• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 20 subscribers
  • Views 583 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

What can I do to an HC86 to stop the glitches?

Stephen Tomporowski
Prodigy 230 points
Other Parts Discussed in Thread: SN74LV86A

I am dealing with an old design, from the '90's, which uses an HC86 with feedback to square up a sine wave.  Back in the day, the Motorola chip they specified worked fine, but even then it had to be Motorola.  Now with process improvements and shift of production all  over, we end up with a significant number of HC86 chips that spit out a large amount of spikes.  In this case, better sensitivity causes malfunctions.

The basic circuit is a tuned RLC (inductive sensor) feeding through a capacitor to one of the inputs of the HC86 with a 301k feedback resistor from output to input.  The other input is tied to +5V.  We're running the whole system at 5V.  The tuned RLC is fed by a 1 Mhz clock signal (from one of the other gates of the HC86).  As the inductive sensor varies with distance, the signal to the HC86 varies with phase and amplitude, but we are only doing this to extract the phase information. 

Right now I have 3 systems that have multiple spikes on the trailing edge.  To add to the misery, the number of spikes varies with the sensor displacement.  Sometimes we get lucky and the spikes only arrive outside the range of interest. 

So, it seems that I need to desensitized and slow down my HC86.

Luckily this product will be obsoleted soon, but not soon enough.

Any ideas will be appreciated!

Thanks....Steve

  • 0
    TI__Mastermind 46480 points

    Hi Steve

     It sounds like you sould be using a schmitt trigger input device.

    First, be sure the part has a bipass capacitor right in the Vcc pin.  0.1uf to 2uf should work.

    Another thing you could do is switch to the SN74LV86A. These are very quiet parts and usually clean up noisy systems.

     

     

  • 0 in reply to Chris Cockrill
    Prodigy 230 points

    Yes, the power supply is bypassed.

    I'll check out the SN74LV86A and see what it does.  Is there such a beast as a Schmitt-trigger XOR?  Time to search.

    Thanks much....Steve

  • 0 in reply to Stephen Tomporowski
    TI__Mastermind 46480 points

    unfortunately I dont know of a schmitt trigger Xor.

  • 0 in reply to Chris Cockrill
    Prodigy 230 points

    Thanks for the info.  We have the LV version on order to arrive tomorrow.  I'll post what happens.  The big problem is that this is an old product about to go obsolete, probably within the next 6 months to a year.  I would love to revamp the design.  About 4 or 5 years ago we did come up with 3 variations, but we had no support.  In the original design, they locked down the HC86 to be a Motorola only.  I would have thought, or at least I would have been suspicious of a design that would work with only one mfg's parts.  But, as I said, I inherited this beast and the designers are long gone.  On a different product, another inherited design, I got to talk to one of the designers.  When I asked him why he was using a 100khz rated switch at 500khz, he replied "Well, it worked."

    Thanks for all the help!

  • 0 in reply to Stephen Tomporowski
    Expert 1495 points

    I'm not sure, but it seems that LT6994-2 might help solve your problem. Not a TI part, though..

     

  • 0 in reply to Jayant Deshpande
    Prodigy 230 points

    The LT6994 looks like a pretty neat device.  Unfortunately, I'd need a lot less than 1µs minimum clock.  The pulse itself is roughly 200nsec and the ringing adds about 80 nsec to the pulse and pulse rep rate is 500khz.  Thanks for mentioning it.  I could have use for this in other places. 

    Thanks...Steve

  • 0 in reply to Stephen Tomporowski
    Prodigy 230 points

    Just wanted to mention that we did dig up a few old Motorola 74HC86A's in the stockroom to get these three units going.  Because each type of sensor is different and has different  tuning, the type of problem we encounter here is not normal.  It just pops up once in a while and usually on older products like this one.

    74LV86's are due in today.  I'm excited.  If it works, I may convince sales to keep life in this product.

  • 0 in reply to Stephen Tomporowski
    Prodigy 230 points

    Okay, after having a water main break on Friday (I told my boss I had to go home because I had to use the bathroom), I finally replaced the On Semi HC86 with the TI LV86.

    And this is what happened.....

    It work perfectly!  The performance is almost identical to the old Motorola HC86.  The only difference is that the LV part has a faster rise/fall time.  We do have one glitch that enters the pulse, however, that is way outside the measuring range.

    Many thanks to Chris Cockrill and Jack Bourke for their recommendation!