| Electronics > Beginners |
| Bridge Rectifer and AC Ground relationship |
| (1/1) |
| jwhitmore:
Out of my depth as usual, when it comes to electronics, but there's a chance I might learn something. Somebody asked me to have a look at a fancy Italian designer light which don't work. I'm in Europe so it's 240V AC mains. There's a bit of scorch marking under the bridge rectifier chip, ( GBPC2508W ). The lights are powered on/off and dimmed with an Infra Red remote control, so there's an AT90 uC on the board which is obviously monitoring IR inputs and setting outputs of three lights via 3 "15J301" devices (an IGBT I think [1]) So to test what appears to be a dead board I put a multimeter across the power pins of the AT90 uC that gave me a reading of about 1.6 volts which is well below it's required power. I then measured the voltage between mains ground and the ground rail of the PCB, so the negative terminal of the full bridge rectifier and mains ground. I didn't expect a voltage between these to points. Perhaps I'm misunderstanding but I expected the two to be at a similar voltage level. I actually measured 130V. I'm now assuming that this device is toast and my best bet is to try decode the remote which still works and build a new PCB with a new uC. As the unit I have is dead I assume that the uC controls the brightness of the 3 outputs using PWM of the 15J301. The 15J301 appears to be a low side switch as the full bridge rectified voltage is set to the output lights. This looks like an old design as it's all through hole components there might be a better device for switching the output. I should try understand this board though as I can't see how power was ever delivered to the uC [1] http://www.northstardatasheet.com/datasheet/4097.pdf |
| Ian.M:
The board's negative rail will normally be bouncing between the Neutral potential (near ground in *most* EU countries) and approximately -340V (assuming 240V RMS supply), with a 50Hz half wave rectified waveform. The average of a half wave rectified sinewave is 0.318 of the peak, which would be 108V, and your observed 110V would be well within the tolerances to be expected due to mains voltage variation and neutral to ground offset. Watch the video: *before* attempting to scope *ANY* signals in the device! Without board photos, or a schematic its pointless to speculate how the MCU is powered. Low MCU Vcc voltage may be due to a failed supply circuit or an overload due to a failed IC drawing excessive current. Until you've determined whether or not the MCU is drawing excessive current, its premature to condemn the device. |
| 6PTsocket:
I am not that familiar with European power distribution but if there is a bridge rectifier, isn't everything after it referenced to the + and - output of the rectifier. Whey are we looking back at the mains, neutral, etc? I would assume a filter cap charged up to near the peak value of the AC into the bridge and take it from there. Obviously there is a means of votage reduction to feed the control circuitry unless it went through a transformer before the bridge. I must have missed something. Finding why the IC is only supplied with 1.6 volts seems like a straight forward power supply problem, no? Sent from my SM-G900V using Tapatalk |
| jwhitmore:
I'd assumed that the GBPC2508W [1] was a full bridge rectifier. If it's a half now I'm confused. The circuit half rectifies the AC to feed that to the AC 240V bulbs, that means that a 100W build will only give 50W output, as half the wave is chopped. Think I'm going to have to sit down with a jotter and try draw out the circuit. Mean time I'll try find an easier way to dim an AC bulb. I agree with the 'straight forward power supply problem' but at present it's resistors and a few caps, so have to trace it all out to see where the control voltage is coming from. [1] https://www.mouser.co.uk/datasheet/2/308/GBPC2508-1301123.pdf |
| Ian.M:
Ah, I think you have misunderstood me. Yes I did say that for a bridge rectifier: --- Quote from: Ian.M on November 28, 2018, 09:41:05 pm ---The board's negative rail will normally be bouncing between the Neutral potential ... and approximately -340V ... with a 50Hz half wave rectified waveform. --- End quote --- but (assuming the presence of a sizeable reservoir capacitor), the positive rail will also be 'bouncing' with a similar waveform maintaining a difference between V+ and V- (i.e. across the cap) of about 340V (less a bit of ripple due to the load current). If there's no reservoir cap, the positive rail waveform (with respect to Neutral) will be a similar, positive going half wave, with its peaks in the gaps in the negative going halfwave on V-. The difference between V+ and V- is then the full wave rectified waveform you would expect from a bridge rectifier. in both cases the steady DC V (-ripple) or the full wave rectified sine DC V cannot be measured with ground referenced test equipment, and if you attempt to do so, *something* is certain to go *BANG*! |
| Navigation |
| Message Index |