Page 1 of 1

Repair tips

PostPosted: 27 Nov 2011 21:25
by OH2LH
Hello to all JUMA people,

I have been assembling several JUMAs lately and I have found following problem: There has been several short circuits
in PCBs especially TX136 and TX500 and PA-100 control boards. Also one TX 500 main board had short circuit in +14V line. It was the most fatal case since short was so strong that about 1cm of PCB foil was destroyed totally. Last week I found a short in one TX136 control board´s +5 V line. Fortunatelly +5V regulator had protection and nothing was broken. Shorts have been some times very hard to detect since they are almost invisible, very thin copper track under protection laquer. You need good magnifying glass to see those shorts.
So before you connect power to any JUMAs check the +14V and +5V lines against GND with multimeter in resistance measurement range.
Couple of shorts have been in some control line like relay control line and one was in I2C clock line.

The shorts are result of poor cleaning and particle contamination during the PCB manufacturing process.

Let the solder smoke!! Build JUMAs!

73 de Heikki OH2LH

Repair tips

PostPosted: 20 Jan 2012 23:37
by OH2LH
Hello JUMA people!!

Just today (yesterday now) got one TX 136 from JA that was faulty (newly built, never working).

Symptoms: When pressing ON/OFF unit didn´t start,only display back light was lit and upper row of the display showed pixels about half brightness.Releasing ON/OFF button, the unit switched OFF. Sound like CPU problem.

Faults:
1.Since it is not starting (the CPU) i checked first visually CPU pins so that there are no short circuits. OK
Next I checked operating voltages when pressing ON/OFF continuously, OK
Next I checked from CPU crystal pins with oscilloscope that it was oscillating when pressed continuously ON/OFF. OK
Next I discussed with Matti OH7SV about idea that if the CPU was not programmed. Matti advised to check the CPU´s MCLR signal state.
When pressing ON/OFF the MCLR should be high (+5V). Now it was low (0V). So something was dragging to low. Measurement with resistance measurement showed that the MCLR line was short to GND. With magnifying glass I found the short: Programming connector´s (ICD2) pin 1 (located under display) was short to GND since there was too much solder at the pin and the there was some marks that the soldering mask was little bit broken (maybe damaged during the assembly).Removed extra tin and powered ON unit it was OK.

2. During the testing I noticed that the display´s contrast adjustment was not working. Checking the PWM output (contrast is adjusted by changing the PWM duty cycle) from CPU (pin 63 ) it was ok. Then at the input of the IC3-A pin 3 nothing, GND... again short? Yep, similar short as in the first case now R17´s upper end was shorted to GND same way (too much tin).After removing extra tin it was OK.

Testing continues later, now to sleep

73´s de Heikki OH2LH

Re: Repair tips

PostPosted: 05 Feb 2012 18:58
by OH2LH
Hello all JUMA people!!

Just repaired one JUMA PA-100D control board. The problem was quite difficult to find: a short on PCB UNDER the CPU.
CPU pin 25 and GND was shorted just under CPU, feed through hole´s colar had tiny piece of foil connecting them together.

HINT nr1 : before one starts to assemble the boards check all different voltage lines against GND with resistance meter to find possible shorts.
It is much easier before than after installation of the components.

HINT nr2:
Do not use excessive amount of solder when soldering SMD components. couple of shorts were found from one TX136 last week due to the too much of solder tin.

73 de Heikki OH2LH