Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
pin 2 of raduino board seems to have gone to CW TX full time after boot message.
|
Odd issue popped up while working with the transmitter.
Radio went to CW and stayed keyed.? Tried power of and suoon as it booted it went CW and keyed. This is odd as the blue, purple, and green wires are not even in the connector as i"m not using the key. The 4.7K resistor is in place on the green/blue connection pair. If I measure pin2 (blue) to ground I get .1V save for ther is no connection and it ohms open (power off). If I connect blue to +5 it stops but the pin is pulling down 25ma! No ESD likely strap and conductive pad in use. If I didn't know better id swear the input self reprogrammed as an output! I may have to do the programing dance on this before pulling the nano. Firmware is? factory v0.20 on startup. Any thoughts? Allison |
¿ªÔÆÌåÓýAllison you should check your 4.7k resistor solder connections you should be measuring 5vdc on both sides and ground. Pin 3 is the 5vdc supply and the 4.7k resistor is setting pin 2 high. Do you have the CW jack wired up. If so what is plugged into it, that could be what is drawing low enough to switch into transmit.Skip Davis, NC9O? On May 19, 2018, at 22:56, ajparent1/KB1GMX <kb1gmx@...> wrote:
|
|
Hello Alison,
Head scratching really since you don't even have the usual suspect wires in the connector. It it was me I would pull the Raduino off the main board, reload the factory software (the pin is explicitly configured as an input with pull-up resistor).? If unsuccessful I can't see anything than a short on the circuit, which you tested, or a dead Arduino analogue inputs.? That would make number 4 in less than so many days. By the way I am following your driver stages re-design with interest. All the best, 73, John (VK2ETA) |
|
John,
One L or two it still sounds like my name.? ;) I had noticed but could not find the earlier reports but I remember seeing them. The usual checks like is there a short/leakage path to ground come up no. If I put a milliamp meter in the between pin2 (raduino) and 5V it start up open and after the init display with the version number it then goes to 25ma.? It does not power up that way so its not an input failure. That suggests its programmatically triggered. The analog input pc1A1 can be a digital output if so programmed as I'm familiar with the atmega328 and arduino.. That suggest an bug in the code. Next step is reprogram...? That will be a while as I have other things for the next week so time will be short for working on this.? My first pass is to drop some simple code in to run the display and check that pin. I'll report here. Allison |
|
Did the work to test and verify....? For whatever reason the port is stuck to ground so the
easy solution for now is to change the code to ignore the keyer. Analog pin A6 appears to have developed a low resistance to ground about 150 ohms compared to others at 2-5 megohms. That done I get normal behavior save for no CW.? that's ok for the work I'm doing. I plan to rework the code to suit me. If I have to get a few nanos so be it I can just as easily sub a pro-mini with bunch or wires to the right pins as I have more than a few. Why that pin failed is a mystery, the header had no wire at all, the 4.7K was on the back aout of the way.? A7 works fine and could be used for the same purpose if need be. Allison |
|
Curious.
toggle quoted message
Show quoted text
We've had a bunch of recent reports of those Nano IO cells going haywire, Pretty much never occurred previously, just occasional reports but then it would come out that some wire brushed by the IRF510 heatsink (at +12v).? Or something like that. So could be a recent run of especially dainty Nano's. Alternately, could still be static discharge.? Or a soldering iron with AC leaking to the tip? Allison has been careful with her rig, more so than most, working with an antistatic mat and such, And has worked on plenty of devices much more sensitive than the Nano. So seems unlikely here. But, perhaps a good time to mention that a 1k series resistor on any lines to the the front panel (such as PTT, key, encoder) would drastically reduce the chance of static blowing an IO cell. Though that 1k won't work with the analogRead() scheme on A6 for the key by the stock uBitx code. Jerry, KE7ER On Sun, May 20, 2018 at 09:03 am, ajparent1/KB1GMX wrote:
Did the work to test and verify....? For whatever reason the port is stuck to ground so the |
|
That series 1k would also avoid blowing the IO cell if the wire brushed by 12v.
toggle quoted message
Show quoted text
Something to consider for those laying out alternate Raduino boards. On Sun, May 20, 2018 at 10:08 am, Jerry Gaffke wrote: But, perhaps a good time to mention that a 1k series resistor on any lines to the the front panel |
|
Jerry, there was no wire to flail around brush the board only has the 12V input
as everything is 5V or 3.3.? That's why I am concerned it seemed spontaneous. My guess it was ESD despite usual precautions.? Though in general I've never experienced that and its actually very humid due to the rains. However layout and input protection are important. I tend to put diodes from ground to the input and input to whatever applicable +V so the input cannot see ESD or other excursions from off board sources.? Inputs I worry about are key, mic-ptt, button, Encoder A/B inputs. Its got me, curious. Allison |
|
Jerry, and others... Same thinking here.? This week I worked off-group with a person who experienced this problem and came to the same conclusion.? It may be human error but 3 such incidents in the same week is suspicious.? This may make the case for socketing the Nano so it can be removed, tested separately, and if necessary, replaced.? Many Nano purchases from Ebay suppliers do include pins for this but you still need the sockets. If the Nano was socketed it would be possible to remove it, install test code and determine The suggestion of using a 1K resistor in each input to the Nano is a valid one, and it could I have no idea how much cost might be added to the Raduino manufacturing budget if the Nano were to be socketed, but it might be worth considering even if it required adding a dollar or so more to cost of the Raduino and uBITX main board.? There are other persons who have designed their own Raduino board and have commented about it on this forum.? It would be interesting to see if they are experiencing a similar problem. Arv? _._ On Sun, May 20, 2018 at 11:08 AM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote: Curious. |
|
Allison, and others... I have not taken time to look at the software, but a question keeps nagging me that this may be a software problem.? The "pinMode()" command has an option for turning on the internal PULLUP resistor, or not doing that.? What might be the effect of not turning on the PULLUP option for the encoder push-switch, or for one or both of the encoder phase signals?? pinMode(); pinMode(pin, mode); pinMode(pin, INPUT); pinMode(pin, OUTPUT); pinMode(pin, INPUT_PULLUP); or digitalWrite(pin, HIGH); digitalWrite(pin, LOW); If using interrupts to read the rotary encoder it may be possible that the direction LOW-to-HIGH or HIGH-to-LOW was not set properly relative to any pinMode() command. With that in mind I may have to download the source code and take an in-depth look at what might be happening there. Arv _._ On Sun, May 20, 2018 at 11:50 AM, ajparent1/KB1GMX <kb1gmx@...> wrote: Jerry, there was no wire to flail around brush the board only has the 12V input |
|
Jack Purdum
Using: ?? pinMode(pin, INPUT_PULLUP); is functionally equivalent to: ?? pinMode(pin, INPUT); ?? digitalWrite(pin, HIGH); The first version puts it in a known state just as the second two statements do. I think I checked the list file for the statements a long time ago and both generated the same code as I recall, so nothing is gained in terms of resource use from one over the other. Jack, W8TEE
On Sunday, May 20, 2018, 2:27:03 PM EDT, Arv Evans <arvid.evans@...> wrote:
Allison, and others... I have not taken time to look at the software, but a question keeps nagging me that this may be a software problem.? The "pinMode()" command has an option for turning on the internal PULLUP resistor, or not doing that.? What might be the effect of not turning on the PULLUP option for the encoder push-switch, or for one or both of the encoder phase signals?? pinMode(); pinMode(pin, mode); pinMode(pin, INPUT); pinMode(pin, OUTPUT); pinMode(pin, INPUT_PULLUP); or digitalWrite(pin, HIGH); digitalWrite(pin, LOW); If using interrupts to read the rotary encoder it may be possible that the direction LOW-to-HIGH or HIGH-to-LOW was not set properly relative to any pinMode() command. With that in mind I may have to download the source code and take an in-depth look at what might be happening there. Arv _._ On Sun, May 20, 2018 at 11:50 AM, ajparent1/KB1GMX <kb1gmx@...> wrote: Jerry, there was no wire to flail around brush the board only has the 12V input |
|
Arv,? That is interesting but a whole other subject.? The failure has nothing to do with
the encoder which still works. Right now I want to work TX power, the failure is tested with some of my test code and? the A6 failure is definitely some form on shot to ground likely the internal ESD has failed. For now Its a i'll deal with it later thing. Allison |
|
Allison, Thanks for sharing this problem and getting back to us so soon.
I don't see in the datasheet where ESD protection is specified. Could also be different for analog only ports, as there is noise from ESD protection circuit.. At first my mind was going down the path that there could be variants of the Mega328P that Amtel didn't expect to get in the hands of consumers. The only Nano change I could see is that more analog ports are called. #include "../standard/pins_arduino.h"
#undef NUM_ANALOG_INPUTS
#define NUM_ANALOG_INPUTS? ? ? ? ? ?8
I am still curious, that if you changed setup() to just void setup()
{
? Serial.begin(9600);
? lcd.begin(16, 2);
?
? //we print this line so this shows up even if the raduino?
? //crashes later in the code
? printLine1("uBITX v0.20");??
}and did nothing in loop() if you still see the same behavior on A6.? You had mentioned that it didn't seem to power-up with A6 being driven LO, but was driven LO after the splash. All the ports should default to input,? unless explicitly set, right? If the part got zapped even with all your precautions, well it's damaged. It sucks, but your equipped to fix the problem if you need to. If there's a batch of variants out there, there will be more issues arising.? Sure to create havoc among many users. Catch-22, if there is an optional digital port on A6 for potential variations of Mega328P, there's no public documentation on how to handle the port. I know this is a half baked idea, but still plausible that a variation of the Mega328P was spun for a specific customer and that die was not intended for general market. I do very much like the ideas of making the overall circuit more ESD / Voltage Over Stress robust where possible. Rain stopped here, back to fixing a flat tire.? That problem was an obvious one. Regards, Gary |
|
I had my own code to check the A6 input.? It works but the ohmmeter says 150 ohms to ground with power OFF and either polarity (fluke 77). So I can keep going (production version) on other stuff I just commented out the call to cwKeyer(). In short ignore A6 for now.? Keeping it simple its all doe by one line in the VOID LOOP() code. -----section----------- void loop(){? //**? cwKeyer(); */?
? if (!txCAT)
? ? checkPTT();
? ? checkButton();
.... Since CW is not my big thing I may drop it completely from the code and do my personal version from there.? That will be more buttons and No push button on the encoder as? the encode I'm using will not have one. The buttons will be: ? ?Sideband (toggle U/L) ? ?Band (rotating list 160,80,40,20,17,15,10 back to start) ? ?Tune rate (10hz/1khz/1mhz step select rotating list.) ? ?VFO ( A/B toggle) ? ?Spilt (TX on B RX on A, Toggle ON/OFF)?? ? ?Calibrate (internal button, bring up list for setting cal BFO, 1ST, 2ND) ?I haven't decided on digital in or analog. Allison ? |
|
Hi Allison,
Maybe a left field idea, but instead of ESD could it be RF? Inducted/coupled in other parts of the circuit and finding its return path through that pin? I am saying that because I had some serious RF issues with shielded wires from the control section, especially on 21 and 28 Mhz. All resolved with small caps on the encoder and shielded cables. 73, John (VK2ETA) |
|
Not likely as pins 1,2,3? of the connector didn't even have a wire inserted.? I? popped the pins out
rather than have flying leads.?? That and I was on the dummy load and not transmitting. The only way to know is uncap the die and look at the chip under a really good microscope. I'm not going to waste further time on that. It looks typical of ESD from manufacture or handling and could have occured anywhere from? the the Nano build or after that.? I've worked around it.? I do not need CW right now. Allison |
|
Could be a reason that some Nano's are $2.50 and some are $3.00.
toggle quoted message
Show quoted text
I'd hate to think of all the corners that would have to be cut to hit those prices. Maybe $5 from a mainstream vendor would be a good investment here. On Sun, May 20, 2018 at 03:11 pm, ajparent1/KB1GMX wrote: Not likely as pins 1,2,3? of the connector didn't even have a wire inserted.? I? popped the pins out |
Skip Davis
to navigate to use esc to dismiss