On a sunny day (Mon, 24 Feb 2014 08:53:38 -0500) it happened Spehro Pefhany wrote in :
Its is above the edge, can move around, it is supported by teh MOSFET pins, and by a sinle pic bottom left into the PCB.
Well, I won't use it becasue of the RF interference, but also because I think it is too much risk, I want at least my test equipment to be safe, sometimes other voltages appear, once there is an arc it keeps going. I have a 400 kV Tesla on order.... :-)
Home insurer: "So we found you posted to Usenet this thing was unsafe, the house burned down, so why should we pay?" sort of thing. Not worth the 10 $ That MOSFET could be useful...
That was what I was wondering... maybe D1 was put in backwards. If Jan's not going to use it, then flip D1 and see what happens. (Maybe install the blast sheild, before adding power :^)
Would there be sufficient charge sharing between the Source and Gate to create an current impulse in the transformer and with positive feedback from the secondary to then properly drive the gate?
--
Mike Perkins
Video Solutions Ltd
www.videosolutions.ltd.uk
On a sunny day (Mon, 24 Feb 2014 16:32:59 +0000) it happened Mike Perkins wrote in :
When power is applied the Cds will cause a current in the transformer. It is just a few pF, but probably enough to get it going. It is also possible I missed something in the layout. No longer an issue now :-)
So it's a proper fullpack. So you lied. Heatsink has no connection to AC. No electrocution hazard from it at all. A little RF due to package capacitance.
And the 1M goes to +V after all, just as I hinted and Sphero verified.
On a sunny day (Mon, 24 Feb 2014 11:20:44 -0600) it happened "Tim Williams" wrote in :
Yea, possible, still a piece of shit, glad I killed the interfering POS. I am sure you like the track between the opto pins too. Did I mention I found a lose piece of solder in it? Yes I did.
really? I do that in the case of switching to prevent HV and bring the tank into near resonance. If noise reduction was all it did, I would think the chinese would op to leave that out! Just think how much they could save just by leaving out one cap out of those many supplies they make.
It would screw up touchscreens and they'd get tons of returns.
Best regards, Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
On a sunny day (Mon, 24 Feb 2014 20:25:11 -0500) it happened Spehro Pefhany wrote in :
I think the subject line in the thread I started says it all.
500 kHz now. Smaller transformer needed, cheaper. It will not stop there, I expect them to make coils on both sides of the PCB as transformer, and go to 100 MHz or more in the future. That any radio or TV services will no longer work only sells more of those. Extrapolation, I mentioned that before in sci.physics, extrapolation and design. So extrapolation tells us about the future Chinese designs here.
As to the Tim Westcrop or whatever rant, if the 1M is indeed to + primary supply, I also found a small glass diode when ripping it apart that then could perhaps be a zener of sorts against the
1 M to limit gate voltage.
I did draw a quick sketch, not design a circuit diagram, just to see what they (Mao & Co) were up to. If you cannot live with that try posting something useful yourself one day Timmy and Philly.
As to the 'isolated' MOSFET,. that is crap, it is fixed with a parker screw of unknown torque to a lose heatsink strip that misses the pin on the far side (there is a hole in the PCB for it), and the 'glue' just fell of when I moved it a bit,. Metal fatigue of the MOSFET pins when this thing experiences vibration will make it lose and can easily make a short,. The required 13 mm free space is missing at the opto, the thing has no filters in either in-or output, it has no screening anywhere, the seller is a crook as he displays an image of a product with FCC and what not labels but shipped one with a fake CE label, and the thing was build in a bad way. Found some fluid under one 450V cap, thought it leaked, but that dark brown fluid was probably PCB cleaner or whatever.
In short this thing is dangerous, you should not buy from this seller, the thing is a RF jammer, and that alone was the reason I scrapped it, my analog video f*cked up by it on the other side of the house!
I timed myself taking it apart, 8 minutes, not bad.
D1 is most likely a zener with a breakdown voltage somewhere above rectified mains (maybe 400-500V or so). In this case it works by clamping the flyback peaks only, while also doing double-duty to periodically "reset" the RC to a known potential.
On the other hand, are you sure that the switching device in there is really a MOSFET? If so, as others have noticed, it cannot reliably start, since its "gate" terminal has a DC path to HVGND, but no path (DC or otherwise) to HV+. Since all secondaries are initially deenergized on power-up, there is no way to get energy to supply the "gate" from there. Also, it's unlikely to start by avalanche, for the simple reason that it's unlikely to survive avalanche. The flyback pulse (way above mains voltage, especially for a wide
100-230V mains) would cause the turn-off cycle to abort by re-avalanching the "FET" immediately, which would lead to a catastrophic failure.
What's more likely, is that the switching element is a little more sophisticated. It's likely a self-starting SMPS IC that works like a "normal" MOSFET, but provides internal pulse triggering. It's probably not very complex and has no real control circuitry, current-mode or otherwise. More like an SCR-type trigger, lamp dimmer style, that has been miniaturized and added to a MOSFET in the same package.
I could not find any documents for what was scribbled on the schematic, so here's my wild simplistic guess at its innards (all parts except the cap integrated, the resistor possibly being a depletion MOS constant current source).
On a sunny day (Tue, 25 Feb 2014 20:21:02 +0100) it happened "Dimitrij Klingbeil" wrote in :
I dissasembled the thing, and it tured out D1 is a normal diode, FRG105G. with a reverse breakdown of about 600V. I tcould perhaps be used as some sort of zener,
Yes, google WF_5N60_MOSFET.pdf
If so, as others have noticed, it cannot reliably start, since its
Yes it has been pointed out by some the 1M is to the + input supply, not to ground.
Since all secondaries are initially deenergized on power-up, there is
It think the diagram I did draw is OK apart from the 1M that goes to +, not ground.
Your circuit could perhaps work too, have not tried it. Thede cheap converters all use some feedback turn on the transformer for oscillation. I have used that myself, but not at 500 kHz that way... but lower fequency sinewave. Using a sinewave you do not get harmonics to near infinity.
Disregard my diagram. I was thinking, the above could be integrated into the MOSFET to make it a "self-starting switching transistor", while keeping the rest of your schematic exactly the same (including the feedback winding, which takes over once the first pulse has got it started). Too much wishful thinking, you see :)
It turns out that startup was done in a simpler way. The 1M is large, so, as the gate slowly charges up, the FET will go linear and let some random amplified noise provide positive feedback via the transformer windings (and start oscillation before the FET fries itself from rising I_ds).
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.