I don't know where to search next. I have examined all PCB with macro lens and
there is no defects. PIC's can't work neither on main PCB, nor on expansion
board. But why then Atmel works perfectly? This is very strange.
Interesting fact - when i put my arm over the power regulator conection with
PC automatically breaks down. I suppose its because of inductor. I can't find
resistor type and used standart one (vertical).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Atmel chips don't require the HV regulator, so the problem should be there;
check again your circuit, look for mounting errors like
transistors/diodes/capacitors in reverse, resistor value, etc.
There is absolutely no point in trying to program when the hardware test
fails; it will never work.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
2 kasamiko. Yes I tried. PIC16F84A. Voltages are present only for a second,
then they just disappearing. GUI writes or error writing adress, or Connection
timeout.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
So it should be 4,3 - 4,7V. When I starting H/W test, FB is 4,37V, then in
standby mode it is 1,7V. Is it correct? VPP=12,34V, VDD=4,98V; CK, D, PGM
outputs 0,01 - 0,3V in read/write mode. H/W test on step 2 disconnects
programmer from PC. VDD and VPP becomes set to 4,9V. Is it a reset process or
bug in main micro?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Sorry, forgot to measure inverted polarity("+" on the testpoint "-" on the
VDD). There is a stable signal -4,93V on D, CK and PGM outputs. But if I
right, this signals have to be positive.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
1.7V is right if the regulator is turned off, i.e. when you plug the
programmer.
You should measure voltages with respect to GND.
I suspect that your inductor is far too low in value; please check again, it
should be around 100 uH.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I used 100 uH, then tried 220uH(resistor type, buyed day ago). When I used
220uH the output voltage was around 18V(only in reading or writing processes.
Voltage in H/W test was 13.1 max). When used 100 uH I have 12,3(in read/write
process) up to 12,7 in standby(when regulator is on, but Q1 closed). Yesterday
I had 18V again on 100uH(once, next 10 times it was normal). I suppose main
PIC is defective, or firmware error(but why ATMEL works?). Tried to read a
reserve PIC18F2550 with OP firmware in it - Connection timeout. I'm in
despair.
Used parts: resistors 5%, transistors - USSR analogs KT3102 and KT3107,
Inductor - 100uH(cylinder type) or 220uH(resistor type), Quartz -
12Mhz(12000000000006). C1 and C6 - electrolityc on 50V, D3 - 1N5822, Other
capacitors - ceramic 10%. PCB - homemade.
SW1 - for what? Is it resets main micro?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
100uH is better because it saturates less; what is unexpected is that the HV
voltage varies, the regulator should keep it constant.
Measure both VPP and VPPU; there is no stdby, VPP is at 13V when the dcdc is
turned on, VPPU is connected to VPP when entering pic program mode or during
some test steps; a short on vppu could disrupt operation and disconnect the
usb entirely.
Since you used different transistors you should make sure to connect emitters
to emitters etc and don't take a pnp for an npn; check many times the pinout
(is usully in reverse) or better yet don't trust drawings and measure be and
bc drop wit a multimeter; be junction has a higher drop.
C1 has the right orientation and value?
If you have an oscilloscope check the base of q4: it should be switching
during the test; disconnect D3 and verify that q4 collector goes up and down.
The probability of a defective main pic is almost zero: it would have a broken
RC2, but in that case no HV voltage could be generated.
SW1 is not used at the moment.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Transistors orientation and pinout are exact as on BC557 and BR547. So there
is no mistake.
C1 has the right orientation, value is 100uF 10% ( a little bit higher - 107
uF). Oscilloscope will be only in the evening, so I'll check it later.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I can see power-up transient on collector, but I used time 0,2 us instead of
50us. I can't see the form of the signal is base of Q4, it's too fast for my
oscilloscope.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Sorry for the delay.
The DCDC drive signal is at 90kHz, or about 11us, so you should see it clearly
on any type of oscilloscope ever made. Look for a rectangular wave and then
see what's on the collector.
Post a picture please.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
This is definitely not right.
Do the following:
1: disconnect R10 and look again at RC2, it has to be a rectangular signal at
90 kHz
2: check the value of R10, should be 10k
3: try with R10 and Q4 but no diode and no L1
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
To me it looks like your Q4 is a PNP. When the base is low also the collector
goes low; but it may also be a capacitive effect; try a 10k resistor in place
of L1, still without diode. Also please increase the volt/div and shoot a
little closer to the scope.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
No problem, I just have little time lately.
If the volts/div are correct there is some problem with the power supply, it
can't be 5V.
Anyways, even without inductor the signal is not very rectangular on the base
and collector.
Try to see if VCC is stable at 5V on the 18F2550, and if GND is well
connected; I mean, no cold solder and no resistive shorts.
Q4 seems an NPN.
Try to remove all transistors except Q4.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I don't know where to search next. I have examined all PCB with macro lens and
there is no defects. PIC's can't work neither on main PCB, nor on expansion
board. But why then Atmel works perfectly? This is very strange.
Interesting fact - when i put my arm over the power regulator conection with
PC automatically breaks down. I suppose its because of inductor. I can't find
resistor type and used standart one (vertical).
That's a weird problem..have you tried programming a single PIC like 16F628A
while monitoring the programming voltage?
Atmel chips don't require the HV regulator, so the problem should be there;
check again your circuit, look for mounting errors like
transistors/diodes/capacitors in reverse, resistor value, etc.
There is absolutely no point in trying to program when the hardware test
fails; it will never work.
2 kasamiko. Yes I tried. PIC16F84A. Voltages are present only for a second,
then they just disappearing. GUI writes or error writing adress, or Connection
timeout.
By the way. What voltage should be on FB contact? (2 pin of 18F2550)
VFB=VPP*12/34
So it should be 4,3 - 4,7V. When I starting H/W test, FB is 4,37V, then in
standby mode it is 1,7V. Is it correct? VPP=12,34V, VDD=4,98V; CK, D, PGM
outputs 0,01 - 0,3V in read/write mode. H/W test on step 2 disconnects
programmer from PC. VDD and VPP becomes set to 4,9V. Is it a reset process or
bug in main micro?
Sorry, forgot to measure inverted polarity("+" on the testpoint "-" on the
VDD). There is a stable signal -4,93V on D, CK and PGM outputs. But if I
right, this signals have to be positive.
1.7V is right if the regulator is turned off, i.e. when you plug the
programmer.
You should measure voltages with respect to GND.
I suspect that your inductor is far too low in value; please check again, it
should be around 100 uH.
I used 100 uH, then tried 220uH(resistor type, buyed day ago). When I used
220uH the output voltage was around 18V(only in reading or writing processes.
Voltage in H/W test was 13.1 max). When used 100 uH I have 12,3(in read/write
process) up to 12,7 in standby(when regulator is on, but Q1 closed). Yesterday
I had 18V again on 100uH(once, next 10 times it was normal). I suppose main
PIC is defective, or firmware error(but why ATMEL works?). Tried to read a
reserve PIC18F2550 with OP firmware in it - Connection timeout. I'm in
despair.
Used parts: resistors 5%, transistors - USSR analogs KT3102 and KT3107,
Inductor - 100uH(cylinder type) or 220uH(resistor type), Quartz -
12Mhz(12000000000006). C1 and C6 - electrolityc on 50V, D3 - 1N5822, Other
capacitors - ceramic 10%. PCB - homemade.
SW1 - for what? Is it resets main micro?
100uH is better because it saturates less; what is unexpected is that the HV
voltage varies, the regulator should keep it constant.
Measure both VPP and VPPU; there is no stdby, VPP is at 13V when the dcdc is
turned on, VPPU is connected to VPP when entering pic program mode or during
some test steps; a short on vppu could disrupt operation and disconnect the
usb entirely.
Since you used different transistors you should make sure to connect emitters
to emitters etc and don't take a pnp for an npn; check many times the pinout
(is usully in reverse) or better yet don't trust drawings and measure be and
bc drop wit a multimeter; be junction has a higher drop.
C1 has the right orientation and value?
If you have an oscilloscope check the base of q4: it should be switching
during the test; disconnect D3 and verify that q4 collector goes up and down.
The probability of a defective main pic is almost zero: it would have a broken
RC2, but in that case no HV voltage could be generated.
SW1 is not used at the moment.
Transistors orientation and pinout are exact as on BC557 and BR547. So there
is no mistake.
C1 has the right orientation, value is 100uF 10% ( a little bit higher - 107
uF). Oscilloscope will be only in the evening, so I'll check it later.
I can see power-up transient on collector, but I used time 0,2 us instead of
50us. I can't see the form of the signal is base of Q4, it's too fast for my
oscilloscope.
Sorry for the delay.
The DCDC drive signal is at 90kHz, or about 11us, so you should see it clearly
on any type of oscilloscope ever made. Look for a rectangular wave and then
see what's on the collector.
Post a picture please.
I have this signal on the base of Q4 (0,5 us)
Same with 1 us
And this I have before R10 (right on RC2 of 18F2550)
I guess that is not right. Is it because of resistor?
P.S. Sorry for unprofessional answer. I'm just learning.
This is definitely not right.
Do the following:
1: disconnect R10 and look again at RC2, it has to be a rectangular signal at
90 kHz
2: check the value of R10, should be 10k
3: try with R10 and Q4 but no diode and no L1
Results:
1. Got it. Clear rectangular signal.
2. 9,92 kOhm
3. Now I have this on the base of Q4
And this is on the collector
Sorry... Wrong second picture. This is right collector signal...
To me it looks like your Q4 is a PNP. When the base is low also the collector
goes low; but it may also be a capacitive effect; try a 10k resistor in place
of L1, still without diode. Also please increase the volt/div and shoot a
little closer to the scope.
Replaced L1 with 10k resistor. And have the following signals (time 2 us/div)
On the base of Q4 (0,1 V/div)
On the collector (0,5 V/div)
I used 1N5189 for D3. I also encountered some problem on this section C1
should NOT be lower than 25VDC caps...
I used 40V 100uF capacitor C1 and 1N5822 shottky D3.
Sorry for annoying, but does this oscillograms mean that I have a deffective
inductor?
No problem, I just have little time lately.
If the volts/div are correct there is some problem with the power supply, it
can't be 5V.
Anyways, even without inductor the signal is not very rectangular on the base
and collector.
Try to see if VCC is stable at 5V on the 18F2550, and if GND is well
connected; I mean, no cold solder and no resistive shorts.
Q4 seems an NPN.
Try to remove all transistors except Q4.