Data lines from memory chip look weird

Discussion in 'General Electronics Chat' started by Robin Mitchell, Aug 12, 2011.

  1. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    hi everyone,

    I am getting a strange reading from my oscilliscope when i probe the data pins on my memory chip. The address lines are connected to the address lines of a z80A CPU, the data lines of the chip dont connect to anything. WE is held high (as im using it as a ROM), CE is connected to a 74139 decoder and the OE is connected to the z80's RD.

    I checked all the control lines + address lines, all perfect square waves. I check the add lines on the mem chip, same result. I probe the data lines and i should see random square waves (because of the different data) but instead i get a continuous wave that has an odd shape.

    The wave has half a square wave but instead of falling to 0 quickly it decays till its 0 and then goes back up, so think of half a square wave and then a capacitor exponent decay at the end.

    http://mientki.ruhosting.nl/data_www/pic/jalss/rcx_motor_display2.gif

    looks just like the second one :<

    I forgot to put by pass capacitors so i added them, same result :(
    I made sure no pins where floating, same result :(
    If i slow the clock down from the usual ~900 KH down to HZ it works and the data lines display what they should be doing.

    I checked the CLK line from the 4093 and its a clean square wave. Power lines have no fluctuation. I tried different memory chips and different z80 but the problem still persist
     
  2. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    The data lines are probably going tri-state or high impedance, thus they tend to "float" to some value. Noting to worry about, it actually means all is well.
     
  3. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    so when i connect them and it happens again (like last time) to the data pins on the z80, what does that mean?
     
  4. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    As Bob Pease stated (quoting Milligan's Law): "When you are taking data, if you see something funny, Record Amount of Funny."

    What does "it" mean? The funny floating? It's normal. Address and control lines don't tri-state, hence look like perfect square waves. If you zoom into the low frequency "perfect" data lines you will see the exact same slope as they do at high frequency.

    Addendum:When a memory device is not being addressed it will remove itself from the data bus so that another device may connect there.
     
  5. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    Ok so i went a connected the data lines, its working as expected but the data still shows the same signal with the curvy decay. Im not sure why this is happening but it is not affecting the operation :S ???
     
  6. tom66

    Senior Member

    May 9, 2009
    2,613
    213
    Try using a 1k - 10k resistor to ground on each data line.
     
  7. Robin Mitchell

    Thread Starter Well-Known Member

    Oct 25, 2009
    732
    199
    im assuming that stops the data lines from ever floating and help remove the curve?
     
  8. tom66

    Senior Member

    May 9, 2009
    2,613
    213
    It can help discharge the pin's capacitance.
     
Loading...