| Electronics > Repair |
| Question about Dual supply power supplies and shorting |
| (1/3) > >> |
| shockpoint:
Hi all, I have a few general questions hoping someone can shed light on them. I am repairing an old digital piano. Initially the problem was thought to be on the amplifier board that takes +12 and -12V to power the op amps. However, thorough circuit tracing did not show any glaring abnormalities. So my search led me to the digital pianos mainboard, where the +12V and -12V feeds to a network of 5 LC833 op amps The reason I suspect the mainboard is because I found that by disconnecting everything else, reconnecting just the amplifier board and the mainboard's signal/voltage cable caused the same fuses to blow (which is the original condition it came in). I depinned the JST connector and disconnected the +12V, -12V and GND pins and retested, and the new fuses I put in stayed intact. So that is highly suggestive that the fault is in the mainboard. Unfortunately I do not have any schematics for the mainboard as Kawai is rather protective of its proprietary devices I guess So I have a few questions that are more conceptual/general in nature 1. Between the -12V and +12v rails on the main board, I have measured 128.2 Ohms with a DMM. This resistance is not seen on the amplifier board which has 5 other op amps, when the mainboard is disconnected from the amplifier. On the amplifier board, the 12v and -12v rails have a resistance in the 3.8k+ region. Does the 128.2Ohms sound suspicious to those who are experienced in op amp fault finding? 2. the 128.2 ohm also tests as a voltage drop of 0.041V in diode mode when I test either between the -12V and +12V pins, or on the opamp -Vcc and +Vcc pins; I have not seen such a small voltage drop before across op amp power rails and in my own personal experience this is edging towards a short 3. What happens when you connect +12V to -12V? Logically I think it shorts just like if you connected +12V to GND but just wanted to confirm. 4. Is it possible that the 128.2ohms demonstrates the 'wrong side' of the short on the shorted device? by that i mean if the short is on the other side of the SMD I am trying to find, then if I measure on the incoming voltage side or the outgoing voltage side my short may either be 0.001ohms (a true short) or 128.2ohms (not technically a short but only because im measuring in the wrong place). My next step is to begin voltage injection to try and find the shorted SMD because the op amps are in parallel. The other option would be to use a milliohmmeter. |
| Harry_22:
Hi! Try the following: in diode mode check the drop voltage between the +12V buses and the GND, then between the GND and -12V. Compare the values. A short circuit on the power bus is very easy to find if you have a thermal imager. To do this set the laboratory source to, for example, 1V with a current limit of 1A (the values can be adjusted based on further results). Connect it to the buses in correct polarity. Next look for the heat release point. All other methods including searching with a milliohmmeter are more labor-intensive. |
| shockpoint:
--- Quote from: Harry_22 on September 21, 2024, 08:40:38 pm ---Hi! Try the following: in diode mode check the drop voltage between the +12V buses and the GND, then between the GND and -12V. Compare the values. A short circuit on the power bus is very easy to find if you have a thermal imager. To do this set the laboratory source to, for example, 1V with a current limit of 1A (the values can be adjusted based on further results). Connect it to the buses in correct polarity. Next look for the heat release point. All other methods including searching with a milliohmmeter are more labor-intensive. --- End quote --- Hi, thanks for your reply The diode mode between +12V and GND, and -12V and GND compare similarly and they are: 0.565 and 0.537 respectively. Should these values be closer together? With voltage injection, is it voltage set too high that will kill a given IC or is it the current? If I keep the voltage at 1V, with this board having inputs of +12 and -12V, and increase to 2A-3A (in case 1A doesn't get the chips warm enough to show up on the thermal imager), is there a risk of damage by current to the ICs? I have checked though and it doesn't seem that the 12 nor the -12 go to any of the ICs directly. +12V goes to 78LM05 (voltage reg) before it goes to the ICs and there is no short on the voltage regulator |
| Harry_22:
Hi! Seems you have short only between +12V and -12V rails, or between the opamp +Vcc and -Vcc pins. In this case connect plus of laboratory PS to +Vcc and minus to -Vcc for test. One volt limit usually sufficient to detect the short line. You will see that your PS will go into the current limitation mode with around zero volt at output. This is absolutely safe for your device under test. But please start from lowest current limit, for instance 1A to avoid restoring the short circuit. Thermal imager is very sensitive. Below you can see the termal image of leaked capacitor 4mA at 12V. |
| shockpoint:
--- Quote from: Harry_22 on September 22, 2024, 08:36:47 pm ---Hi! Seems you have short only between +12V and -12V rails, or between the opamp +Vcc and -Vcc pins. In this case connect plus of laboratory PS to +Vcc and minus to -Vcc for test. One volt limit usually sufficient to detect the short line. You will see that your PS will go into the current limitation mode with around zero volt at output. This is absolutely safe for your device under test. But please start from lowest current limit, for instance 1A to avoid restoring the short circuit. Thermal imager is very sensitive. Below you can see the termal image of leaked capacitor 4mA at 12V. --- End quote --- Thanks again for your insight. I will test it as you have suggested and post the results. A general question I have about +12V and -12V rails. I noticed you mentioned plus of the Lab PSU to the +12V and minus to the -12V. That makes sense as I am trying to see which component 'completes' my short circuit. But would there be any value in purchasing a power supply that produces both positive and negative voltages that have isolated grounds from each other? such as the HP 3631A? |
| Navigation |
| Message Index |
| Next page |