Chinese 7400 and 4000 Series Parts - Do they hate feedback?

Discussion in 'General Electronics Chat' started by Robin Mitchell, Feb 18, 2016.

  1. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    734
    200
    Dear all,

    So some of you here will be familiar that I have made posts about Chinese imported parts that did not function correctly when used in a specific arrangement.

    I also have two other parts : 74HC14 and 4093
    Now I stock these for the sole purpose of their oscillation applications. They have Schmitt trigger inputs which means that they are perfect for RC oscillators (and that they are inverting so the output is at 180 degrees to the input, a requirement for oscillation).

    Now, I built my CPU card and fitted one of these Chinese parts and the oscillator did not oscillate. Instead it sat around 2.5V (with a 5V supply). But what was more worrying was how hot the chip got....then it exploded! Surprise surprise :(

    Now if anyone here looked at the parts they would spot that they are not official parts in a second!

    "But Robin, its probably your circuit configuration. Im sure your at fault that these parts are fine". Well here is the schematic for the two oscillators:
    [​IMG] [​IMG]

    And with the CPU card I replaced the 74HC14 part with one of my older ones (which is an official part). And guess what happened... The LEDs blinked, the IORQ line on the Z80 was clocking away and the oscillator...was working :)

    So I have a theory and need some guys here to help me with this (those who have experience or knowledge on semi-conductors). I reckon that what is happening is that either some kind of input protection diode is failing or that the internal "MOS" transistors (if they are :/ ), are sitting around the linear region which causes them to explode. Maybe there is poor silicon design so latch up is really easy to get. I also believe this because when the power supply is reduced below 3V feedback circuits have a better chance of working such as the 4013 toggle layout. Maybe the lower voltage is not high enough to cause latch up and ruin the device.

    Any thoughts? Anyone here ever seen a chinese silicon piece under a scope?
     
  2. dl324

    Distinguished Member

    Mar 30, 2015
    3,250
    626
    My suspicion is that they relabeled a regular inverter for the HC14 and the RC time constant is keeping the inverter in it's linear region long enough to destroy it. You can conduct threshold voltage tests to confirm or deny.
     
    Robin Mitchell likes this.
  3. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,365
    A resistor from output to input of a CMOS inverter constitutes negative feedback. This puts the inverter into the linear region at 2.5V for 5V supply. The circuit becomes a linear inverting amplifier. This is a well known phenomenon and is exploited to make simple linear amplifiers with digital components.

    In your case with your experience with the questionable 74HC14, two things appear to be evident.

    1) The Schmitt trigger function is lacking, allowing the part to be stable in the linear region.

    2) The totem pole output stage has been fabricated to allow both high-side and low-side drivers to be turned on hard. Excessive current is drawn. The chip gets hot and blows.
     
    Robin Mitchell and nsaspook like this.
  4. dannyf

    Well-Known Member

    Sep 13, 2015
    1,828
    365
    I have never seen this and it is not supposed to happen. The only thing I can think of is that you may have used a non St input part. It will oscillate at much higher frequency.

    One possibility, incredible remote if I may add, is that the upper and lower output devices are somehow conducting at the same time. Tough to envision how that could be the case.

    Without saying some measurement and in life, hard tobaay for sure.
     
    Robin Mitchell likes this.
  5. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,348
    Hello,

    Also what value did you use for the resistors?

    Bertus
     
  6. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    734
    200
    @bertus

    10K ohm resistor and a 100pF capacitor
     
  7. dl324

    Distinguished Member

    Mar 30, 2015
    3,250
    626
    That time constant shouldn't cause any harm. I've used similar values for power on reset using CMOS inverters.

    EDIT: I take that back. If it is a normal inverter, it's being self biased to VCC/2 and will cause excessive power dissipation.
     
    Last edited: Feb 18, 2016
  8. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,348
    Hello,

    The 10K is a to low value for cmos.
    In the following picture taken from the TI datasheet a value of 50 K to 1 M is given:

    4093_oscillator.png

    Bertus
     
  9. GopherT

    AAC Fanatic!

    Nov 23, 2012
    6,061
    3,834
    Your circuit is fine for TTL but the electronics site below says you need an extra resistor for CMOS at higher frequencies. Explanation of sorts is in the first paragraph.

    http://www.electronics-tutorials.ws/waveforms/generators.html

    Whereas the NXP datasheet for 74hc14 and 74HTC14 don't require it (Way below)

    image.jpg

    image.jpg
     
    Last edited: Feb 19, 2016
  10. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    734
    200
    Too all those who say the value is too low : It is not too low for the 74 chip and for the 4093 the chip would at least oscillate and I do have a working 4093 chip with a 10K resistor that works fine. Its only the cheap chips that fail, they don't even oscillate!
     
  11. dannyf

    Well-Known Member

    Sep 13, 2015
    1,828
    365
    I can confirm that: I have used ones in the sub-1K range, without any problem.

    I think you seem to be on a crusade against "cheap" chips. I don't think you have established that those chips are fake, and I think you have provide little evidence to support that they don't work in this set-up.
     
  12. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    734
    200
    @dannyf sounds like you want pictures and video. I will see what I can do!
     
  13. dannyf

    Well-Known Member

    Sep 13, 2015
    1,828
    365
    I don't want anything. I just thought for anyone to claim something, the burden is on the claimant to support his/her claims. You have made your claim here but I don't think you have sufficiently support it with variable facts.
     
  14. dannyf

    Well-Known Member

    Sep 13, 2015
    1,828
    365
    I tried a few chips I have: 1k feedback resistor + 4.7u cap. 5v supply. three chips, a knock-off HC00, a Fairchild HC00, and a TI AHC132. They all oscillated happily. Current consumption is 5.4ma, 12.7ma, and 7.1ma, not in that order.

    Which do you think is a knock-off?

    :)

    I think I have some CD4k series somewhere and will try when I find them.
     
    absf and GopherT like this.
Loading...