Author Topic: 62.5kHz ringing Noise on signal  (Read 3189 times)

0 Members and 1 Guest are viewing this topic.

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
62.5kHz ringing Noise on signal
« on: May 07, 2019, 10:08:47 pm »
Hi,

The project that I am working on is to transmit video stream through visible light. The schematic of receiver circuit is shown in attachment.

Received signal which is transmitted through visible light is converted into voltage with help of OPA657 then I clamped the signal to ground level and finally buffered.

This schematic works in breadboard design.

However, In PCB design there is always ringing on signal which disturbs the signals color.

PCB layout is also in attachment.

62.5kHz ringing on signal is shown in DS1Z_QuickPrint135.png.

Is there any suggestions?

Thank you in advance.
 

Offline stj

  • Super Contributor
  • ***
  • Posts: 2195
  • Country: gb
Re: 62.5kHz ringing Noise on signal
« Reply #1 on: May 07, 2019, 10:30:19 pm »
interesting, the frequency falls into the range of most modern switching psu chips.
have you checked the enviroment for "contamination" ?
 

Offline Andy Watson

  • Super Contributor
  • ***
  • Posts: 2138
Re: 62.5kHz ringing Noise on signal
« Reply #2 on: May 07, 2019, 10:32:26 pm »
The ringing has a repetition rate of 62.5KHz but the frequency of the actual ringing looks to be much higher. Those high-bandwidth op-amps are very sensitive to capacitance around their inputs - you've got some "long" traces runing parallel to the ground plane. "Visible light" - does that mean that one of those inputs is connected to a photodide ? - Remember to factor the capacitance of the diode into your design.
Also, how are you probking the signal? Your oscilloscope probe will add a few pf to the circuit - it's often useful to isolate the probe by placing a 100 resistor at the probe tip. And use the short stubby probe ground clip too.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: 62.5kHz ringing Noise on signal
« Reply #3 on: May 07, 2019, 11:35:07 pm »
Are you sure it's not coming from the 555 -9V generator, although its frequency should be about 48kHz with those values. Its 3200uF cap seems a very large, probably too large to suppress fast switching spikes.

I've just realised that 15,625*4 = 62,500 which might help to give a clue, or not, to where it's coming from.
« Last Edit: May 08, 2019, 01:55:18 am by StillTrying »
.  That took much longer than I thought it would.
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1048
  • Country: ca
Re: 62.5kHz ringing Noise on signal
« Reply #4 on: May 08, 2019, 02:35:37 am »
Isn't 15,625 Hz a horizontal scan rate for a television operating on 50 Hz?  Could there be a CRT or LCD display close by generating the 62.5 kHz as a multiple of 15,625 Hz?  It soulds like the circuit may be sensitive to stray light of some sort.

« Last Edit: May 09, 2019, 05:06:51 pm by duak »
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 62.5kHz ringing Noise on signal
« Reply #5 on: May 08, 2019, 03:07:05 am »
The supposed "ringing" is two or three pixels wide. How are we supposed to be able to judge what the signal is?

Also, 62.5kHz appears to be the repeat rate, not the ringing frequency.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: 62.5kHz ringing Noise on signal
« Reply #6 on: May 08, 2019, 03:11:29 am »
The layout makes me think that there is excessive parasitic capacitance between the inverting input and ground lowering the phase margin of the operational amplifier causing it to oscillate.  This is a problem if the ground plane and flood fill is not cut away around the inverting input.

The impedance at the inverting input in parallel with the parasitic capacitance to ground creates another pole inside the feedback loop which lowers phase margin.  One solution to this is to add a compensating capacitance in parallel with the feedback network.
« Last Edit: May 08, 2019, 03:15:30 am by David Hess »
 

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 7854
  • Country: au
Re: 62.5kHz ringing Noise on signal
« Reply #7 on: May 08, 2019, 07:06:11 am »
Hi,

The project that I am working on is to transmit video stream through visible light. The schematic of receiver circuit is shown in attachment.

Received signal which is transmitted through visible light is converted into voltage with help of OPA657 then I clamped the signal to ground level and finally buffered.

This schematic works in breadboard design.

However, In PCB design there is always ringing on signal which disturbs the signals color.

PCB layout is also in attachment.

62.5kHz ringing on signal is shown in DS1Z_QuickPrint135.png.

Is there any suggestions?

Thank you in advance.

The bursts of ringing are synchronous with the video signal, so they are related to that signal.
The repetition frequency of the bursts is 4 times 15.625 kHz (the PAL or SECAM line rate).

From your schematic, it appears you are using Composite Syncs to switch your DG419 to ground to act as a keyed clamper.
That is not normal practice, which is to derive a clamp pulse from your line sync pulse (and from the equalising pulses & serrations  in the field group.)

Clamp pulses are of a standard duration, (somewhat shorter than that of a sync pulse) & also usually delayed so they fall during the back porch of the signal requiring clamping.(sometimes the clamping is performed during the sync pulse duration)

Using the whole composite sync waveform will make the clamping fairly erratic.
It could be that this causes various spurious signals to appear.

If you use the burst/ back porch output you will have a closer approximation to how it is done in Broadcast TV, although the pulse duration will be a bit long.

If you do use this signal, you will need to put a 4.433MHz trap in series with your connection to pin (1) of the DG419, so you don't mangle the burst.

For fault finding I would suggest you remove the connection to the DG419.
You should now have a composite video signal which has lost its DC reference.

If it still has the interference, disconnect the connection to the LM1881.
Again, if that device was the cause, you will have a composite video signal without its DC reference.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: 62.5kHz ringing Noise on signal
« Reply #8 on: May 08, 2019, 03:12:56 pm »
Hi,

The project that I am working on is to transmit video stream through visible light. The schematic of receiver circuit is shown in attachment.

Received signal which is transmitted through visible light is converted into voltage with help of OPA657 then I clamped the signal to ground level and finally buffered.

This schematic works in breadboard design.

However, In PCB design there is always ringing on signal which disturbs the signals color.

PCB layout is also in attachment.

62.5kHz ringing on signal is shown in DS1Z_QuickPrint135.png.

Is there any suggestions?

Thank you in advance.

The bursts of ringing are synchronous with the video signal, so they are related to that signal.
The repetition frequency of the bursts is 4 times 15.625 kHz (the PAL or SECAM line rate).

From your schematic, it appears you are using Composite Syncs to switch your DG419 to ground to act as a keyed clamper.
That is not normal practice, which is to derive a clamp pulse from your line sync pulse (and from the equalising pulses & serrations  in the field group.)

Clamp pulses are of a standard duration, (somewhat shorter than that of a sync pulse) & also usually delayed so they fall during the back porch of the signal requiring clamping.(sometimes the clamping is performed during the sync pulse duration)

Using the whole composite sync waveform will make the clamping fairly erratic.
It could be that this causes various spurious signals to appear.

If you use the burst/ back porch output you will have a closer approximation to how it is done in Broadcast TV, although the pulse duration will be a bit long.

If you do use this signal, you will need to put a 4.433MHz trap in series with your connection to pin (1) of the DG419, so you don't mangle the burst.

For fault finding I would suggest you remove the connection to the DG419.
You should now have a composite video signal which has lost its DC reference.

If it still has the interference, disconnect the connection to the LM1881.
Again, if that device was the cause, you will have a composite video signal without its DC reference.


I build the circuit on breadboard, it worked fine with DG419 thus i don't think that is the problem.

Are you sure it's not coming from the 555 -9V generator, although its frequency should be about 48kHz with those values. Its 3200uF cap seems a very large, probably too large to suppress fast switching spikes.

I've just realised that 15,625*4 = 62,500 which might help to give a clue, or not, to where it's coming from.

I omitted 555 circuit and fed my bench power supply to give negative voltage to test this however there was still ringing.

The layout makes me think that there is excessive parasitic capacitance between the inverting input and ground lowering the phase margin of the operational amplifier causing it to oscillate.  This is a problem if the ground plane and flood fill is not cut away around the inverting input.

The impedance at the inverting input in parallel with the parasitic capacitance to ground creates another pole inside the feedback loop which lowers phase margin.  One solution to this is to add a compensating capacitance in parallel with the feedback network.


Ground plane is in B.Cu and opa657 is in F.Cu in my PCB board. Do I need to still cut out in back side of the PCB to lower the capacitance? I cut the PCB to isolate ground plane around opa657 however there is still noise?

Do I need to design PCB without ground plane?

Isn't 15,625 Hz a horizontal scan rate for a television operating on 50 Hz?  Could there be a CRT or LCD display close by generating the 62.5 kHz as a multiple of 15,625 Hz?  It soulds like the circuit may be sensitive to light of some sort.



It is sensitive to light but same circuit on breadboard worked fine in same environment.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: 62.5kHz ringing Noise on signal
« Reply #9 on: May 08, 2019, 05:28:53 pm »
I like this bit.  :)

"to transmit video stream through visible light."

"It sounds like the circuit may be sensitive to light of some sort"


Check that the 62.5kHz pulses are not in the transmitted signal at the TX side.
On the RX I'd have a look around all the pins of the LM1881 to see if the 62.5kHz switching is there.
« Last Edit: May 08, 2019, 05:50:05 pm by StillTrying »
.  That took much longer than I thought it would.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: 62.5kHz ringing Noise on signal
« Reply #10 on: May 08, 2019, 06:24:17 pm »
I designed another PCB without ground plane. Can you comment on the design? Do you think there will be interference on signal?
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: 62.5kHz ringing Noise on signal
« Reply #11 on: May 08, 2019, 07:22:44 pm »
"Do you think there will be interference"

Quite likely, it could even be worse until you find where the 62.5kHz pulses are coming from.
.  That took much longer than I thought it would.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: 62.5kHz ringing Noise on signal
« Reply #12 on: May 08, 2019, 08:27:18 pm »
It could be parasitic capacitance due to lm1881 and DG419 since all of the integrated circuits including OPA657 has same ground plane around their pins and LM1881, DG419 are probably switching according to 62.5kHz/4 = 15.625kHz as others have stated.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: 62.5kHz ringing Noise on signal
« Reply #13 on: May 08, 2019, 09:41:51 pm »
The layout makes me think that there is excessive parasitic capacitance between the inverting input and ground lowering the phase margin of the operational amplifier causing it to oscillate.  This is a problem if the ground plane and flood fill is not cut away around the inverting input.

The impedance at the inverting input in parallel with the parasitic capacitance to ground creates another pole inside the feedback loop which lowers phase margin.  One solution to this is to add a compensating capacitance in parallel with the feedback network.

Ground plane is in B.Cu and opa657 is in F.Cu in my PCB board. Do I need to still cut out in back side of the PCB to lower the capacitance? I cut the PCB to isolate ground plane around opa657 however there is still noise?

Do I need to design PCB without ground plane?

No, but the place to start is to remove the ground plane from proximity to the inverting inputs to lower the parasitic shunt capacitance.  This is a common problem with high speed operational amplifiers.

As a test, add a small amount of capacitance in parallel with the feedback network.  Up to 10s of picofarads is usually enough.  This will cancel out any pole caused by the parasitic capacitance to ground.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: 62.5kHz ringing Noise on signal
« Reply #14 on: May 08, 2019, 10:13:00 pm »
Adding capacitor in parallel with feedback loop helps a lot but its not enough there is still distortion on signal.

So can we conclude that ground plane was the problem?
« Last Edit: May 08, 2019, 10:18:27 pm by syntax333 »
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: 62.5kHz ringing Noise on signal
« Reply #15 on: May 09, 2019, 12:20:35 am »
"So can we conclude that ground plane was the problem?"

It depends on how the 'distortion' changed with the extra cap. If the frequency and timing of the 'distortion' changed quite a bit the problem is likely right there. My crystal ball says only the amplitude of the pulses lowered a bit so it's probably somewhere else.
« Last Edit: May 12, 2019, 01:19:27 pm by StillTrying »
.  That took much longer than I thought it would.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: 62.5kHz ringing Noise on signal
« Reply #16 on: May 12, 2019, 01:20:40 pm »
Did the OP find where the 62.5kHz pulses are coming from. r?
.  That took much longer than I thought it would.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: 62.5kHz ringing Noise on signal
« Reply #17 on: May 13, 2019, 04:22:16 pm »
Contrary to what I thought, carving the PCB board to isolate OPA657 did not help. After testing circuit with working version on breadboard I found out the problem was the switching buck converters on the Transmitter side. However, it was probably due to common ground for PSU (Counterfeit LM2596 has switching frequency of 62.5 KHz  :palm:), Analog and Digital signals. I didn't know that I should separate ground of analog circuitry from grounds of Digital and PSU. Probably counterfeit PSU switching caused noise on the ground plane thus on the analog signal. After I changed buck converters to linear 78XX regulators on both transmitter and receiver the problem was solved. In addition, I also changed 555 circuit which could cause also noise on ground plane due to its switching, to dedicated negative voltage supply IC (ICL7660).

Thank you all for your valuable suggestions which helped me to find the problem.
« Last Edit: May 13, 2019, 04:26:27 pm by syntax333 »
 
The following users thanked this post: StillTrying

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 22436
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: 62.5kHz ringing Noise on signal
« Reply #18 on: May 13, 2019, 11:28:21 pm »
Oh, those stupid regulator modules?

Counterfeit or not, doesn't really matter, so much as their layout is terrible, and filtering nonexistent.

You can fix that a bit by connecting to the module with three wires, which in turn connect to your board in a single point (well, not actually a single point obviously, but as near to a single point as is practicable).  Two ceramic caps tie the two power leads together with the ground (eliminating high frequency differential noise), and the module having no additional connections eliminates common mode noise.  (Well, perhaps.  Ideally, the module sits inside a metal box, with the single-point connection being made at a single hole in said box.  Or something topologically equivalent to that; it can be practical to realize with traces and pours on a PCB for example, for some degrees of "topology".)

If that's not enough filtering, then chokes can be used in series with the power lines, then another pair of ceramic caps, to get a pair of pi filters.  And so on.  All the while, make sure ground fully surrounds the filter, on top and bottom, well stitched with vias.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf