1) Check this contact very well. Pay attention if there is no burnout of the interlayer bushings in the printed circuit board.
2) Connect a 100 W light bulb as a load, only apply the load after turning on the device, and not before.
3) If the voltage subsided, or rather, decreased to 0, then the contact between the rack and the board is bad. Connect the lamp to the rack. Repeat measurements.
4) If the voltage sags, climb the oscillator onto the motherboard and check the signal on L6386: between 8 and 9 feet, between 12 and 13 feet.
a) If there are pulses, then this is not firmware or calibration. Two options
b) Both output diodes are broken. You can check with a tester.
c) The amplifier stage from the complementary pair on the module is faulty (decoupling L6386 with MOSFET keys)
Further, a huge complex branch for further troubleshooting of this machine. But I will not dump everything, I will listen to your measurements.
Black Fire wrote:
forget about the potential for failure of the power supply unit and whistling.
My second year is whistling robust after a slight fall. Is it going to die soon?
Exuberant wrote:
My second year is whistling robust after a slight fall. Is it going to die soon? 
We're talking about a different whistle. Inverter series 130/160/170 whistle slightly when off (features of the auxiliary power supply unit)
In your case - but who knows how much - maybe it will outlive us (you can't tell without looking)
as an electronics engineer, I would advise when reworking switching power supplies
use microcircuits that have the same operating frequency and are turned on
according to typical schemes from datasheets found for example on ">,
otherwise whistling, static chirping and output voltage jumps are possible
in case of interference during welding in such a converted unit with unpredictable consequences
for an expensive power unit,
for example, in the recommended replacement NCP1055ST100T3G-TNY278PG, the first microcircuit
operates at 100, and the second at 132kHz (like NCP1055ST136, operating at 132kHz),
hence the question is whether it works without problems
over the entire operating range of voltages and temperatures?
and another very important is the snubber chain (according to the datasheet), the limiter in Russian, connected in parallel
primary winding - the elements of this circuit operate in maximum mode
and it is desirable to replace them with high-quality and powerful ones, which manufacturers often do not do (do not even install!)!
electrolytic capacitors of rectifiers of secondary windings change to jamikon, green,
with low series resistance (LowESR), which will save you a bunch of additional problems,
rinse thoroughly from the flux and, after checking, cover the installation with heat-resistant varnish, even furniture,
the tsapon peels off in some hot places, although where it is not hot, it will
Basically, devices with defective microcircuits do not react in any way to attempts to turn on, but there were rare cases when Fubag IN160 inverter shutdown with warm-up, after a few seconds it turned on, then turned off again, and clicked. In this case, it turned out to check the microcircuit by turning on the device and, while clicking, gently leaning a cotton swab moistened with alcohol against it, the clicking immediately stopped and the inverter worked normally until the alcohol dried out.
Sometimes, after opening the device and external examination, such defects of the NCP1055B microcircuit are visible, one might say, with the naked eye.
Often the microcircuit breaks down, by itself, without affecting the vital surrounding elements. Something she has personal problems with the tightness of the case. In such cases, it is sufficient to replace only the microcircuit. We will consider this option.
First of all, we will clear the place from the killed microcircuit.
It makes no sense to put it in the SOT-223 case again, since they are unreliable. Better to somehow contrive. and receive the FEATHER BIRD. no set the microcircuit to DIP-8.
Comparing the conclusions of this microcircuit by the datasheet, we bend them in this way.
If everything is done correctly and the surrounding components are intact, which is better to be sure by using the circuit in advance and checking them, then the device turns on immediately and works great.
On the forums, you can find options for replacing the NCP1055B with the TNY268 with a slight alteration of the circuit.
remont22
Posted 02 April 2011 - 01:57
PWM NCP1055P100.
From another forum.
I fixed another deadhead, with an "improved" duty feeder on the NCP1055 that explodes all the time.
In general, this product can be exchanged for a TNY268 in a DIP-8 package. I soldered it onto a small breadboard, and connected it to the circuit with short conductors. How to do this is clear from the comparison of the documentation for these 2 mikruhi. The bypass kander (10mk) is changed to 0.1mk. A source.
Which option to choose is at your discretion.
Attention!
Be careful when repairing the welding inverter with your own hands, the responsibility lies with you.
Repair of welding inverters from Fubag and other manufacturers.
