|
A frequent subject, dont see a path for me to pursue. Built qcx kit, December 2017 ship, group 2 board, made the three mods suggested (did RF choke pin 20 to 7 instead of Schottky). Start up was perfect all steps worked fine, heard signals, 20 meters,? noticed only 1.5 watts on transmit but assumed I would get to that later with output tweaking of toroids etc., otherwise fine. Turned it off, turned it on, nothing. Brightest screen just gives me the two rows of squares, etc. Voltage measurements compared to the table are all over the place. Checked 20 MHz osc. with scope, works fine. 5v, 12 v, seem to be where they should be, clk 0 and 1 are 9 mv with a dvm, no scope on that, various odd voltages on ICs etc, but assume without the LCD and microprocessor no point in worrying. None of the issues I have read about so far seem to get me started. I did replace both small chokes with my own older type (look like a 1/4 watt resistor) since both easily self destructed on removal, but no mods, no fancy step. Just need some direction and I will go there I guess.?
Cap
W0CCA
|
Hi Cap
Have you followed the QCX troubleshooting guide?
Jan
|
Yes, I have pretty much exhausted what I can read on the no startup issue, i.e. the discussion on moving the inductor to pins 20/7 etc. It would appear issues beyond the no readout are moot without readout though. Since the unit worked just as it was supposed to at first turn on and all that happened was wouldnt turn on again, I am not looking at any downstream issues yet.?
Cap
|
Didn't early firmware releases have an issue? I remember my first one when I turned on the power, just the backlight came on. But if I turned it off and immediately turned it back on, all was well. I don't remember what Hans said was the issue.
Jack, W8TEE
From: "capallen@..." <capallen@...>
To: [email protected]
Sent: Monday, January 29, 2018 4:10 PM
Subject: Re: [QRPLabs] QCX microprocessor startup
Yes, I have pretty much exhausted what I can read on the no startup issue, i.e. the discussion on moving the inductor to pins 20/7 etc. It would appear issues beyond the no readout are moot without readout though. Since the unit worked just as it was supposed to at first turn on and all that happened was wouldnt turn on again, I am not looking at any downstream issues yet.?
Cap
|
Hi Cap,
A couple of pointers.
Read Han's Trouble shooting guide from beginning to end, do just
skip sections because it doesn't at first sight apply to your
symptoms.
As stated the chances are its an assembly problem not a component
failure, although you have had it working before.
what value RF chokes did you substitute for? The value has changed
in the latest version to 47uH and now
inductor L5 is connected to “Vcc” rather than +5V. A subtle
distinction.
If you want to keep your build inline with the current
versions.
Verified you still have 5v on pins 7 & 20 f the MCU, to discount
track damage caused by the rework.
As Han's points out, for the MCU to complete initialisation at start
up it also requires acknowledgement from the Si5351 IC via I2C bus.
Closely examine solder and check continuity between the two.
Has IC1 also got it's Voltage supply & 27MHz Osc running?
(Take care when probing this IC as it's known not to tolerate
accidental shorts,
and replacement is not easy, but not impossible with care /
experience with SMD rework.)
Alan
toggle quoted message
Show quoted text
A frequent subject,
dont see a path for me to pursue. Built qcx kit, December 2017
ship, group 2 board, made the three mods suggested (did RF choke
pin 20 to 7 instead of Schottky). Start up was perfect all steps
worked fine, heard signals, 20 meters,? noticed only 1.5 watts on
transmit but assumed I would get to that later with output
tweaking of toroids etc., otherwise fine. Turned it off, turned it
on, nothing. Brightest screen just gives me the two rows of
squares, etc. Voltage measurements compared to the table are all
over the place. Checked 20 MHz osc. with scope, works fine. 5v, 12
v, seem to be where they should be, clk 0 and 1 are 9 mv with a
dvm, no scope on that, various odd voltages on ICs etc, but assume
without the LCD and microprocessor no point in worrying. None of
the issues I have read about so far seem to get me started. I did
replace both small chokes with my own older type (look like a 1/4
watt resistor) since both easily self destructed on removal, but
no mods, no fancy step. Just need some direction and I will go
there I guess.?
Cap
W0CCA
|
Cap
Don't rule out something simple like a poor connection, possibly your movement of the rig broke loose a poorly soldered connection to controller or display. Examine these connections and reflow them. Also be aware that lcd adjustment pot has a rather narrow sweet spot where it displays. Don't jump to conclusion that your malady is same as someone else's.
Curt
|
remember to make sure your display clears the toroids underneath, you may need to move the display tabs down towards it’s own board to clear them.
toggle quoted message
Show quoted text
On Jan 29, 2018, at 6:03 PM, wb8yyy via Groups.Io < wb8yyy@...> wrote:
Cap
Don't rule out something simple like a poor connection, possibly your movement of the rig broke loose a poorly soldered connection to controller or display. Examine these connections and reflow them. Also be aware that lcd adjustment pot has a rather narrow sweet spot where it displays. Don't jump to conclusion that your malady is same as someone else's.
Curt
|
Very much appreciate answers from far and wide. On use of my scope which seems good enough for the purpose, an old 60 MHz Tektronix, I see a very small signal on the "freq" test point but nothing on pin 8 or 10 of IC3 and with a magnifier see nothing amiss with the solder joints of IC1. Completely redid all solder joints, see voltage fine at 20 and 7 on IC2. Leaves me at a loss. Hard to believe how it all disappeared given the very straightforward testing and alignment it went through just fine.?
Cap
|
Cap,
The Freq Test Point is an
Input* to the units own built in frequency
counter.
* It also reads CLK2 signal, in calibration mode.
Your QCX does not boot far enough to access those.
The Si5351doesn't appear to be outputting anything from what you
have described in your checks so far.
The next step is to find if this is the cause, or is something else
preventing it.
Volts & 27MHz Osc on IC1 ?
Without the basics of having IC1 communicating with the MCU it will
hang on startup,
and not do any more.
Then we can move forward, to seeing if there is activity on the I2C
bus etc.
Can you post clear images of your PCB Top with & without LCD,
& Bottom.
Alan
toggle quoted message
Show quoted text
Very much
appreciate answers from far and wide. On use of my scope which
seems good enough for the purpose, an old 60 MHz Tektronix, I see a very small signal on the "freq" test
point but nothing on pin 8 or 10 of IC3 and with a
magnifier see nothing amiss with the solder joints of IC1.
Completely redid all solder joints, see voltage fine at 20 and 7
on IC2. Leaves me at a loss. Hard to believe how it all
disappeared given the very straightforward testing and alignment
it went through just fine.?
Cap
|
Alan.many thanks for hanging in. With scope pin 10 ic 3 , clk1, clk2, all show no signal. Voltage at d1 = 3.5. What are you looking for with pics? Can try to do better at that spot. Very difficult to get meaningful pics on board need a loupe to look at. Afraid if need to think about replacing ic1, game, set, over with my skills. Still stumped since went through all setup/startup steps perfectly before shutoff.
Sent from my Samsung Galaxy smartphone.
toggle quoted message
Show quoted text
-------- Original message -------- From: "Alan de G1FXB via Groups.Io" <g1fxb@...> Date: 1/30/18 4:26 AM (GMT-07:00) Subject: Re: [QRPLabs] QCX microprocessor startup
Cap,
The Freq Test Point is an
Input* to the units own built in frequency
counter.
* It also reads CLK2 signal, in calibration mode.
Your QCX does not boot far enough to access those.
The Si5351doesn't appear to be outputting anything from what you
have described in your checks so far.
The next step is to find if this is the cause, or is something else
preventing it.
Volts & 27MHz Osc on IC1 ?
Without the basics of having IC1 communicating with the MCU it will
hang on startup,
and not do any more.
Then we can move forward, to seeing if there is activity on the I2C
bus etc.
Can you post clear images of your PCB Top with & without LCD,
& Bottom.
Alan
Very much
appreciate answers from far and wide. On use of my scope which
seems good enough for the purpose, an old 60 MHz Tektronix, I see a very small signal on the "freq" test
point but nothing on pin 8 or 10 of IC3 and with a
magnifier see nothing amiss with the solder joints of IC1.
Completely redid all solder joints, see voltage fine at 20 and 7
on IC2. Leaves me at a loss. Hard to believe how it all
disappeared given the very straightforward testing and alignment
it went through just fine.?
Cap
|
Hi Cap,
No need the for game over thinking....
Man designed it,
well HAN's designed it.? ;-)
Man built it = you.
Man / you used it, even if it was only for a short time.
It's a positive thing that it did work, it proves you are 99.9%
there.
EVERYTHING is repairable, just some are easier than others.
Plenty of times engineers walk away & grab a coffee, or other
stimulant of choice....
(You would be VERY unlikely to find the problem to be with a new
component,
there something called infant mortality in electronic components.
Some findings are true. Much not, but used by people who don't
understand the big picture
and use it for their own end's.
WAY OFF TOPIC....)
FACT:- one reason SMD's came about, was to enable machines assemble
PCB's cheaply, quickly & RELIABLY.
They are our friends......
Quickly back to IC1
You have reported the supply volts as 3.5. All OK.
I didn't spot you have reported the check on the 27MHz Osc?
measure on a scope the crystal leadout, dangle a wire from the ant
jack of a RX and listen for 27MHz,
failing that, DVM and it should measure close to half supply
volts. Seeing it on a scope is the best check.
To move on to the next step,
are the MCU & Synth talking? I2C bus on IC1 pins 4 & 5
The best place to measure is at their connection to the pull-up
resistors R3 & R4
I have not got a QCX to measure first hand, and report what it
looks like & scope settings.
It will be a digital waveform, so expect a voltage swing close to
0-5v.
Is there anything, a quick burst, or a train of data on both pins?
If not, is at the MCU end, IC2 pins 27 & 28
RE:- photo's,
It's not a personal attack, Han's did find poor SMD machine
soldering on one kit, out of 2500?
Sometimes it just takes a fresh set of eye's to spot something,
If you have a camera phone, or better still a DSLR many have a macro
function and with correct lighting you can get good images of PCB's.
(Some of us HAVE to use them daily, to inspect soldering on SMD's
after middle age.....
When I started and had good eye sight, it was valves &
components mounted on tag strip, and used a 120w solder gun.
Now I have tired eyes, the 'biggest' components are less than 1 inch
square and have +120 legs.
That's progress...)
if you want to, perhaps buddy up with an Elmer or
find somebody at a local club that does his own construction to take
a fresh look at it.
The thing to do is not to do anything you may regret later, one of
my beef's is when you take on a job from another engineer who has
damaged the PCB tracks,
or worse still deliberately cuts tracks to isolate an area of
circuit for "fault finding"...
Until we find out what's wrong, don't warm up the soldering iron
yet.
Every one has talents and weaknesses, You probably never met me on
the air sending CW.
If you do, it will be machine sent.... ?? :-[
Alan
On 01/02/2018 04:04, CAP ALLEN wrote:
toggle quoted message
Show quoted text
Alan.many thanks for hanging in. With scope pin 10 ic 3 ,
clk1, clk2, all show no signal. Voltage at d1 = 3.5. What are
you looking for with pics? Can try to do better at that spot.
Very difficult to get meaningful pics on board need a loupe to
look at. Afraid if need to think about replacing ic1, game, set,
over with my skills. Still stumped since went through all
setup/startup steps perfectly before shutoff.
Sent from my
Samsung Galaxy smartphone.
-------- Original message --------
Date: 1/30/18 4:26 AM (GMT-07:00)
Subject: Re: [QRPLabs] QCX microprocessor startup
Cap,
The Freq Test Point is an
Input* to the units own built in frequency
counter.
* It also reads CLK2 signal, in calibration mode.
Your QCX does not boot far enough to access those.
The Si5351doesn't appear to be outputting anything from what you
have described in your checks so far.
The next step is to find if this is the cause, or is something
else preventing it.
Volts & 27MHz Osc on IC1 ?
Without the basics of having IC1 communicating with the MCU it
will hang on startup,
and not do any more.
Then we can move forward, to seeing if there is activity on the
I2C bus etc.
Can you post clear images of your PCB Top with & without LCD,
& Bottom.
Alan
Very much
appreciate answers from far and wide. On use of my scope which
seems good enough for the purpose, an old 60 MHz Tektronix, I see a very small signal on the "freq" test
point but nothing on pin 8 or 10 of IC3 and with a
magnifier see nothing amiss with the solder joints of IC1.
Completely redid all solder joints, see voltage fine at 20 and 7
on IC2. Leaves me at a loss. Hard to believe how it all
disappeared given the very straightforward testing and alignment
it went through just fine.?
Cap
|
Alan, I did look at the output of the? clock chip with my scope, which is only a 60 mhz Tek and sweeps max about 15 mhz, but you do see rf envelopes up to 50 mhz easily and I see nothing on clk 1, clk 2, or pin 10 of the I/Q chip. So I think nothing going on there. Nice sine wave on the 20? mhz clock.?
I see nothing on IC1 pin 27,28 except 2.8, 3.5 vdc respectively. Nothing on the scope resembling a wave form.?
Cant imagine a pic showing much about IC1 as I use a jewelers loupe to see it at all, with the loupe about 2 cm away !! But I see no bridges and it looks good to me.?
What is this "chip reset" I have seen mentioned here on the blog? Is that something more than on and off?
Cap
|
Alan, I did look at the output of the? clock chip with my scope, which is only a 60 mhz Tek and sweeps max about 15 mhz, but you do see rf envelopes up to 50 mhz easily and I see nothing on clk 1, clk 2, or pin 10 of the I/Q chip. So I think nothing going on there. Nice sine wave on the 20? mhz clock.
I see nothing on IC1 pin 27,28 except 2.8, 3.5 vdc respectively. Nothing on the scope resembling a wave form.
Cap You will have to look carefully, only when commanded do short pulses occur. Cant imagine a pic showing much about IC1 as I use a jewelers loupe to see it *at all*, with the loupe about 2 cm away !! But I see no bridges and it looks good to me.
A decent camera can be a lot better than a loupe. What is this "chip reset" I have seen mentioned here on the blog? Is that something more than on and off?
How about page 140 of the A4 manual?? 73 Alan G4ZFQ
|
Would seem I need a menu on the screen to get to that reset item ? Wonder if there is a simpler way, i.e. a pullup resistor to ground on the reset terminal? If that causes it all to go I certainly wont do it.?
|
Would seem I need a menu on the screen to get to that reset item ? Cap, Sorry, I was forgetting your problem. Let's look at your first post.. It seems the only answer is here I guess you've already done that a few times. 2 rows of blocks - no Si5351 communication. 73 Alan G4ZFQ
|
Hi Cap,
Reading Back....
Your unit originally worked OK?
You did 3 mods, and then on it has two rows of solid squares on the
LCD?
TEMPORARY reverse ALL the modifications
you have made.
what value are the replacement chokes you are using?
(If they have a too high DC resistance they may make the Unreliable
microcontroller start-up at power-on problem worse?)
Please replace with the originals or jumper with a piece of wire and
try again?
Its helpful if you can keep the past replies to your thread attached
so others can follow the progress of the thread.
regards Alan
On 02/02/2018 07:13, Alan G4ZFQ wrote:
toggle quoted message
Show quoted text
Would seem I need a menu on the screen to
get to that reset item ?
Cap,
Sorry, I was forgetting your problem. Let's look at your first
post..
It seems the only answer is here
I guess you've
already done that a few times.
2 rows of blocks - no Si5351 communication.
73 Alan G4ZFQ
|
jjpurdum
Alan de G1FXB