1
\$\begingroup\$

I'm very much an amateur tinkerer when it comes to electronics. My formal training amounts to the digital electronics requirements from my CS degree in the early 90s. I'm also a bit of a blinkenlights fan, having grown up in the 60s and 70s.

All of that said, I have some older (< 10MHz) computers that I'd like to create some "front panel" displays for. To this end, I'd like to "tap" the address lines and bring them out to LEDs. I'm concerned about (intuitively... not with any informed knowledge):

  • Feedback into the bus lines from the additional line driving a transistor that drives an LED each.

  • Drawing too much power through the address line in an effort to drive the LED.

My thoughts are to create a simple high impedence buffer that drives a transistor that ultimately powers each LED.

Am I way overthinking this, are my concerns valid, and am I missing anything important that I should be worried about but haven't thought of?

\$\endgroup\$
  • 4
    \$\begingroup\$ Addresses are very often transitory. Unless you can freeze the bus, you may be disappointed with what the LEDs display. \$\endgroup\$ – glen_geek Jun 14 at 14:20
  • \$\begingroup\$ I appreciate that. I was planning to tap both address and data lines, which should prove more interesting. @wossname has a nice suggestion on a sample and hold approach that might be useful. \$\endgroup\$ – David Hoelzer Jun 14 at 15:35
  • \$\begingroup\$ Might we inquire as to what breed of host computer you are experimenting with? \$\endgroup\$ – Wossname Jun 14 at 20:46
  • \$\begingroup\$ A variety of 6502 and 680X0 based systems. The first guinea pig will likely be a PET 8032 since it's the slowest. :) \$\endgroup\$ – David Hoelzer Jun 14 at 21:06
  • \$\begingroup\$ @DavidHoelzer, sounds like you have some interesting projects ahead of you. There are plenty of retro-computer enthusiasts around here who would like to help so keep asking questions if you get stuck. Good luck! :) \$\endgroup\$ – Wossname Jun 14 at 21:17
2
\$\begingroup\$

You could use a latch chip to "sample and hold" the data lines.

Something like the 74HC573 might do the trick. Connect your data lines to the inputs of the latch chip and connect your LEDs to the outputs. Then have a way to toggle the latch pin on the 74HC573 at a slow rate (perhaps 1 Hz would suit your purposes) to refresh the display. You can simply tie the ~OE pin ("Output Enable, active low") to 0V. You can probably find a slowly toggling signal in the system somewhere or you can generate one yourself with something like a 555 timer. The latch chip's datasheet will show a diagram of the pulse (duration and direction) that you need to send to the LE pin ("latch inputs, active high").

If you didn't have a way to sample and hold the values, then it is likely that the diode would simply appear dimply lit all the time, which would look unimpressive. Toggling the refresh of the display slowly allows each LED to light up brightly and gives the human eye time to see it clearly.

The 573's inputs have a high impedance and therefore will not adversely affect the bus lines you are interfacing with. Although you should be careful to keep your "tap" wires a short as possible to avoid interference.

Your concerns are valid, placing LEDs directly onto these lines would likely cause all kinds of problems and may potentially damage the device. This is because data / address lines are not intended to be power sources to high current devices like LEDs, they are intended to be voltage-only signalling.

Latches are one of the simplest chips to use and will make a fun project for you I'm sure. :) 573's often have a nice wide voltage range (2 to 6 V) which makes life easier when playing with older computers.

\$\endgroup\$
  • 1
    \$\begingroup\$ An HC part may not function if the bus is truly TTL compatible, or driven by older NMOS processors...the \$V_{OH}\$ of the bus may not be high enough. \$\endgroup\$ – Elliot Alderson Jun 14 at 15:01
  • 1
    \$\begingroup\$ Hmm.. Perhaps a simple timer circuit that's driven off of the system clock... Ten samples per second or fewer would likely be great. \$\endgroup\$ – David Hoelzer Jun 14 at 15:36
  • \$\begingroup\$ I have to investigate whether or not it's possible for me to halt the clock on this system.. I haven't yet looked to see if DRAM is used and, if it is, if the refresh is driven from the system clock or not. \$\endgroup\$ – David Hoelzer Jun 14 at 15:37
  • \$\begingroup\$ @ElliotAlderson, yes you're right. I suggested an HC but there are many alternatives, perhaps a 4000 series part might be better for some vintages of host computers. \$\endgroup\$ – Wossname Jun 14 at 20:41
  • \$\begingroup\$ Actually, I think an HCT part might be a good choice. \$\endgroup\$ – Elliot Alderson Jun 14 at 20:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.