Possible Noise on power rail

oracacle

Senior Member
I have been playing with the a nextion display. For this I have been re-imaging an older project which has an audio detection circuit. Original project:- http://www.picaxeforum.co.uk/showthread.php?26768-High-speed-photography
All works fine until the display is dimmed, at which point the signal from the audio detector goes high. I cant actually detect any issues with noise on the power line.
If the display is disconnected from power while dimmed the audio detector returns to normal. if brightness is set to 100, again thing return to normal.
If I have a separate power supply for the audio detector things work normally (wall wart for main circuit, batteries for audio with shared 0v). This is what led me to believe there is some noise on the power line.
There is no change if the display is left powered on and disconnected from the H-serial pins with the analogue inputs connected to ADC 5 and 6 as the project makes use of comparator interrupt.
The circuit is currently being prototyped on an axe091 using a 20x2. I have tried adding an 100nf and 100uf capacitor across the power rails next to where the nextion meets the board which had no affect on the noise causing the issue.
I was hoping to keep component count as low as possible as a result was going to run the system on 3 AA batteries with no regulator.

Although the system will function just fine without the display dimming it wold be better to do so as its intended to be used in near darkness and light emitted from the screen may well affect the final image. The original just hide the display (send 254,8) and then restoring is when needed (254,12).

This is not something that I have had to deal with in the past so am in new territory and any help would be appreciated.

sound detection circuit: http://www.picaxe.com/docs/picaxe_sound.pdf
 

inglewoodpete

Senior Member
It is likely that the sound detection circuit is sensitive to RF interference. There are ways of combatting this (Eg ferrite bead/s metal shielding) but it will vary according to your particular situation.

You're on a steep learning curve. To help you get an idea of the problem, place an AM radio near the circuit while the interference is being generated.
 

oracacle

Senior Member
I don't even own a radio, let alone an MA radio.
I didn't even think of RF interference.
I was going to stack the board that contained the audio and picaxe circuit under the display with the batteries below that.
As it currently stands the audio circuit is about 20cm away for the display.

Even with RF in mind I would have thought that the same RF would still be an issue even when running from separate power supplies.
 

techElder

Well-known member
The backlight on many of these devices is a fairly high voltage. I have a similar display in a programmable thermostat for my HVAC unit.

When the display dims, I can hear the power supply "humming", but when it is brightest I hear nothing.

Perhaps the Nextion has a similar system?

Does the "noise" problem react the same way if you adjust the sensitivity of the audio amp?

I don't see much in the way of filtering in that audio amp circuit. Remember that "high frequency noise" can be rectified in that circuit just as easily as audio frequencies.

If your circuit works without being hard wired to the display board, then you might assume "noise" is coming over the lines. I would certainly add more filtering in the audio amp circuit (even on the digital lines connecting to the PICAXE.)
 

oracacle

Senior Member
this is the output at the Q2 side of the diode with the brightness set to 5%.
When set to around 95% I get a strong square pulse of 170us every 3.3ms.
the output remain the same until nearly zero and the output drops to zero. Changing set zero has not effect until output becomes 0
I was thinking about putting a high frequency band pass filter between RV1 and C3.

Edit: I get a dip in voltage every 3.3ms on the junction of RV1, C2 and Q1
 

Attachments

Last edited:

oracacle

Senior Member
I decided yo put a 0.2 non inductive resistor between ground and the display and reading taken from between the 2. I found that set to 100 brightness it draws about 115mA. this fluctuates when the display is dimmed. there is voltage swing of about 45mV (or 225mA) between ground and peak draw.
In the attached image trace 2 (in green) is voltage reading from the 0.2ohm resistor. Trace 1 (in yellow) is from the junction of C2, RV1 and Q1. the dip is about 0.940-1V and coincides with the fluctuation of power from the screen.
 

Attachments

techElder

Well-known member
So, 3.3mS is about 300 Hz. That probably is the display backlight power supply.

If you have "long" ground lead between audio amp and AXE091, then you could be developing a small voltage across that lead through the power lead caused by pulse current from the backlight power supply. In your last image, that's what I was thinking when I saw the broad green trace.

I know you've stated that you put a 100uF cap on the power at the display, but go ahead and swamp it out with more capacitance just to see if you didn't go far enough. If the problem is from the power supply switching current, then you'll have to supply that surge current from somewhere. Better from the capacitor than from the main power through "all that wiring" with resistance.

However, also remember that C1 in your audio amp is essentially a high pass filter. Better to keep "stuff" out before it gets amplified by the circuit.
 

oracacle

Senior Member
I built a copy onto the bread board so ty the word of short.

I did another little test and measured the current that the audio pulled, I found that here is the a matching dip. The current bring used was tiny, which got ne to thing about just delivering a contact current supply, it only needs about 20mA. I will give bigger capacitors later.

I will get some more form numbers too
 

oracacle

Senior Member
I have just had a quick try and found the 2000uf electronic and 470nf ceramic (I have some 1000uf ones here) goes some way to fixing things. I haven't put the scope on it yet, but it seems like a lot of capacitance and quite bulky as I may will need more to fix the issue completely.
 

techElder

Well-known member
I suppose I was suggesting a way to isolate and point to the problem. Perhaps there's a better solution.
 

oracacle

Senior Member
A solution has been presented, final components still need to be found with some testing. But a 330r between power supply and power rail for audio circuit with a 470n capacitor to ground on each of the adc sensors solved the problem. The value of 220r and 47uf were suggested but didn't have any, I will get some for testing
 
Top