Latching LEDs question

Discussion in 'The Projects Forum' started by Otaku, Dec 1, 2010.

  1. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    Hi all,

    I'm designing a test circuit that sequentially checks 12 TC junctions for correct polarity. I'm using a 555 oscillator and a pair of cascaded 4017's to sequence the tester. The tester looks at just one junction at a time.
    BTW, thanks to all who helped me get two 4017's to cascade and reset correctly. That part of this project is working great.

    As each TC junction is heated, the mV output is sent to an amplifier and then to a LM339 quad comparator. When that amplified voltage exceeds the LM339's reference voltage (2V), the output goes to a 4081 AND gate along with the output from the 4017 (this verifies the specific junction under test). I can use the output (5VDC) from the LM339 to light a LED. So far, so good.

    Here's where I'm stuck: As each junction test finishes, I need to keep the LED lit until all the junctions have been tested so that the assembler can see that the test sequence is complete.

    Obviously, what I need to have is an IC that, when it receives a high output signal from the comparator, switches on the LED (either directly or via a transistor) and holds it in that state until the operator resets the tester. I've done a fair bit of research for the device I need, and it looks like a 4042 quad latch may be the answer. I've tried to find a circuit example that uses the 4042 to perform the LED latching function, but haven't found anything yet.

    I don't have a schematic at this time (still learning to use ExpressSCH). I've searched the forums here for a thread that addresses this specific issue but haven't found one as yet. I really just need to know what to do once I have the 5VDC output from the LM339. Can anyone point me in the right direction? As always, all help is much appreciated!
     
  2. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    A problem with the 4042 is the common clock for all stages. Another approach would use low power SCR like 2N5060 as latch & LED driver, one set for each output. This assumes there is a 4081 gate for each stage. Opening V supply resets all SCRs.
     
  3. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    Thanks for the heads-up on the 4042, Bernard. Yes, there is a 4081 gate for each stage. I plan to have a momentary switch on the test fixture that let's the user reset the LED latching device(s) after each test sequence is complete.
    I checked the datasheet for the 2N5060 SCR and see that it calls out a max gate voltage of 0.8V and a max gate current of 200 uA. The 4081 gate outputs will be ~12VDC, 1A (maximum output of the regulated power supply I'm using). Will I blow the SCR's with these values? Should I use current-limiting resistors on the 4081 outputs?
     
  4. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
  5. JDT

    Well-Known Member

    Feb 12, 2009
    658
    85
    Or do it in logic with an S-R latch. Look at 74HC279. Or cross-coupled gates. Or the S and R inputs of a 4013. CMOS logic will drive a small LED directly.
     
  6. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    Partial schematic of LED driver-latch as used in a multi zone alarm system from 1970s.
     
  7. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    Way cool. Thanks, all. The schematic is very informative; it's pretty much what I'm trying to do.
     
  8. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    Don't forget to put a simple switch (or transistor) to break the current to ground. SCRs have no reset, the current path must be broken.
     
  9. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    My plan is to use a NC momentary switch to reset the SCR's. I'll have those in hand on Monday.
    The manufacturing guys are jumping on the feature-creep wagon. They're asking about having some kind of an audible alarm in the event of a TC failure. Time for more homework!
     
  10. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    That could be done somewhat simply without a redesign. Use couple 8 input NAND gates (12 inputs from SCR outputs, the rest tied to last step of counter) to sound a piezo buzzer at the end of the test and before reset if one or more LEDs do not have a "HIGH" signal.
     
  11. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Having dealt with many factory worker types, I'm going to potentially/predict additions to your feature-izing of this.

    How will the DUT (Device Under Test) be connected to this test fixture? if it is a long bar clamp that connects all the wires at nearly the same time, the bar could have another contact that could be used to start the test, and opening the clamp could be used for "reset", so there aren't extra buttons to worry about, possibly speeding the process up.

    On the topic of fast testing, adding a 2nd output option to each comparator for a "Bad TC" to alert (Red LED/Buzzer) before the end of testing. This is compared to the NAND solution I mentioned above, which would only be enabled by the final clocks from the 4017 to do set off the buzzer. The NAND gate would light up RED if testing failed, or all of the LEDs would be green showing they are functioning. "See Red, it's dead" kind of thought.

    I guess it depends on the frequency of scanning, what time is needed for the system to test all 12 TCs? If the first one fails, and they are all one unit, it could stop testing there and sound the alarm.

    The problem with this is no matter how much you idiot proof it, somebody comes up with a better idiot.
     
  12. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    True, idiot-proof is a nice ideal but tough to pull off LOL.
    The DUT (a radio-frequency tissue ablation catheter) will be inserted into a socket that holds the TC heaters and has a set of conductive pads that contact the electrodes (NiTi needles). Four of the twelve TC's are embedded in the NiTi tubes for monitoring the needle temps. A Patient Interface cable connects the device to the test circuit I'm developing. The LEDs are red/green bi-colors and the red dies will always be lit. When a successful test occurs, the SCR will light the green die for that particular TC. The green dies on these LEDs completely overpower the reds so there's no mistaking a good test. If the test of a particular TC fails, the green die won't light. After the test is finished, a reset button will shut off the SCRs and all LEDs will again be red. Right now, it looks as though the test sequence will take about 30 secs to sequentially heat TC's 1 - 12 to the point where the voltage output exceeds the threshold in the comparators.
     
  13. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    Do you still need an audible alarm at time of failed test?
     
  14. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    Yeah, kind of looks that way. What Prod. wants is an audible alarm at the time of the first failed test. The remaining tests, if any, should run out to see if there are any other failures (as indicated by a red LED).
     
  15. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    You could go for obnoxious, like a continuous alarm, or a simple beep. If you go for obnoxious I'd put a push button or switch for alarm off, your fixture will last longer that way. Voice of experience.
     
  16. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    Obnoxious is what my boss has in mind, just short of making the box dance off the table. There is a problem with this, though. It will take ~3 secs (est) to heat the TC to the point where the voltage signal passes the comparator's threshold (2VDC). If I use an AND gate (on the comparator output) to trip the alarm, the alarm device will activate when the tester begins heating the TC under test and then shut off if the TC passes. So you'd have this thing throwing out a ~2 second beep at the start of each test. That might not be a bad thing, unless it's a very loud beep, and then it just becomes annoying.
     
  17. Wendy

    Moderator

    Mar 24, 2008
    20,765
    2,536
    Your boss isn't going to beat it with a ballpin hammer to get it to shut up. I'd put a button to turn it off, if only for a minute or three.
     
  18. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    I do not have it down in black n white yet, but what i had in mind was to re-shape the 555 clock so output would be high for say 3 sec with a 20 ms reset time. If test fails after 2980 ms, the 4017 pulse, 20 ms pulse & not failed test[ AND inverted] are inputted to 3 input NAND gate, inverted & , output fed to big 12 input OR or NOR & inverted hopefully resulting in a + pulse out for a failed test. Use another SCR for memory & signal driver.
     
  19. Otaku

    Thread Starter Active Member

    Nov 19, 2008
    128
    2
    I planned to have the reset button turn off the alarm at the same time it clears the latched green LEDs. The bi-color LEDs would revert to red at that point. The boss only uses the hammer on me LOL.
    Bernard, interesting concept. The 555 is currently designed for an adjustable high (1 sec to 10 sec) with a fixed low time of ~500 msec. This design can be changed as needed, of course.
    The estimated heating time for the TC test is still just a guess. I'm also designing the holding fixture, and the embedded TC junctions will be subject to a bit of heat sinking from contacting the needle guide slots. I'm not certain what that effect will be just yet.
     
  20. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    How big are the leads? I was thinking they were something like 16 gauge wires, but the description makes me think they are smaller. Would a 28 or 40 pin ZIF socket work for the test clamp? Move lever to open, insert, move lever to close..


    --ETA: THIS WAS MY TWO THOUSANDTH POST!
     
    Last edited: Dec 6, 2010
Loading...