Sounds like an opportunity to dive deep into a hardcore repair, if nothing else. This is where the fun starts.

Firstly, it's good to find out why it failed in the first place, so it doesn't happen again. My main suspicion is that the tank parameters are outside of what the unit expects. Since the capacitance is within the supported range of the unit, it could be that the inductance is too large for the given capacitor. Do you have a way to measure the resonant frequency of your tank circuit? It should ideally be well within the accepted 750 - 1150 kHz operating range of the unit. Also consider that loading the coil will modify the resonant frequency, generally upwards.

I had a look at the MOSFETs in my unit, but they also have the markings lasered off, so finding the correct part, or at least a suitable replacement, could be a bit tricky. The devices appear to be in a TO-264 package which rules out a lot of options, so that's a start. IXYS parts seem like a good candidate, as they are one of the few MOSFET manufacturers with a wide lineup in this package, part number prefix IXFK.

I'd bet more than one MOSFET is bad, at least the other one in the same bridge leg is likely toast. Also check the diodes if they are fine, the series blocking ones could have taken a beating but diodes tend to be pretty resilient so they could have survived. If the remaining two MOSFETs are fine, you can measure some parameters to help find a suitable replacement. A good start is to measure the Rdson (push an amp through the MOSFET when it's turned solidly on, and measure Vds). This along with the gate capacitance will allow you to get the sizing of the MOSFET right.

As far as performance goes, it looks like you have a lot of stray inductance between your capacitor and coil, maybe as much as in the work coil itself. This will lead to a lot of your tank VARs not being coupled to the workpiece.

Thankyou for the time with replying, sorry for such a delay in my response.

When this failure happened there was surging of the power been delivered to the work coil, this was apparent by the info displayed on the unit but before I could do anything the unit failed.

I do expect as you had said that the coil was outside the expected range / setup and this caused an error or the unit got mixed up during power delivery ; likely down to my poor tank setup.

I will at some point try and figure out the MOSFET needed and replace the full bridge ; Hopefully the gate driver side is all in working order ; I am still unsure why the pre charge power resistor failed the way it did , as if it had exploded as one end had become unattached to the PCB and the body is broken. This may be a bodge to replace as it sits behind the main filter capacitors.

Do you have any pointers on how to make this work head of this unit, I understand the need to keep stray inductance down as much as possible but I am unsure where to go with my current head. Power is feed by 2 pairs of litz wire and two twisted pairs for feedback. The leads to the work head are not twisted ; Would this make a difference ?

My other option and one I am leaning to is to eventually locate an original head and remake the bridge as then there should be no issues with the feed and tank head setup.

thanks again.

Thanks for the info here.

I had an attmept at making a custom head but it seems I ran into a failure . I built a head around a 200nF cap / Litz wire feed and such like the origional cable ; Head performace isnt great compared to the normal head.

My issue was everything was running fine, then the power was pulsing up and down ; crackle sound and a bang, upon looking a large resistor that is precharge from what I can tell had burnt out and after looking and tacking the bridge out ( Really nice its modular ) It seems one of the mosfets has failed and is split across its front. Other mosfets look intact but its hard to say what was taken with the faiure. I do hope it was the one mosfet.

They have the marking removed so I have no idea what mosfet to replace it with, any suggestions there ?

/>


 

Wanted to share a current project if anyone was interested.

I am very much the fan of old display tech and wanted to create a display with easy to acquire neon indicators so after some planning and trials these are the first full prototypes of my displays.
They consist of an 8*8 array of neon indicators each with its own opto-triac and resistor, the brains of the unit consists of an Arduino based 328P UNO controlling a MAX7219 on the back of each of the 8*8 displays, The Logic is 5V driven and the neons are driven from 120V 50Hz.

I am using premade libraries for the 7219 as my programming skills are quite poor.

The first version of these displays I mistakenly used zero cross based opto triacs and while this worked there was an unpleasant flicker due the multiplexed nature of the display, after switching to non zero cross triacs the outcome was a smooth illumination and displaying/scrolling of the display.

Its abit long winded with all the soldering so has taken some time to make these by hand but I am very happy with the outcome, the current build will be a 4 board display with a future 10 board display planed.

Thanks for looking ^

/>

Very nice van d'graph generator impressive output ^^

I was looking to make a small one someday , With the rollers you have used nylon and PTFE ? What one sits at the bottom / vs the top ?

It is almost certain that it is NOT oil. MOCs are film capacitors and depending on the film technology its either a dielectric or impregnating fluid.

Its hard to find a datasheet that really specifies it, as its not a oil, you will not see any "Contains no PCB" on them.

They are foil and paper capacitors , with an thin oil like liquid filling normally so in this case some form of impregnating fluid and quite a few state no PCBs

I have non MOC foil/paper caps for PFC that are filled with caster oil .

Thanks

I never measured the primary current when i tried this as it was just a quick setup and for sizing the capacitor it was just trial and error for the most part : )

Do you have a supplier for the silicone oil ? , Currently I use transformer grade mineral oil with my X ray transformers , but think changing over may be a better solution for longer life ?


When I vacuumed my small transformer , It took around 7 days straight and I used a small vibration module to help move the trapped air , Worked really well .

After the 7 days , No more air bubbles could be seen and i stopped the vacuum , Seems to work fine currently : )

When I made a few ZVS drivers I always ran the gate supply from 12V and varied the voltage into the transformer primary to control the output .

Seemed to work well , though not sure how reliable it would be in the long run

one reason for the 4 windings on a full bridge GDT is that even though the two mosfets/igbt are in phase they are not both at the same potential

One switch will be sitting near the supply voltage while the other will be at ground with respect to its source so you need windings for each to 

I have done both using manual PFC with high voltage transformers and it dose work within a range , as said knowing the inductance of the circuit helps in selecting a capacitor .


I have also have built a system that uses a industrial PFC controller to control 12 capacitors and will actively try and maintain a PFC , I have had good results with test trials .


On a high voltage transformer setup with inductive ballast and auto PFC i was able to hit 0.98 PFC during arcs , there are limits to this as the PF controller was never intended for this and is abit slow , but the principle is there .