with desired voltage, even if it doee require two resistors. 	
	


   best yet, they're really cheap and

   0.5% tolerance. Vcc is generally

   tight enough to serve both as power

   and reference.



 

   Bob . . .



                    //// 

                   (o o) 

    ===========o00o=(_)=o00o========= 

    < Go ahead, make my day . . .   > 

    < show me where I'm wrong.      > 

    =================================

  

    In the interest of creative evolution

    of the-best-we-know-how-to-do based

    on physics and good practice.

nuckolls.bob(at)aeroelect
Guest

Posted: Thu Aug 24, 2023 1:10 pm Post subject: Next generation Crowbar OVM

 	  Quote: 	
		    	  Quote: 	
		  > While the FET may indeed be suited to the 'crowbarring' of a circuit breaker, how do we insure that it maintains a saturated, on condition from the time it switches ON until the breaker opens?

 > 

 > The gate voltage cannot fall below the minimum value for necessary conduction while the breaker is deciding to toss in the towel... probably 15 to 50 milliseconds.



 Yeah, good point.  When the FET turns on, the supply is pulled HARD toward 0V, which kills the FET's gate drive. 	


    Also good eye. Vcc for the electronics will need to be

    diode isolated from the effects of the crowbar pull down.

    The capacitor across Vcc will be large enough to hold

    the crowbar action ACTIVE long enough to trip the breaker. 	


   Your comment prompted further inquiry into suitable

   replacements for the SCR I had been using. Don't

   know what I entered 'wrong' in the search parameters

   but another look produced this little gem:



  https://www.ween-semi.com/sites/default/files/2022-02/TYN30-600TF.pdf 



   Very sensitive gate current (10 mA). Specifically

   cited for 'crowbar service'. Price is right too.

   I've got some on order.



   This illustrates the value of collaborative dialog.

   I might well have peddled down the FET pathway

   with success as yet to be demonstrated. Your

   comment prompted another look into the 'devil-

   we-know' technology with what appears to be

   a superior solution that's more compatible with

   the very low-power, comparator-timer configuration.

   Don't know how I missed it the first time at-bat.



   I'm headed out with the grandma and grandson for

   Oregon to get my little brother married off . . .

   and to let grandson stick his toes in the Pacific.



   We'll be pretty much out of pocket until Sept 6

   although I can peek in from time to time through

   the web browser portal.



   In the mean time, I'll have plenty of opportunity

   to conduct some asphalt engineering.  I'll knock

   out an ECB to see how well we can make this

   fourth generation CBOVM sing, dance and do

   dishes.



   Thanks for your input! 



 



 

 

   Bob . . .



                    //// 

                   (o o) 

    ===========o00o=(_)=o00o========= 

    < Go ahead, make my day . . .   > 

    < show me where I'm wrong.      > 

    =================================

  

    In the interest of creative evolution

    of the-best-we-know-how-to-do based

    on physics and good practice.

finn.lassen(at)verizon.ne
Guest

Posted: Thu Aug 24, 2023 1:43 pm Post subject: Next generation Crowbar OVM

On 8/24/2023 5:10 PM, Robert L.       Nuckolls, III wrote:

     

      	  Quote: 	
		                 	  Quote: 	
		            	  Quote: 	
		  >             While the FET             may indeed be suited to the 'crowbarring' of a circuit             breaker, how do we             insure that it maintains a saturated, on condition from the             time it             switches ON until the breaker opens?

             > 

             > The gate voltage cannot fall below the minimum value             for necessary             conduction while the breaker is deciding to toss in the             towel... probably             15 to 50 milliseconds.

             

             Yeah, good point.� When the FET turns on, the supply is             pulled HARD             toward 0V, which kills the FET's gate drive. 	
         

           �� Also good eye. Vcc for the electronics will need to be

           �� diode isolated from the effects of the crowbar pull           down.

           �� The capacitor across Vcc will be large enough to hold

           �� the crowbar action ACTIVE long enough to trip the           breaker. 	
       

       � Your comment prompted further inquiry into suitable

       � replacements for the SCR I had been using. Don't

       � know what I entered 'wrong' in the search parameters

       � but another look produced this little gem:

       

        https://www.ween-semi.com/sites/default/files/2022-02/TYN30-600TF.pdf       

       

       � Very sensitive gate current (10 mA). Specifically

       � cited for 'crowbar service'. Price is right too.

       � I've got some on order.

       

       � This illustrates the value of collaborative dialog.

       � I might well have peddled down the FET pathway

       � with success as yet to be demonstrated. Your

       � comment prompted another look into the 'devil-

       � we-know' technology with what appears to be

       � a superior solution that's more compatible with

       � the very low-power, comparator-timer configuration.

       � Don't know how I missed it the first time at-bat.

       

       � I'm headed out with the grandma and grandson for

       � Oregon to get my little brother married off . . .

       � and to let grandson stick his toes in the Pacific.

       

       � We'll be pretty much out of pocket until Sept 6

       � although I can peek in from time to time through

       � the web browser portal.

       

       � In the mean time, I'll have plenty of opportunity

       � to conduct some asphalt engineering.� I'll knock

       � out an ECB to see how well we can make this

       � fourth generation CBOVM sing, dance and do

       � dishes.

       

       � Thanks for your input! 

       

       

       

       

                

           � Bob . . .

             

                  	
     I just ordered a couple of these:from DigiKey:

      SCT640B

     1655-2116-ND

     SCR 600V 40A TO220B     

$2.27 each, more than twice the TYN30-600TF 

          

The TYN30-600TF has a lower trigger voltage (1V) compared to the       SCT640B (1.5V) but 35mA gate current but same as the S4025 (1.55V       and 35mA) we're trying to replace.     

BTW, I was able to order the MBS4991 on eBay, so still possible       to make the original OVP in small quantities.     

Finn

nuckolls.bob(at)aeroelect
Guest

Posted: Thu Aug 24, 2023 3:42 pm Post subject: Next generation Crowbar OVM

 	  Quote: 	
		  

 1655-2116-ND

 SCR 600V 40A TO220B 



 $2.27 each, more than twice the TYN30-600TF 



 The TYN30-600TF has a lower trigger voltage (1V) compared to the SCT640B (1.5V) but 35mA gate current but same as the S4025 (1.55V and 35mA) we're trying to replace.



 BTW, I was able to order the MBS4991 on eBay, so still possible to make the original OVP in small quantities.



 Finn 	




   Let us know how your experiment works out!



 



 

   Bob . . .



                    //// 

                   (o o) 

    ===========o00o=(_)=o00o========= 

    < Go ahead, make my day . . .   > 

    < show me where I'm wrong.      > 

    =================================

  

    In the interest of creative evolution

    of the-best-we-know-how-to-do based

    on physics and good practice.

Display posts from previous:

	Matronics Email Lists Forum Index -> AeroElectric-List	All times are GMT - 8 Hours
Page 1 of 1

Jump to:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum