Problems from walkie-talkies

Go To Last Post
35 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

We just learned that our controllers in the field exhibit "weird" behavior when a walkie-talkie is keyed for transmit near them.

-- Commercial, name-brand units (e.g. Kenwood, Motorola) on legal frequency bands.
-- Power setting unknown at this time, but probably/perhaps the max for this type of unit.

We'll be doing more investigation. The request here is for ideas on what might be affecting our AVR apps.

This particular unit is in a stainless-steel case, but there is a window for display/keypad.

AVRs inside running at 4MHz-8MHz. Could the crystal be affected? What would typically be affected by this type of external noise?

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The high energy from a walkie-talkie is very difficult to block. It prefers to travel over the cables ... kind of a gateway ;)
How to solve it: can you post a picture ? What cables are going in/out ?

In noisy environments, a full swing crystal oscillator is a must. Nasty when it's battery operated.

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ferrite beads on hi impedance lines like a/ds? Smaller value pullups to make their impedance lower for less antenna-like behavior?

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

All the cases of this I've seen first hand the interference was limited to the analog portions of the circuitry. All the digital stuff (processor, logic gates, etc.) seemed to be unaffected and there were no software crashes, resets, etc. Of course, the analog signals going all over the map made the overall machine malfunction but the micros chugged along nicely. So I guess my suggestion is to try to make all the sensor inputs as noise immune as possible.

+1 on the ferrite beads.

Letting the smoke out since 1978

 

 

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My experience mirrors what has been said. 100pF caps, ferrite beads etc will usually show an improvement.

Someing else to be aware of - these walky talkies usually have a "rubber ducky" antenna. These tend to break internally and cause some spurious emissions along with poor range. If some walkie talkies are worse than others (assuming the same type) then this may be an explanation.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Same here as the rest of the folks that responded.

++ for Ferrite beads.

One regularly overlooked possibility is... Use good quality shielded wiring on anything leaving the box, especially the analog, that you suspect will be problematic to RF and the typical electrical noise in an industrial environment.

Another possibility is that the RF is coming in thru the power supply mains. I've seen this problem several times. I don't really know if the RF messes with the power supply, itself or if it simply finds it's way in through lengths of wire inside of the box.

A good quality Corcom noise filter, right at the entrance of the box, is a must for most high noise industrial electrical environments.

I've attached Corcom's web-link for your convenience.
http://www.corcom.com/

You can avoid reality, for a while.  But you can't avoid the consequences of reality! - C.W. Livingston

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Power setting unknown at this time, but probably/perhaps the max for this type of unit.

Does the user of your product have control over the walkie-talkies (HTs)?

If this was a fast food service and the counter people were talking to the drive through, for example, you might be able to adjust the output power level on the HTs.

If it is police/security or others that only occasionally are in the area, and need "full power" for their job, then this isn't a solution.

Note that many HTs have a high power, (VHF typically 5W, UHF typically 4 W), and a low power (often 1/2 or 1 W) mode.

El-cheapo units often do not have this feature.

Even though the HTs are "high quality" Kenwood and Motorola units, type certified for the frequencies and user, for this kind of interference that won't really matter. A slight freq drift over operating temperature range and voltage, or a bit of spurious harmonic energy transmission doesn't matter. Your micro project isn't a highly tuned circuit. You are just being swamped by the RF EMI as noted above.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I had field units suddenly get spurious resets - due to 802.15.4 transmitter in same unit. Only some units. After great pains, what cured it was a small (like 20pf) capacitor, very short leads, right at the reset pin of the chip, to ground plane. A larger cap didn't work.

Walkie talkies (handheld portables as they're called) like the OP mentioned are much higher powered - 5W if user selects hi power mode. RF will conduct in through wires that enter a steel enclosure. Aluminum is almost RF transparent.

Long ago, I owned a VW hatchback with one of the first (analog) engine fuel injector controllers. When I, as a ham radio guy keyed my radio, the injectors cut off. Same happened once when stopped at a light, next to a police car who's guy was talking.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

digitalDan wrote:
All the cases of this I've seen first hand the interference was limited to the analog portions of the circuitry.
The higher impedance parts of the circuit.
Some op amps include a RFI or EMI filter to reduce this effect.
The following states the likely impacts are on amplifiers, data converters, and regulators:
Precision Analog Designs Demand Good PCB Layouts by Rick Downs (Texas Instruments)
Go to "Analog receptors: electromagnetic energy".
IIRC the offset shift occurs due to rectification of the RF energy.
Tips and Tricks for Minimizing EMI/RFI in your Circuit Design – a Great Tutorial from the Analog Apps Team by G Hann (Texas Instruments, Jan 22 2010) states a presentation "Tackling EMI/RFI at the Board and System Level" by Thomas Kuehl; I couldn't find it.

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Aluminum is almost RF transparent
How about stainless-steel like the case this unit is using?

I don't have any experience with stainless-steel enclosures however, because of the fact that stainless-steel is non magnetic (unlike mild steel), it may not provide good shielding.

Wild guess here.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I dare say it is not the material as such but more the application. Aluminum forms an oxide layer so an Al to Al interface may not be too good. Stainless has a similar property but there are many grades of stainless both magnetic and non.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What frequency are the radios transmitting on?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Aluminum is almost RF transparent.

Really? Interesting. Maybe couldn't put 2.4g RF module inside aluminum case without badly detuning the antenna.

(Sorry, has nothing to do with keeping RF from disrupting your device.)

The largest known prime number: 282589933-1

It's easy to stop breaking the 10th commandment! Break the 8th instead. 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have developed proximity switches in thin faced stainless steel. It is transparent at about 80 KHz, so the problem is likely the cables. A few large clamp-on ferrites applied near the enclosure would be my first try.

It all starts with a mental vision.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks, all, for reading and responding.

-- These industrial controllers have been in the field for a few years now. Total count in three digits.
-- The controllers are somewhat modular in configuration. The particular installation in question has the "main board" and "display board" in the enclosure in question. The short summary is that connections are AC power in, and a shielded 4-conductor twisted-pair cable going to remote module(s)--Gnd, 24VDC, and RS485.
-- So there is no analog in this controller box.
-- The display board is "smart" with a Mega164 and it is mounted close to the enclosure window on standoffs. The main board has a Mega1280 mounted below the display board with connections via riser cards.
-- Both AVRs are running at 7.3728MHz via crystal; fuses are set for "full swing".

The situation has just been reported so we are just starting to dig into it on the bench. We did try the walkie-talkie thing a few years ago...yes, it affected operation but only if the radio was triggered right next to the circuit boards, as if it was held inside the enclosure. Not a typical scenario. ;)

The unit in the field (with video) has the triggered radio waved in front of the controller enclosure while triggered. An immediate effect can be seen on the display board.

Re the AVR's /RESET -- all of our AVR designs have a pullup R and a small C to Gnd.

There is more work to do on the bench. Reset causes are logged. The first tests seem to indicate that the AVR is reporting Brownout as the reset cause.

Re radio power--the application is generally a food processing plant, with lots of stainless-steel tanks and such. So I'd speculate the maintenance staff wants/needs to keep radio power cranked up to whoopee levels.

We are getting a sheet of transparent film that is "EMI shielding" to see if it helps.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch wrote:
-- So there is no analog in this controller box.
"Digital Engineers don't believe gates are differential amplifiers" - Dr. Howard Johnson
Interesting that brownout.
Possible RFI via the regulator.
Ref.
Why Digital Engineers Don't Believe in EMC by Dr. Howard Johnson (Spring 1998).

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The brown-out is an interesting clue.

AVR's are tough to get into a "Latch-up"-state. But there are many IC's that are much more sensitive to latchup. I found that recently on a GLCD module. An overshoot of 1V, just 20ns duration, on the E line made the display latchup: which resulted in drawing much more current (factor 10 or so)

In your application the RF from the WT's can bring one or more IC's in latchup. Power drops, AVR browns out. And btw: typical behaviour of IC's in latchup: they need a complete power down - power up cycle to recover. Reset doesn't help to recover.

What is latchup ? http://en.wikipedia.org/wiki/Lat...

Hope this helps ...

Nard

Edit: gchapman wrote:

Quote:
Interesting that brownout.
Same thoughts ...

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

And btw: typical behaviour of IC's in latchup: they need a complete power down - power up cycle to recover. Reset doesn't help to recover.

(Remember that we are early-on in this investigation...field report and re-creation on the bench. Little investigation so far.)

The controller "comes right back" after an event with the radio. No power cycle or external reset needed.

A test with a fast-food intercom radio (set for less than max power) doesn't cause problems when at a normal distance; even fairly close to the face of the enclosure.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
The controller "comes right back" after an event with the radio. No power cycle or external reset needed.
I understand. But is the complete system working again after an event, in a proper way ? So is it "just" a re-boot you're facing ?
If so, possible RFI via the regulator (as suggested by gchapman)

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch,

The way you are describing the physical layout you are probably getting EMI impingement thru the display window. But it's also possible the various supply and RS-485 cables are carrying the RF into the circuitry.

A quick test to determine which is the culprit would be to block the window with, or replace it with, a piece of aluminum or copper securely connected to the main stainless steel casework. But, can you tell if the unit is messing up in this non-visible situation? Perhaps thru actions of the RS-485 remotes?

If the problem disappears with the solid metal window, try overlaying the glass window with various metallic mesh screens. These will probably affect visiily/useability of the display. Still, they will give you an idea of the severity of the problem and the degree you will need to go to shield the window.

Your transparent EMI film may do the trick. In military applications this type of shielding is accomplished with "ITO Glass". This is "ordinary" sheet glass coated with indium tin oxide - a transparent and conductive compound used in LCD construction. If the window is your impingement entry point, the ITO glass is probably your best final solution. Not sure where to get it, but I can find out if you get to the point where you want to give it a try.

If these window experiments prove negative, the cables are the likely culprit. In this case the more conventional cures like entry caps, ferrite beads, EMI power filters and the like, will solve the problem if applied in appropriately copious amounts.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

But, can you tell if the unit is messing up in this non-visible situation?

LOL--is this the case where we need to put that dude's cat in the box and decide if it is alive or not? :)

Quote:

Your transparent EMI film may do the trick. In military applications this type of shielding is accomplished with "ITO Glass". This is "ordinary" sheet glass coated with indium tin oxide - a transparent and conductive compound used in LCD construction. If the window is your impingement entry point, the ITO glass is probably your best final solution. Not sure where to get it, but I can find out if you get to the point where you want to give it a try.


Interesting.

A complication with shielding: There is a capacitive keypad mounted below the display. I'm just guessing that it may not be "happy" with ITO glass or shielding film or whatever.

(Should be interesting, eh?)

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Capacitive keypad .... that's interesting indeed :)

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

theusch,

"below the display" - Does that mean between the display and the cover glass? Or, a separate touch screen adjacent to the bottom edge of the LCD display proper? Give me a better verbal "picture" of the geometry.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My guess is with Plons and Chuck. The most wire is in the RS485 network, and I have found RS485 transceivers can latch up. About 20 years ago I did ESD and EFT, on a few different RS485 transceivers. Not all are created equal. I seem to remember, one that was rated at 15KV for ESD was bullet proof in all tests.

The others when latched would current limit a 1A SMPS.

It all starts with a mental vision.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

"below the display" - Does that mean between the display and the cover glass? Or, a separate touch screen adjacent to the bottom edge of the LCD display proper? Give me a better verbal "picture" of the geometry.


2x4 capacitive touch "buttons" on a circuit board glued to the inside of the enclosure window below the display board. Largish pads, about 20mm x 20mm.

http://digital.bnpmedia.com/prin...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Funny link, Lee ;) Spam ? :)
Those buttons make nice antenna's. Ouch.

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

Funny link, Lee Wink Spam

It is a link to a PDF that has a picture of the controller (Hydrite Optimax). Indeed the link looks funny; I tested it in Firefox and it opened the page OK. Let me see if I can do better...

Attachment(s): 

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I didn't know I had to look for Hydrite Optimax .... now I do, and it makes sense.

Quite an area that is exposed to RF power. Do you use Atmel's Qtouch SW ?
SOP to prevent RFI is a LPF using a resistor (or inductor) and a capacitor to ground. Transzorbs are optional. Unfortunately LPF a no-go for a capacitive touch panel. OTOH, you could give it a try using 47 to 100 pF, close to the AVR. Sensitivity of the touchpanel will decrease of course.

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

First we need to do more work to determine what is happening when the event occurs. Then see what can be done in that area. At this point it could be any/all/none of above-mentioned target areas. Cabling, power supply, AVR /RESET, touch pad, display, latchup, ...

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Plons wrote:
Unfortunately LPF a no-go for a capacitive touch panel.
Atmel states that the QTouch design "... form a natural low-pass filter for incoming fields; the roll-off frequency of this network is ...".
Atmel also states a possible avenue for RFI (protection diodes).
Secrets of a Successful QTouch(tm) Design (Atmel; go to page 9 and "4 EMC & ESD").

"Dare to be naïve." - Buckminster Fuller

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks for the link.

Nevertheless, the wire or track for the touch runs from the exposed frontpanel to the Qtouch or AVR. And *that* may upset that IC. Directly, or via Gnd and Vcc. The problem is that the RF has a way to get in ! And once in, it can do bad things. Will do ...

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tessa and Tina, You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Interesting thread, I was semithinking of offering touch switches for a possible upcoming project (prototype stage now, real switches).

But it is likely that the unit may need to work near "Aussie bush" strength CB radios so I may reconsider.

It may become expensive putting the unit through EMI/EMC testing.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

John, you may be able to get a pre-compliance check where the lab will do the required test but with less rigour and documentation. You basically get a yes or no answer. Yes, it has a good chance of passing or no, it dies at XX value.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If the only thing that has changed ie no repairs were made or replacement units or cables, I would check all the earths for the units. I then would check the cables entering the units. ensure any shielding on cables is attached at least one end(faraday shielding).

if you suspect RF energy is getting in to the circuits, use a good scope on all the lines connected to the effected board. as stated above RF decoupling at all entry points to the case should be considered.

One point you mentioned was the display/keyboard opening, this opening can act as a tuned cavity at certain frequencies. if this unit is not new I doubt it being the cause but is a feature that needs consideration during case design now days.

regards
Nigel Fenton
DE G0JKN

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Measuring Structural Resonances by Douglas C. Smith describes how to measure the resonant frequencies of a structure (box, PCB, etc.).
The resonant frequencies are possible avenues for EMI or EMC incompatibilities.
There are other interesting articles at his web site:
spark gap wideband EMI sources
at his home page, search for "probe", search for "resonance"
destruction of electronics by a mobile phone

"Dare to be naïve." - Buckminster Fuller