Quantcast

KX1 build issue after Part 1 of build

classic Classic list List threaded Threaded
8 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

KX1 build issue after Part 1 of build

Frank Cox
(this is a repost after accidentally adding this to another thread)

I'm building a KX1 and finished Part 1. I encountered problems during testing.

First I noticed when scrolling through the MENU items some some segments of the LED display don't light on some menu item names. However the same segments will light on another menu item so it doesn't seem to be a problem with the display itself. Also, tuning with the encoder and band switch displayed as expected.

But now the display started going dark during tests and not coming back on after button clicks or encoder turns. When it does this the 5v and 6v lines still have correct voltages and nothing feels hot to the touch. The only way to get it to come back is to turn it off and wait a while.

I've carefully checked parts placement and solder joints and can find no problems so far, and it passes all the part 1 voltage tests.

Don, W3FPR suggested reflowing all the solder joints. Anything else?

73, Frank KK6WUU


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: KX1 build issue after Part 1 of build

Don Wilhelm
Frank,

It sounds like the MCU is getting "mixed up".  Try doing a Reset to
Factory Defaults.
See page 69 in the manual.  It is a 2 step process.
1) Hold BAND and MENU together and power on - release the buttons after
the LED test.  E09 should be displayed.
2) Turn power off, then back on and you should see E10.  Tap any switch.

73,
Don W3FPR


On 4/28/2016 3:01 AM, Frank Cox wrote:

> (this is a repost after accidentally adding this to another thread)
>
> I'm building a KX1 and finished Part 1. I encountered problems during
> testing.
>
> First I noticed when scrolling through the MENU items some some segments of
> the LED display don't light on some menu item names. However the same
> segments will light on another menu item so it doesn't seem to be a problem
> with the display itself. Also, tuning with the encoder and band switch
> displayed as expected.
>
> But now the display started going dark during tests and not coming back on
> after button clicks or encoder turns. When it does this the 5v and 6v lines
> still have correct voltages and nothing feels hot to the touch. The only way
> to get it to come back is to turn it off and wait a while.
>
> I've carefully checked parts placement and solder joints and can find no
> problems so far, and it passes all the part 1 voltage tests.
>
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: KX1 build issue after Part 1 of build

K9MA
I recently built a KX-1, and had the following problems (none of which
were my fault):

1.  Factory installed solder bridge on the DDS chip pins, shorting out
clock input.

2.  Bad DDS chip causing really bizarre behavior.  Unrelated to above
solder bridge and definitely not a bad connection.  Elecraft had never
seen that before.

3.  Bad solder joint on pre-installed surface mount mixer chip on
carrier board.

It works fine now.

Based on my experience, I would be very suspicious of every surface
mount IC connection.  Look at voltages and waveforms right at the
surface mount package lead.  Unfortunately, you need a microscope and a
very fine-tipped soldering iron to re-solder them, and it's very easy to
create new solder bridges.  Elecraft support suggested a piece of number
20 wire wrapped around the tip of a soldering iron.  I used a Metcal
soldering station, but replacing the DDS chip was still difficult.

73,

Scott  K9MA

On 4/28/2016 06:27, Don Wilhelm wrote:

> Frank,
>
> It sounds like the MCU is getting "mixed up".  Try doing a Reset to
> Factory Defaults.
> See page 69 in the manual.  It is a 2 step process.
> 1) Hold BAND and MENU together and power on - release the buttons
> after the LED test.  E09 should be displayed.
> 2) Turn power off, then back on and you should see E10.  Tap any switch.
>
> 73,
> Don W3FPR
>
>
> On 4/28/2016 3:01 AM, Frank Cox wrote:
>> (this is a repost after accidentally adding this to another thread)
>>
>> I'm building a KX1 and finished Part 1. I encountered problems during
>> testing.
>>
>> First I noticed when scrolling through the MENU items some some
>> segments of
>> the LED display don't light on some menu item names. However the same
>> segments will light on another menu item so it doesn't seem to be a
>> problem
>> with the display itself. Also, tuning with the encoder and band switch
>> displayed as expected.
>>
>> But now the display started going dark during tests and not coming
>> back on
>> after button clicks or encoder turns. When it does this the 5v and 6v
>> lines
>> still have correct voltages and nothing feels hot to the touch. The
>> only way
>> to get it to come back is to turn it off and wait a while.
>>
>> I've carefully checked parts placement and solder joints and can find no
>> problems so far, and it passes all the part 1 voltage tests.
>>
>>
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
> Message delivered to [hidden email]


--
Scott Ellington  K9MA
Madison, Wisconsin, USA

[hidden email]

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
Message delivered to [hidden email]
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: KX1 build issue after Part 1 of build

Frank Cox
In reply to this post by Don Wilhelm
Thanks Don,

I tried doing a reset and it came up E09 as it should but after a short time the display went dark and came back on with only the E illuminated extra bright. This is another symptom I've noticed that I haven't mentioned before. When the display is going off while I'm actively giving inputs (so not an idle timeout), any one digit my come back on extra bright, with or without other digits illuminated normally.  

The above happened while the circuit was warmed up. When I first tried to this morning cold it worked as described in my first post for about half a minute. Then it started going dark as I said. If I turn it off a short time it will "work" for a while and start going dark again. If I turn it off longer it takes longer to act up.

This makes me think of heat but nothing gets warm to the touch. I checked with an IR thermometer and the active components warm a few degrees from cold but none more than the others.

I also started today with a battery supply to eliminate possible issues which the switching supply I was using.

Frustrating.  
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: KX1 build issue after Part 1 of build

Frank Cox
In reply to this post by K9MA
Thanks for that Scott,

I did a careful visual of the SM components with magnification after I read your post. Everything looks ok.

I did have a moment of elation when I saw how close the solder pads of C39 are to the pins of U2 but looking at the schematic they should be connected.

Still trying.

73  
Rod
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: KX1 build issue after Part 1 of build

Rod
This post has NOT been accepted by the mailing list yet.
In reply to this post by Frank Cox
I was just reading about this problem and I have a suggestion.

The PIC micro is used to multiplex the LED display as well as read
the encoder and battery voltage. I don't have access to the code but it probably
goes something like this.

1) Send data to LED0
2) Read the encoder
3) Send data to LED1
4) Read the encoder
5) Send data to LED2
6) Read the encoder
repeat 1 -6

The battery might also be in this loop - it doesn't really matter. If you execute this loop
very fast, the display will not "appear" to flicker. However, if you want, say 3 ma per segment,
you must drive it to 9 ma since the time spent on each LED is 1/3 of the scan time.

If the pic stops "scanning", the LED current will be three times as much. Either the micro
is going off and executing code where it shouldn't or the clock for the micro is stopping.
I would look carefully at the 3.995 MHz crystal. Perhaps "bumping" C24 or C25 with another cap
might get it going once it has stopped. Many times, simply measuring it will start or stop it.

Do you have a scope?

72

Rod
 
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: KX1 build issue after Part 1 of build

Frank Cox
In reply to this post by Frank Cox
I found a solution. I noticed that pressure on the power connector would sometimes cause a reboot. Reflowing a second time with more solder in that area cleared up all my problems.

My theory is that short interruptions in power made the system glitch and create the symptoms. Remember that the display would go dark and not come back on even though I could see 5v and 6v power on the board.

The other display issues were due to me not understanding the 7-segment alphabet used. For example the strange looking T. Once the display stayed on more than 30 seconds I was able to work it out with email help from support.
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: KX1 build issue after Part 1 of build

Frank Cox
In reply to this post by Rod
Thanks Rod,

That fits with my ideas which I put in my last post where I talk about the solution to the problem. But I didn't have a reason for the one digit being brighter. Your idea would complete my  theory.

73,

--Frank
Loading...