Unable to find/cure intermittant PSoC3 Reset Cause

[Anonymous]

 Using CT8CKIT030 for development purpose.

   

Getting an intermittant Reset that is not as a result of my code (working with defaults) and experiencing variables being set to values which indicate a Reset has occurred.

   

CyResetStatus returns all zeroes. Understand that a PRES is not included in the CyResetStatus save; thus a PRES is the probable culprit.

   

Using an ADC DeltaSig which turns on PRES. I have an external 4096Vref that I monitor to provide an emperical correction of the ADC results and also use the Offset with a volts conversion to fine tune any variations (an oscilloscope shows Vdda has variation and I use the ADC Vdda to Vddd {set at 5 volts] to reduce granularity. So basically I don't feel that the ADC PRES activation is necessary in my case.

   

Steps taken:

   

WDT not activated (turned off my code). No change noted.

   

Next attempt was to also include a 9v battery with the wall power jack contacts jumpered to make the battery functional with the plug inserted. Results were inconclusive.

   

Next attempt was to include a UPS to prevent utility power glitches to not cause a Reset (UPS was capable of backing up a small load and should switch in ~40 seconds). Results were encouraging but not a total success.

   

Latest attempt was to add a 47ufd 50v ceramic cap to Vin on an external board. This showed the most improvement but was not 100% effective.

   

Can the 5v Vdda and/or Vddd also have a filter capacitor added ala the Vin cap?

   

All sugestions for a resolution are wecome.

   

I can not live with my system occaisionally Resetting as it is a Data Acquisition AND Control project.

[HeLi_263931]

The bigger memory is helpful when looking at your issues I think. For the same time period you can use a higher sample rate, giving you more detail. Or you can use the same rate and measure for 12 times the duration. Also, AFAIK the Rigol has segmented memory, which can be very helpful.

   

Regarding the frequency limit: its only an issue when your signal is faster than 50MHz. The bandwidth limit has nothing to do with the Niquist effect (the sample rate is still 1Gs, so for a 50MHz signal you have 20 samples per period, the same as for the Hantek). And even then the signal will still display fine, just with lower amplitude. If it still is a concern for you, head over to the EEVBlog forum, look for the articles concerning the rigol scopes and be happy 🙂

   

One idea for four channels: connect two of the to the PSoC. Then pulse them in a unique sequence during startup, and use that one for a pattern trigger on the scope. With the long memory, you can then look at whats happening before this startup occurs (the Rigol can combine flank, level and duration conditions on multiple channels).

[HeLi_263931]

Oh, and AFAICS the Rigol should cost about the same as the Hantek. But maybe it depends on your location...

[ETRO_SSN583]

There are a number of good reviews/teardowns on eevblog, youtube, on lowcost

   

scopes.

       

Regards, Dana.

[Anonymous]

 Dana,

   

The eevblog was one I was looking at. But without a full understanding of the DSO basics, I was not able to fully understand the posts to make a decision.

   

I found two web sites that helped clarify the DSO specifications and operation:

   

http://www.testandmeasurementtips.com/calibration/data-acquisition/the-oscilloscope%E2%80%99s-acquis...

   

https://www.yokogawa.com/ymi/tutorial/tm-tutorial_01.pdf      Touts their products but is educational.

   

My HanTek DSO5102BM choice is now based on better knowledge. Downloaded and read the manual. I have one on order.

   

I am wary of buying new models of any product, so I am passing on the new Rigol DSO.

   

I am trying to get a failure on my bench (w/o operational sensors, etc.) by running the LVIA and LVID up to 4.96 volts (VDDA is ~5.1v on both my CY8CKIT-030 boards); but have not had success in causing a reset. I think the 4.71v on the real system is increasing the reset frequency; but the "MyResetStatus" is still not telling my why they are occurring. MyCases is still open.

   

Thanks,

   

Bruce

[Anonymous]

 Dana et.al.,

   

DSO shows EMI is the cause. Screenshot attached. Hits both VDDA (Blue) and VDDD (Yellow).

   

The second CY8CKIT-030, without any auxiallary boards attached, and only powered by the 9 v battery, also shows the EMI when in the proximitry of the unit. My sensors may make the EMI effect a more significant problem.

   

The screenshot shows EMI from switching on a ~50 watt fan motor located about 3 foot away..

   

Removing 1 probe and shorting it shows the probe is not picking up the EMI.

   

Any suggestions about curing this?

   

Thanks,

   

Bruce

[Anonymous]

 Can't get the riht screenshot attached.

   

Website goes bonkers,

   

A new attachment.

   

Bruce

[Anonymous]

Add ferrite to I/Os may reduce this. (We use this on all our boards)

   

If using A/C power, add power line filters.( We required approved power supply that already passed all the EMC/EMI test)

   

If is coupled via EM wave( ie without any connection to control/monitor the motor but still have those interference), then shield your boards.

[Bob_Marlowe]

Something else to consider: Remove the EMI generation at the source. Check if there are Cs to do that and...

       

Bob

[ETRO_SSN583]

Normally large L spikes occur on disconnect of the L, not turn on. For

   

that fast recovery diodes or zeners at site of L can help.

       

Use your scope, trigger qualify on runt pulse, and step the sweep up

   

into ~ 10 nS / Div type of range. Basically to see what the actual spike

   

looks like. Right now you cannot tell it thats an L decay or a C charge

   

based spike.

       

You have to determine if this is a ground bounce problem at the pins.

       

For very fast pulses like this a LPF at inputs can remove their effects.

   

That can be R-C, L-C...... Also where you have Hi-Z pins, but don't

   

actually need Hi-Z, terminate the input with a low based R to surpress

   

coupling. Keep in mind you have to evaluate noise margin which leads

   

to determining allowable R.

       

Regards, Dana.

[Anonymous]

 Dana et.al.,

   

Gives me some clues to keep the EMI from getting to the board.

   

The board only test shows it to be an air based coupling. External sensors will most likely make it a tougher nut to crack.

   

One source is a fractional induction fan motor and it produces the EMI both turning on and off (strange). The screenshot is of this fan.

   

Because I can't control the environment it will live in, I need to try to keep it from effecting the board.

   

Alternatively, IF I could to figure out how to peacefully co-exist with intermittent resets, it would be a more elegant and robust solution. My Data Acquisition and control is currently based on a 1 second sample period so there may be wiggle room.

   

Because of the size of these posts, and I think co-exist is a new topic, I will open a new topic tomorrow:

   

"Reset Recovery Considerations"

   

I will persue trying to keep it off the PSoC3 kit board. I will use this post topic for my findings. Don't want to ignore any suppression fixes.

   

Been a long road and bumpy travelled. Without everybody helping, I don't think I could have gotten to this point.

   

Thanks,

   

Bruce

[ETRO_SSN583]

Could be fan is capacitor start, maybe thats the turn on transient.

       

Regards, Dana.

[Anonymous]

 Dana,

   

The ~ 50 watt fractional horsepower fan motor is in an enclosure so I am not 100% sure it is an induction motor.

   

Fans require almost no starting torque. Finding that it is a capacitor start motor would leave me wondering why the manufacturer added the higher cost to the unit.

   

The DSO at a nanosecond setting should answer the type motor used. I will run a test at that setting.

   

Incidently, the first DSO screenshot I mistakenly included is of VDDA (Blue) and GlobalSignalRef (yellow) brought out to a resistive digital output pin. The CY8CKIT-030 will run on only USB +5v power, but VDDA and VDDD will be at less than 5v. By setting the LVIA and LVID trip level to 4.71v, and starting with +12v wall power (left side of screen) and then removing the wall power, the dropoff can be seen leading to a reset. VDDD can be seen on the output pin prior to the reset, but after the reset the output pin reflects the GlobalSignalRef. I was setting up the DSO reset triggering setting on my bench prior to running it on my actual unit (allowed me to learn the DSO single trigger by an easily repeatable test). The spike at about +3.2 Milliseconds is probably as a result of the reset being a recurring condition and since it is after the first reset I have not persued it as a problem.

   

But keep the suggestions coming,

   

Bruce

[Anonymous]

 Dana,

   

Here is a screen shot at 2 Usecs. WOW, not what I was expecting for EMI.

   

Regards,

   

Bruce

[HaHa_285681]

Wow! Thats not only one big bang!

       

Bob

[Bob_Marlowe]

Another forum software bug, last post was mine!

       

Bob

[Anonymous]

 Bob & Dana et.all,

   

This particular fan motor is part of a humidifier, the source of electricity and water and therefore danger. So it and the PSoC3 wall power is powered via a GFCI for safety reasons. Don't have a convenient source of non GFCI power, but I think when I get time I will explore if the GFCI very low current detection circuitry is to blame.

   

It sure has me scratching my head!

   

Regards,

   

Bruce

[ETRO_SSN583]

Bruce, nice looking signal integrity......:) 8 uS rep rate, 125 Khz, switching

   

power supply ? Conducted line ? LCD Monitor ? Turn persistence on that

   

setup to see what max max is.

       

You could build sniffers to try and pinpoint source of problem.

       

readingjimwilliams.blogspot.com/2011/12/app-note-70-part-3.html

       

www.eevblog.com/forum/projects/diy-magentic-field-probes/10/

       

www.bcarsten.com/

       

Regards, Dana.

[Anonymous]

 Dana,

   

This is happening when I switch off/on my humidifier. Can happen either way.

   

A sniffer seems unneccessary in this case.

   

Bruce

[ETRO_SSN583]

Ultrasonic transducer, large C load to switch.

       

www.senscomp.com/pdfs/using-piezo-sensors.pdf

       

Regards, Dana.

[Anonymous]

  Hello All,

   

CyResetStatus MyCase conclusion.

   

After 32 days, and numerous MyCase Responses, the final "tally":

   

"LVI/HVI reset cannot be detected from CyResetStatus Variable.  ...  We sincerely apologize for the flaw in the documentation. We thank you for bringing this up. We have got PSoC Creator System reference guide corrected which will be released with next version of PSoC Creator software."

   

I apoligize for any of my Forum Posts that could have led others down the garden path with me.

   

Regards,

   

Bruce

[Janetok]

1. PRES Activation: Even if you feel the ADC PRES activation isn't necessary for your specific use case, it may still introduce instability if it's causing some kind of power fluctuation. If you can, try to disable it or bypass it to see if that changes the behavior.

2. Filtering and Decoupling: Adding decoupling capacitors close to the power pins (Vdda, Vddd, and Vin) of a chip is always a good idea to filter out high-frequency noise and provide transient current support. This is standard practice in most designs. The value of the capacitor can vary, but 0.1µF to 10µF are commonly used values.

Now, addressing your specific question about adding a filter capacitor to the 5v Vdda and/or Vddd:

• Yes, you can (and should) add decoupling capacitors to both Vdda and Vddd. This can help stabilize the internal power rails of the chip and provide a reservoir of charge for transient loads. Typically, you would see a combination like a 0.1µF ceramic capacitor in parallel with a larger capacitor (like 1µF to 10µF) for both Vdda and Vddd.

• Place these capacitors as close as possible to the power pins for maximum effectiveness.

• Make sure you're using capacitors with appropriate voltage ratings. A capacitor rated for a much higher voltage (e.g., 50V for a 5V rail) will work, but can be overkill and might not provide the best high-frequency response. Stick to capacitors rated slightly above your operating voltage, like 6.3V or 10V for a 5V rail. Good to know how step recovery diode functioning.

Finally, while adding capacitors can help, ensure that your power supply source itself is stable and can provide the necessary current during peak loads. If you have an oscilloscope, monitor the voltage rails during operation to see if there are any noticeable drops or spikes which could be causing the resets.