How do I eliminate noise in an 8085 circuit?

Thread Starter

keiichicom

Joined Apr 26, 2022
76
Hi,

I am still relatively new to microprocessor circuits. Please tell me how I can reduce or eliminate the noise introduced into the circuit below?

I have a half-decent power supply with a stable output voltage(input about 12 volts, output about 5 volts):

power board.jpg

5volts.jpg


But there is noise in the circuit as you can see in the scope screen pix below.

This noise ends up getting amplified at other pins when I try to make other circuits.

8085 Free Run Circuit

Pulling down the data bus outputs, and pulling up the Ready signal, forces the CPU to execute NOP commands indefinitely:

8085freerun.png

Waveform at VCC

ps_noise.jpg



Waveform at A15 (msb address output) where LED is attached, note the noise at the top of the pulses:
sig_noise.jpg



Thank You
 

nsaspook

Joined Aug 27, 2009
13,435
Those are reflection rings. It's something you will likely always see on a breadboard with fast edges on digital pulses and long random traces to components without a proper ground plane to constrain signal energy.
 

DickCappels

Joined Aug 21, 2008
10,232
Step #1: move your project from that terrible plastic breadboard.

You probably need a ground plane as @nsaspook mentioned and some good low impedance traces for the power input, power decoupling, and signal return.

The best time to reduce interference is during PCB layout when you can visualize signal paths and ground loops.

Once the board is laid out, and the boards are made, if you see interference you have a nice platform upon which to figure out where the noise is coming from and how it is getting into the “victim” circuit.

De-noising on one of those plastic bread boards is, in my book, a futile time wasting effort.
 

MrChips

Joined Oct 2, 2009
30,983
What you see at the top of the waveforms are very likely artefacts produced by your breadboard, wiring, probe and oscilloscope.

Hence they are induced effects caused by your measuring equipment themselves.

In the digital world, you have to learn what is real and needs correcting and what you can ignore.
 

Thread Starter

keiichicom

Joined Apr 26, 2022
76
Those are reflection rings. It's something you will likely always see on a breadboard with fast edges on digital pulses and long random traces to components without a proper ground plane to constrain signal energy.
Wow this forum is so great, thanks for the quick reply. Understood, willdo.
 

Thread Starter

keiichicom

Joined Apr 26, 2022
76
Step #1: move your project from that terrible plastic breadboard.

You probably need a ground plane as @nsaspook mentioned and some good low impedance traces for the power input, power decoupling, and signal return.

The best time to reduce interference is during PCB layout when you can visualize signal paths and ground loops.

Once the board is laid out, and the boards are made, if you see interference you have a nice platform upon which to figure out where the noise is coming from and how it is getting into the “victim” circuit.

De-noising on one of those plastic bread boards is, in my book, a futile time wasting effort.
Thanks, I will get a pcb made. I have prototyping board with soldered wires but pcb will be better. That is exactly my issue, not sure where the noise is coming from. I first noticed the noise in another circuit measuring the sod pin(trying to bitbang data) in another circuit...
 

Thread Starter

keiichicom

Joined Apr 26, 2022
76
What you see at the top of the waveforms are very likely artefacts produced by your breadboard, wiring, probe and oscilloscope.

Hence they are induced effects caused by your measuring equipment themselves.

In the digital world, you have to learn what is real and needs correcting and what you can ignore.
Thanks for the tip. It may be the case, i another circuit it had a real effect and my circuit didn't work.
 

Thread Starter

keiichicom

Joined Apr 26, 2022
76
Very unlikely.
What is the frequency of the quartz crystal?
I suspect your circuit did not work for a different reason.
It turns out you were right, my clock showed that it worked: a 6 MHz crystal and 3MHz on clock output pin, but there was also a weird high frequency signal on the SOD and reset out pins that varied in the 5-6 MHz range which should not be there at all while running the free-run circuit. I switched chips to 8085ah-2 and put a 12 MHz crystal(it should be 10Mhz but I am still looking for it) and the problem went away. I think that the clock signal earlier somehow affected the above mentioned pins.

I am not too concerned about noise mentioned earlier, I will take the advice of others and put the whole thing on PCB.
I will be bit-banging the SOD pin for serial communications.

Thanks.
 
Last edited:

MrChips

Joined Oct 2, 2009
30,983
Quartz crystals (XTAL) are very sensitive to PCB layouts.
Here are the basic rules:

1) Keep all traces between the XTAL and MCU as straight and short as possible.
2) XTAL circuits may require two loading capacitors (about 22pF) and a 10MΩ resistor across the XTAL. Follow the recommendations provided in the MCU application notes.
3) The two external loading capacitors need to terminate on a solid ground plane.

You might have one of the two situations:

1672164574717.png
 

Thread Starter

keiichicom

Joined Apr 26, 2022
76
Quartz crystals (XTAL) are very sensitive to PCB layouts.
Here are the basic rules:

1) Keep all traces between the XTAL and MCU as straight and short as possible.
2) XTAL circuits may require two loading capacitors (about 22pF) and a 10MΩ resistor across the XTAL. Follow the recommendations provided in the MCU application notes.
3) The two external loading capacitors need to terminate on a solid ground plane.

You might have one of the two situations:

View attachment 284003
Thanks, I will try the circuit on the right and follow your other advice.

On another note, I also ensured that I had a proper working reset circuit and pull up / pull down resistors on the various required pins... reference: Michael Schenk's Mini 8085 Circuit

p.s. I had only seen the app note about 20pf capacitor required between pin 2 and ground for <= 4MHz crystals...
 
Last edited:

Thread Starter

keiichicom

Joined Apr 26, 2022
76
Thanks, I will try the circuit on the right and follow your other advice.

On another note, I also ensured that I had a proper working reset circuit and pull up / pull down resistors on the various required pins... reference: Michael Schenk's Mini 8085 Circuit

p.s. I had only seen the app note about 20pf capacitor required between pin 2 and ground for <= 4MHz crystals...
It worked! my clock is now stabilized, even while using the 6Mhz crystal and I now see 6Mhz oscillations on both crystal pins, thanks.
 
Top