3
\$\begingroup\$

good board - back

Good board - bottom view


good board

Good board - top view


bad board - back

Bad board - bottom view


Bad board top view

Bad board - top view


I am new here and mainly a hobbyist when it comes to electronics. I work more with electricity than electronics.

Sorry about the Spanish comments on the schematic. I did that when I sent it to the original designer of the circuit.

Please I need some answers here.

I have three identical circuit boards all with the same circuit configuration.

Now two work perfectly and one does not.

On the one that does not work, I have replaced 99% of the components and still no joy. See the diagram.

When I test the voltages on pin 2 and 3 of the LM3900 in the circuit, the output changes but is still unstable.

However the ICs on the good boards are stable.

I did not design this circuit, I am only trying to fix the board.

I have replaced the IC more than once and still get the same issue with the output 2. It should be that output 2 is off and so is output 1 until the supply voltage changes either going higher or lower.

Adjusting P1 and P2 does not help in this case.

HP REG This is an edited copy of the whole schematic. Some connections are not shown except those pertaining to the op-amp

This is an edited copy of the whole schematic. Some connections are not shown except those pertaining to the op-amp

The voltages shown are for a good board in green and the bad board in red.

Remember the IC is new, and I have tested the bad board with at least 7 new chips and get the same results. To me, something is really wrong with the design. I have checked the bad board all over for shorted connections and found nothing.

If needed, I can post the full diagram. I used Autotrax dex as my editor so I need to export the schematic as a Jpeg etc. It would be nice if pdf files were allowed.

\$\endgroup\$
  • \$\begingroup\$ You should point out which components you have replaced(testing the new ones for correct values in the process of course), or which you haven't if that's easier. Also including pictures of the bad board may help someone spot something, like a layout concern that will only cause 33% of fabbed boards to fail. \$\endgroup\$ – K H Mar 15 at 22:35
  • \$\begingroup\$ Ok I will add the pictures of the front and back \$\endgroup\$ – Edmund Campbell Mar 15 at 22:44
  • \$\begingroup\$ Maybe throw in pics of the good board while you're at it. I assumed you metered or double checked the color code on every resistor and diode as you replaced it and replaced only with matching components, but if you got a resistor with the wrong magnitude band or a diode backwards someone might notice that too. \$\endgroup\$ – K H Mar 15 at 22:50
  • 1
    \$\begingroup\$ Am I correct that there is no bypass capacitor on the power leads to the op-amp? Always, always, always use bypass caps -- 100nF from each power pin to ground is the standard (and correct, 99% of the time). Keep the leads as short as possible. \$\endgroup\$ – TimWescott Mar 15 at 23:11
  • 1
    \$\begingroup\$ Do you have an oscilloscope? What leads you to believe it's oscillating and not doing something else bad? If you have an O-scope, please post pictures of the scope traces for an op-amps output and inputs. \$\endgroup\$ – TimWescott Mar 15 at 23:13
0
\$\begingroup\$

The circuit sort of looks okay. Personally, I think the designer is running far too much current into the Norton amp inputs.

Regardless, I would look at the following possible problem points:

1) One possible cause of your problems is leakage currents on the board. If the chip is socketed, remove the socket. If the chip is soldered directly into the board, remove the chip. Scrub the board thoroughly both top and bottom.

Use a meter set to the highest resistance range that it has and probe between each pin of the chip and the bare board in-between the pins. You should see no signs of conductivity. Reinstall the socket / chip and test again.

2) The voltage on pin 2 looks way low. It should be near 1 Vbe drop (0.6 - 0.7 Vdc) above the chip Gnd pin.

See of that pin voltage changes value as you vary the position of the wiper on P12. If it doesn't change, see if P12 is defective. You can check this by metering the voltage right on the wiper and varying the wiper position.

\$\endgroup\$
  • \$\begingroup\$ I have just tried the board on another phase unit that had a good board and it works as it should so I think the problem is the transformer that supplies low voltage to the board. I will confirm that now. The designer told me it was a 24 volt transformer but I have a feeling it should be lower. Need to confirm the voltage and see what I get. Then I will let you all know. It is amazing how such a simple thing can cause such a great headache, and believe me I have been fighting with this for over two weeks now. \$\endgroup\$ – Edmund Campbell Mar 16 at 1:25
  • \$\begingroup\$ Ok it appears that I have found the problem. The transformer that supplies the bad board had to be replaced and I bought a new 24 volt transformer. However its output voltage is 31 volts and the original one put out 27 volts. So on the other side of the rectifier I get 20.35 v dc from the new transformer and 17.25 from the original one. What I now need to do is to calculate a new value for R24 so I get the same voltage drop across it to get it equal to the other boards. \$\endgroup\$ – Edmund Campbell Mar 16 at 1:49
  • \$\begingroup\$ So it seems I need to just add an 18 ohm resistor is series with R24. Across R24 which is a 120 ohm resistor on the original transformer I get a voltage drop of 10.62 volts so using Ohms law I = V/R:- I get a current of 88.5 mA. Now with the new transformer I get a voltage drop of 12.6 volts but my current is 100mA so to balance get the voltage drop I need to bring the voltage down to 6.6 volts I need to drop the voltage by 13.75 volts at 100mA I need a 137.75 ohm resistor, so to get close I just add an 18 Ohm resistor in series and hopefully Bobs my uncle and it will work. \$\endgroup\$ – Edmund Campbell Mar 16 at 2:20
0
\$\begingroup\$

Just to let everyone know, the issue has been solved. I had to increase the value of R24 and that was it. First for some reason I put the value of R24 as 120 ohms but it was supposed to be 10 kilo ohms. I only had to increase its value to 12 kilo ohms and everything worked as it should. Thanks for all the input.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.