|
I noticed that the display contrast changes quite strongly (characters becoming dark) when transmitting, causing the display to flicker when sending CW. Is this a general problem, or is there something wrong with my unit?
Otherwise vy happy with the little rig, amazing design!
vy73, Peter HB9TVK
|
Flickering of the backlight brightness is caused by voltage drop in
the +12 power supply rail. In the QCX design the LCD backlight is
powered from the +12 rail through a resistor rather than from the +5
regulated supply; see the partial circuit diagram detail of the
microprocessor on page 115 and the explanation of why it is wired that
way on page 116. The input voltage will always drop on key down
because of the increased current demand; only a little if you're using
a bench power supply (because of resistive losses in the wires and
increased voltage drop in the reverse polarity blocking diode),
possibly a lot if you are powering the rig with batteries because the
output of the batteries will also sag. That effect is easy to see, and
it's not unique to the QCX; the backlight on the display of my K2 also
does it.
But that should not cause the contrast to change. The LCD panel itself
(other than the backlight) is powered by the +5 regulated supply, and
the display contrast setting is also derived from that supply through
a resistive divider (trimmer R47). If you are running the rig from a
low supply voltage it's possible that the 7805 regulator is falling
out of regulation, causing the +5 rail to vary excessively; normally
the regulator should keep it steady. You could also be experiencing RF
feedback that affects the operation of the regulator.
The first thing to try is to power the rig from a stiffer source:
either a good bench supply or a large battery (one with multiple amp
hours of capacity), with an input voltage to the rig of 10V or more.
If that doesn't clear up your display flicker you have a problem
somewhere; a significant change of contrast is not normal. If you have
an oscilloscope handy, probe the +5 rail and observe the voltage as
you key the transmitter; you should see only very small changes.
You're likely to observe a quick transient drop on key down and a
quick rise on key up if you're looking at times in the microseconds or
nanoseconds, and then the regulator quickly brings it back where it's
supposed to be. Those microsecond spikes are too short for you to see
the effect on the display.
toggle quoted message
Show quoted text
On Sun, Oct 29, 2017 at 12:33 PM, <peter@...> wrote: I noticed that the display contrast changes quite strongly (characters
becoming dark) when transmitting, causing the display to flicker when
sending CW. Is this a general problem, or is there something wrong with my
unit?
Otherwise vy happy with the little rig, amazing design!
vy73, Peter HB9TVK
|
LCD Display flicker and contrast change is something that can be evaluated with
a simple test platform made up of an LCD and an Arduino.? Run the "Hello World"
program from Arduino IDE examples to get some text on the display.?
Experiments here seem to indicate that use of a potentiometer for setting display
contrast may not be necessary.? The more modern 1602 style displays seem to
work fine with just a 2.2K resistor from pin-3 (the contrast control pin) to ground.
If that is acceptable for your application then it will save the cost and board space
of the potentiometer.?
Drive to the LCD backlight can be made adjustable by using the Arduino PWM
output to drive the base of a emitter follower pass transistor between +5 volts and
the LCD backlight pin.? This then becomes a variable that can be set by software.
A more esoteric solution might involve a light sensor driving an ADC input to the
Arduino and software for automatically adjusting LCD backlight based on ambient
light intensity.? Put your finger over the sensor if you temporarily need higher
backlight output.?
Arv? K7HKL
On Sun, Oct 29, 2017 at 11:31 AM, Shirley Dulcey KE1L <mark@...> wrote: Flickering of the backlight brightness is caused by voltage drop in
the +12 power supply rail. In the QCX design the LCD backlight is
powered from the +12 rail through a resistor rather than from the +5
regulated supply; see the partial circuit diagram detail of the
microprocessor on page 115 and the explanation of why it is wired that
way on page 116. The input voltage will always drop on key down
because of the increased current demand; only a little if you're using
a bench power supply (because of resistive losses in the wires and
increased voltage drop in the reverse polarity blocking diode),
possibly a lot if you are powering the rig with batteries because the
output of the batteries will also sag. That effect is easy to see, and
it's not unique to the QCX; the backlight on the display of my K2 also
does it.
But that should not cause the contrast to change. The LCD panel itself
(other than the backlight) is powered by the +5 regulated supply, and
the display contrast setting is also derived from that supply through
a resistive divider (trimmer R47). If you are running the rig from a
low supply voltage it's possible that the 7805 regulator is falling
out of regulation, causing the +5 rail to vary excessively; normally
the regulator should keep it steady. You could also be experiencing RF
feedback that affects the operation of the regulator.
The first thing to try is to power the rig from a stiffer source:
either a good bench supply or a large battery (one with multiple amp
hours of capacity), with an input voltage to the rig of 10V or more.
If that doesn't clear up your display flicker you have a problem
somewhere; a significant change of contrast is not normal. If you have
an oscilloscope handy, probe the +5 rail and observe the voltage as
you key the transmitter; you should see only very small changes.
You're likely to observe a quick transient drop on key down and a
quick rise on key up if you're looking at times in the microseconds or
nanoseconds, and then the regulator quickly brings it back where it's
supposed to be. Those microsecond spikes are too short for you to see
the effect on the display.
On Sun, Oct 29, 2017 at 12:33 PM,? <peter@...> wrote:
> I noticed that the display contrast changes quite strongly (characters
> becoming dark) when transmitting, causing the display to flicker when
> sending CW. Is this a general problem, or is there something wrong with my
> unit?
>
> Otherwise vy happy with the little rig, amazing design!
>
> vy73, Peter HB9TVK
>
|
Using an AVR output for variable backlight control is a neat idea. But
I think it's impossible to do on the QCX because there aren't any pins
left - every I/O pin on the microcontroller is already in use.
There might be a need for that contrast control that doesn't show up
in bench testing. LCD contrast varies somewhat with temperature, so if
you take the rig into the field on a very hot or very cold day you may
need to adjust the trimmer.
toggle quoted message
Show quoted text
On Sun, Oct 29, 2017 at 1:48 PM, Arv Evans <arvid.evans@...> wrote: LCD Display flicker and contrast change is something that can be evaluated
with
a simple test platform made up of an LCD and an Arduino. Run the "Hello
World"
program from Arduino IDE examples to get some text on the display.
Experiments here seem to indicate that use of a potentiometer for setting
display
contrast may not be necessary. The more modern 1602 style displays seem to
work fine with just a 2.2K resistor from pin-3 (the contrast control pin) to
ground.
If that is acceptable for your application then it will save the cost and
board space
of the potentiometer.
Drive to the LCD backlight can be made adjustable by using the Arduino PWM
output to drive the base of a emitter follower pass transistor between +5
volts and
the LCD backlight pin. This then becomes a variable that can be set by
software.
A more esoteric solution might involve a light sensor driving an ADC input
to the
Arduino and software for automatically adjusting LCD backlight based on
ambient
light intensity. Put your finger over the sensor if you temporarily need
higher
backlight output.
Arv K7HKL
On Sun, Oct 29, 2017 at 11:31 AM, Shirley Dulcey KE1L <mark@...>
wrote:
Flickering of the backlight brightness is caused by voltage drop in
the +12 power supply rail. In the QCX design the LCD backlight is
powered from the +12 rail through a resistor rather than from the +5
regulated supply; see the partial circuit diagram detail of the
microprocessor on page 115 and the explanation of why it is wired that
way on page 116. The input voltage will always drop on key down
because of the increased current demand; only a little if you're using
a bench power supply (because of resistive losses in the wires and
increased voltage drop in the reverse polarity blocking diode),
possibly a lot if you are powering the rig with batteries because the
output of the batteries will also sag. That effect is easy to see, and
it's not unique to the QCX; the backlight on the display of my K2 also
does it.
But that should not cause the contrast to change. The LCD panel itself
(other than the backlight) is powered by the +5 regulated supply, and
the display contrast setting is also derived from that supply through
a resistive divider (trimmer R47). If you are running the rig from a
low supply voltage it's possible that the 7805 regulator is falling
out of regulation, causing the +5 rail to vary excessively; normally
the regulator should keep it steady. You could also be experiencing RF
feedback that affects the operation of the regulator.
The first thing to try is to power the rig from a stiffer source:
either a good bench supply or a large battery (one with multiple amp
hours of capacity), with an input voltage to the rig of 10V or more.
If that doesn't clear up your display flicker you have a problem
somewhere; a significant change of contrast is not normal. If you have
an oscilloscope handy, probe the +5 rail and observe the voltage as
you key the transmitter; you should see only very small changes.
You're likely to observe a quick transient drop on key down and a
quick rise on key up if you're looking at times in the microseconds or
nanoseconds, and then the regulator quickly brings it back where it's
supposed to be. Those microsecond spikes are too short for you to see
the effect on the display.
On Sun, Oct 29, 2017 at 12:33 PM, <peter@...> wrote:
I noticed that the display contrast changes quite strongly (characters
becoming dark) when transmitting, causing the display to flicker when
sending CW. Is this a general problem, or is there something wrong with
my
unit?
Otherwise vy happy with the little rig, amazing design!
vy73, Peter HB9TVK
|
Thanks Shirley for your detailed suggestions!
It is indeed the contrast, not the backlight that is flickering. The +5V rail remains stable during key down (4.830V key up, 4.832V key down, measured directly at the LCD connector). The input to the contrast pin (VO) however rises almost 150mV (from 1.047 to 1.193V) upon key down, thus changing the contrast.
I checked with the oscilloscope and I see some serious HF leaking into VO: Instead of 1V DC, I'm getting a 7 MHz (it's a 40m kit) sine between 0.5 and 2V (1.5Vpp) on the VO input:
?
Any ideas? I'll check again all soldering, maybe replacing the pot by a fixed resistor...
vy73, Peter HB9TVK
|
Just like mine peter. when transmitting, there was a few milli-volts
change on the contrast pin.? I modified the PCB by
1) replace the pot with a lower resistance? say 5k pot.
2) Cut the track connecting the +5v to the pot.
3) Solder a resistor between the +5V and the end of the pot that was
connected to 5V. The value should be 3 to 4 times the value of the
potentiometer.
This makes the pot much easier to adjust, and the lower value is
much more stable when changing from RX to TX.
73s??? Vin.
toggle quoted message
Show quoted text
I noticed that the display contrast changes quite strongly
(characters becoming dark) when transmitting, causing the display
to flicker when sending CW. Is this a general problem, or is there
something wrong with my unit?
Otherwise vy happy with the little rig, amazing design!
vy73, Peter HB9TVK
|
Thanks Peter, I knew that lowering the impedance of the contrast
control worked, but did not investigate why.? Now I know.
Thanks.???? 73?? Vin
toggle quoted message
Show quoted text
Thanks Shirley for your detailed suggestions!
It is indeed the contrast, not the backlight that is flickering.
The +5V rail remains stable during key down (4.830V key up, 4.832V
key down, measured directly at the LCD connector). The input to
the contrast pin (VO) however rises almost 150mV (from 1.047 to
1.193V) upon key down, thus changing the contrast.
I checked with the oscilloscope and I see some serious HF leaking
into VO: Instead of 1V DC, I'm getting a 7 MHz (it's a 40m kit)
sine between 0.5 and 2V (1.5Vpp) on the VO input:
?
Any ideas? I'll check again all soldering, maybe replacing the pot
by a fixed resistor...
vy73, Peter HB9TVK
|
Thanks Vin for your suggestion. I think I found the root cause in the meantime:
I tried lower values for the pot, and also Arv's suggestion to replace the pot with a 2K2 resistor to ground (this works very well btw.). Still massive contrast difference on key down. Filtering the RF with a parallel capacitor did not help, as there was still a DC offset.
I decided to mount the LCD away from the PCB, as I suspected it to pick up RF from the PA stage just below. Immediately, the problem went away. When I mounted the LCD back on, you guess it, the problem still was gone!
I checked the space between PCB and LCD, and noticed that a flap of the LCD frame was cutting into the windings of L2:
I assume this was the original cause for the problem. I bent the flap away some more to stay clear of L2.
Now I can enjoy my qcx even more! Thanks to all for your help and suggestions.
vy73, Peter HB9TVK
|
Hi Peter,
I ended up bending that same flap on mine. I did it because it hit the top of the toroid when I tried to sandwich the display down. Anyway glad you figured it out.
toggle quoted message
Show quoted text
On Oct 30, 2017, at 3:46 PM, peter@... wrote: Thanks Vin for your suggestion. I think I found the root cause in the meantime:
I tried lower values for the pot, and also Arv's suggestion to replace the pot with a 2K2 resistor to ground (this works very well btw.). Still massive contrast difference on key down. Filtering the RF with a parallel capacitor did not help, as there was still a DC offset.
I decided to mount the LCD away from the PCB, as I suspected it to pick up RF from the PA stage just below. Immediately, the problem went away. When I mounted the LCD back on, you guess it, the problem still was gone!
I checked the space between PCB and LCD, and noticed that a flap of the LCD frame was cutting into the windings of L2:
<Mon Oct 30 23_28_44.jpg>
I assume this was the original cause for the problem. I bent the flap away some more to stay clear of L2.
Now I can enjoy my qcx even more! Thanks to all for your help and suggestions.
vy73, Peter HB9TVK
|
On Mon, 2017-10-30 at 15:46 -0700, peter@... wrote: I assume this was the original cause for the problem. I bent the flap
away some more to stay clear of L2.
Now I can enjoy my qcx even more! Thanks to all for your help and
suggestions. I will be using some of these for my 30m QCX (when I get time to build it) I think they are intended for making Arduino shields. I've got various sizes of M3 standoffs to use to get a greater spacing between the LCD and the board. I might "retro-fit" some on my 17m QCX as it has the tall C47 capacitor problem. PeterO G0DZB
|
Shirley Dulcey KE1L
Joel Caulkins/N6ALT