Post by: CCX191 on May 30, 2017, 07:16:58 PM
It has been about 3-4 days of building the circuit. I have very minimal understanding on circuitry/ electronics so I will be posting some of my progress here and some questioning.
My first question might be quite stupid but.. how do I power the circuit? I have done the entire driver except the bridge and would like to test my GDT. At which point where my power is supposed to be connected?
At the image attached is the circuit so at the top left I can understand that I will put in 12VDC and the ground as negative. So how about the arrows pointing upwards with 12VDC do I put in 12v in there too? And then the ground arrow at the interrupter is where I connect my negative?
Post by: futurist on May 30, 2017, 10:26:44 PM
Keep in mind that you have to have GDT connected, otherwise you'll blow up the MOSFET drivers
Post by: CCX191 on May 31, 2017, 02:23:37 AM
Post by: futurist on May 31, 2017, 10:24:43 AM
How do you plan to test the driver?
Post by: Mads Barnkob on May 31, 2017, 10:59:13 AM
As futurist points out, it is important to have a load connected to the MOSFET driver ICs, if they are running open loop they can oscillate at a very high frequency and burn themselves down. Connect the GDT and also connect the MOSFET gates to the secondary side of the GDT, this way you will have a more realistic waveform to check.
I normally use a simple 555 timer or a signal generator for the feedback circuit/antenna to test the driver alone.
Post by: CCX191 on May 31, 2017, 11:30:56 AM
Edit: And also does that mean the interrupter does not require another power in?
Post by: Mads Barnkob on May 31, 2017, 12:38:31 PM
Ah thanks! I just burnt my UCC3732Xs so I guess I will get back when I get my replacement
Edit: And also does that mean the interrupter does not require another power in?
I had it all mounted in the same enclosure, as you can see in the pictures on my website, they shared the same 12V 2000mA power supply.
Post by: CCX191 on May 31, 2017, 02:03:36 PM
Post by: futurist on May 31, 2017, 02:22:02 PM
Don't worry about blowing up components, it happens to everyone - doublecheck your connections before powering anything
Also be very careful when powering the bridge with mains power, it is deadly
I would buy few spare UCCs, they are not that expensive
https://www.aliexpress.com/item/Free-shippin-5pcs-lot-UCC37321P-UCC37321-DIP-8-Bridge-Drivers-new-original/32553627093.html
Do you have an oscilloscope available?
Post by: CCX191 on May 31, 2017, 02:33:20 PM
Edit: I have two bridge rectifiers, GBU1004(400V 10A)& KBPC3504(400V 35A) and I've ordered a 1500uf 450v capacitor which one should I use? And the variac used is 2000W.
Post by: Mads Barnkob on June 01, 2017, 10:57:14 AM
The DSO138 has a low sampling rate and should not be used for signals over 100 kHz, so it has some real limitations, but it is also cheap, but you should be able to build a SSTC with help from it :)
I suggest that you use the 35A bridge rectifier, because of the poor power factor of a SSTC, the 10A might get too hot.
Variac is perfect, variacs can even be overloaded by 300% for some minutes :)
Post by: CCX191 on June 02, 2017, 12:18:21 PM
In Mads post I see that there is a metal sleeving on the secondary and another wire, is it due to the fact that one connects to mains ground and the other to "ground" or the striking rod?
I also have some questions regarding the audio interrupter. I see that it uses a 9v supply which I assume I can get a separate power source or a step down circuit for that right?
Edit: Does polarity matter in the GDT? If so how do I determine :/
Post by: Mads Barnkob on June 04, 2017, 11:56:15 AM
While I wait for my variac, capacitor and UCCs to arrive I've decided to use a metal box (painted) to house everything. I suppose I have to avoid the the Tesla coil from striking it's driver & bridge and from it's the EM field So I assume that everything should be insulated. Is there anything I should be aware on the enclosure? Like grounding and etc.
In Mads post I see that there is a metal sleeving on the secondary and another wire, is it due to the fact that one connects to mains ground and the other to "ground" or the striking rod?
I also have some questions regarding the audio interrupter. I see that it uses a 9v supply which I assume I can get a separate power source or a step down circuit for that right?
Edit: Does polarity matter in the GDT? If so how do I determine :/
You should ground everything to the same ground point, in a star ground manner, just google "star ground".
It is correct that I had a black wire going to house mains ground, but today I would not do that again! The copper mesh sleeves goes to a strike target that I used for measuring spark length, so it is just everything grounded together. I believe its better to make a artificial ground from aluminium foil, kitchen foil, to cover a whole table where the Tesla coil stands on top, instead of using house ground.
I think I just used a 7809 regulator for the audio interrupter supply. From the 12VDC supply for everything else.
Polarity is very important in a GDT, its a regular transformer. You need to drive the MOSFETs with opposite polarity to make it switch any current into the primary circuit. Before you wind the transformer, mark all the wires black in the same end, then when you twist the individual wires to a stand-alone winding, you have a black mark on one of the ends.
If you take a look at my DRSSTC1, you can find this picture of the GDT: (https://highvoltageforum.net/proxy.php?request=http%3A%2F%2Fkaizerpowerelectronics.dk%2Fwp-content%2Fgallery%2F2009_08_27_-_kaizer_drsstc_i%2FIMG_2683.jpg&hash=52b0cf62a5ff6eee6b515fd492214208ebbd7402)
Post by: CCX191 on June 04, 2017, 02:24:29 PM
Hopefully my variac can arrive by this week :P
Edit: Just redid my GDT (https://highvoltageforum.net/proxy.php?request=http%3A%2F%2F%5Battachment%3D1%2Cmsg359%5D%5B%2Fattachment%5D&hash=bed48047a287a72276e3d44040c38b49e7afc601)
And oh how do I enable CW mode? By simply detaching the interrupter and put in 12V ?
Post by: Mads Barnkob on June 05, 2017, 08:56:18 AM
Polarity of the primary side of the GDT does not matter, it only matters that the polarity of the MOSFETs are opposit so that it switches the correct MOSFETs on to lead current through the primary coil.
I know I wrote on the schematic that disconnecting the interrupter made it run in CW mode, but other people have had problems recreating this behavior, so to run CW mode you have to keep the UCC driver ICs enabled, check the datasheet of your drivers if you have to pull it to GND or +12V to keep them on all the time.
Post by: CCX191 on June 07, 2017, 08:30:01 AM
Here's a picture.. am I even using the right capacitors?
Post by: Mads Barnkob on June 07, 2017, 08:41:25 AM
It looks like a regular MKP (polypropylene) capacitor, which is the right type to use, but it is rather small.
It is very hard to see in your picture what is what, could you please upload some more detailed pictures of your whole setup, test and coil etc.
Post by: CCX191 on June 07, 2017, 08:51:33 AM
This is how the whole driver + bridge looks like without the capacitor attached yet
[ Invalid Attachment ]
This is kinda how my bridge looks like
[ Invalid Attachment ]
Post by: Mads Barnkob on June 07, 2017, 02:26:13 PM
Also, ground that box as the first thing, to save yourself from getting shocked if there is any other wrong wiring.
Post by: CCX191 on June 07, 2017, 02:53:02 PM
Post by: Mads Barnkob on June 07, 2017, 03:02:18 PM
You have to double check your circuit to the schematic then, also use some larger capacitors for the primary circuit.
Post by: CCX191 on June 07, 2017, 03:57:15 PM
[ Invalid Attachment ]
Post by: Mads Barnkob on June 07, 2017, 04:00:54 PM
.68 = 0.68 uF
K = 10% tolerance
63 = 63 VDC
You need capacitors rated for at least 400 VDC in the primary circuit.
Post by: CCX191 on June 09, 2017, 06:39:50 AM
Oh and talking about the safety of using the audio moudulator, I see that you have a metal mesh over it to protect it from interference etc but would it work if i connect it to my phone since i see that you've used only a keyboard? Should an optocoupler be implemented or can just be omitted? :)
Currently 3D printing parts for me coil and toroid ;D
Post by: Mads Barnkob on June 09, 2017, 08:12:22 AM
I see.. My capacitor replacement will be arriving in an hour or so and for now I will work on the audio modulator. I'm wondering about the audio input section.. In terms of (+) and (-) I will be using a 3.5 mm audio jack but in terms of polarity is (+) the left channel and (-) ground or (+) left, (-) right?
Oh and talking about the safety of using the audio moudulator, I see that you have a metal mesh over it to protect it from interference etc but would it work if i connect it to my phone since i see that you've used only a keyboard? Should an optocoupler be implemented or can just be omitted? :)
Currently 3D printing parts for me coil and toroid ;D
It is not a stereo input, only mono, so its either right or left channel as + and then ground, but notice that the ground of the audio is not tied directly to ground of the circuit!
Never connect anything directly via cable to a Tesla coil that you can not be without, this is why I used a cheap kids toy, I would not cry if I destroyed it :) I once destroyed a ipod nano from connecting it directly to a plasma speaker, back EMF destroyed the USB interface in it.
Optical isolation is always a good idea!
What kind of 3D printer do you own/have access to?
Post by: CCX191 on June 09, 2017, 04:13:51 PM
As of now I haven't got my toroid mounted since I've ran out of print filament. I've tested my coil and it has successfully lit a florescent light but it's rather dim. A spark is not produced even with a tweezer as a breakout point but I guess I will try again later once the toroid is installed. A small spark can be drawn though about 3mm-1cm that I've tried max
The thing is I've already tripped my house twice from drawing too much power from the wall :/ maybe at about 150VAC-200VAC. Not sure if I can do anything about this.
The printer I have is an Anet A8 that I bought from gearbest for about 155USD. Pretty cheap and a very worthy investment in my opinion, though some upgrades must be done like a change of better belts as the stock ones are horrible and bearings if you can't bare with the noise. I use SketchUp​ to model my things
Here was a setup I have earlier this afternoon
Post by: Mads Barnkob on June 11, 2017, 09:27:10 AM
So my audio modulator doesn't seem to work probably I've wired it incorrectly so I will just check that up.
As of now I haven't got my toroid mounted since I've ran out of print filament. I've tested my coil and it has successfully lit a florescent light but it's rather dim. A spark is not produced even with a tweezer as a breakout point but I guess I will try again later once the toroid is installed. A small spark can be drawn though about 3mm-1cm that I've tried max
The thing is I've already tripped my house twice from drawing too much power from the wall :/ maybe at about 150VAC-200VAC. Not sure if I can do anything about this.
The printer I have is an Anet A8 that I bought from gearbest for about 155USD. Pretty cheap and a very worthy investment in my opinion, though some upgrades must be done like a change of better belts as the stock ones are horrible and bearings if you can't bare with the noise. I use SketchUp​ to model my things
Congratulations! You got sparks, so something is working, now it is just to get it working properly :)
A SSTC will work fine without a topload, sparks will just be shorter, but not as short as those you describe.
Try to reverse the phase of the primary coil, this is done by switching around the two leads from your bridge to the coil, a SSTC can actually operate out of phase, this would fix the small sparks, if that is the problem.
You should not trip your breakers whe only producing such small sparks, maybe you have a defective part that will first fail with a short circuit at high input voltage? Else it is also related to the above phasing issue.
Interesting and cheap little machine, I might try to get one, learn the basics and play around with a cheap machine instead of just going for a expensive first and maybe damage it :)
Post by: CCX191 on June 11, 2017, 10:04:19 AM
As of now my driver is practically broken I accidentally put 240v straight into the bridge without the driver turned on. As a result I did some testing(edit: I think it shouldn't really matter)... That there is nothing coming out from my GDT so I assume my UCCs are dead once again :'( (edit: tested again seems to be kinda fine)
Though for my scenario of putting 240v into the bridge without the driver turned on would it kill the IRFP460 too?
For now I just have some spare UCC37322 but not UCC37321( they are shipping from China now) so I guess I can resume to retry it next week :'(
Edit: my UCCs seems fine though :/
Edit: After some inspection and blowing the fuse on my variac... I'm pretty convinced that my bridge is screwed up so I guess I will redo it when I'm free. Schools gonna start again :'(
Today's setup with a temporary topload fixture using MDF
Post by: Mads Barnkob on June 12, 2017, 11:33:39 AM
Always driver on first, then power electronics.
Be sure to not just check the MOSFETs, but also check diodes, resistors and zeners.
Your topload looks good, that will do fine for testing!
Beware of metal objects like screws and long bolts like the one in the corner, they will affect the resonant frequency of secondary circuit and will absorb some power, worst case you can have a spark jump from topload to the long bolt and then arc into your driver/bridge underneath the plate.
The screws holding the secondary in place are also subject to be induction heated from the primary coil, use nylon screws if possible, or glue the white plastic ring on the wood instead.
I hope you get new parts and find some time between school days to finish your coil and make some good long sparks :)
Post by: CCX191 on June 12, 2017, 01:01:51 PM
I just redid my bridge and blew the fuse on my variac twice... So I guess there is something faulty with some of the parts. How do I test the MOSFETS though?
Is there any possibility if I just turned on the bridge and it might kill the driver? Since it's just connected by the GDT it seems unlikely? I'm not quite sure what's going now with my whole driver and bridge.
My first bridge worked fine though only if I have used the right DC blocking caps. But I guess it's a learning curve.
Post by: Mads Barnkob on June 14, 2017, 10:00:47 AM
Check this video on MOSFET testing with a DMM
/>
A GDT provides galvanic isolation between driver and bridge, and you will often see exploded bridges where no harm have been done to the driver section, because of the GDT. It is mostly transistors, gate resistors, zeners and diodes in the bridge that takes damage from a fault.
Post by: futurist on June 16, 2017, 11:53:34 PM
And don't worry about burnt components - FET is commonly interpreted as "Flame Emitting Transistor"
Post by: CCX191 on June 23, 2017, 07:15:10 AM
School was very taxing this week and I've managed to rebuild my driver for the fourth time two days ago and then consequently blow it up the next day. After some inspections one of the FETs was not switching properly and a resistor blew for the FET that blew up.
My first MOSFET that blew up :D
It sounds like a sort of milestone/ achievement now.
Post by: futurist on June 23, 2017, 02:05:34 PM
Post by: CCX191 on June 23, 2017, 02:07:43 PM
Post by: futurist on June 23, 2017, 06:33:28 PM
I would suggest asking someone who is familiar with the electronics to help you, it's hard to tell what went wrong without seeing the setup live
Post by: CCX191 on June 23, 2017, 06:55:19 PM
I concluded that the MOSFET weren't running as 3 MOSFET's heatsink were relatively warm except for one.
I'm still waiting for my DMM to arrive to test the MOSFETS as I use a cheap analogue one . Though I borrowed my friend's cheap DMM to test and shown that 5 MOSFETS from China were shorted but I have to return it shortly after.
I just went to the component shop today and bought 8 IRFP460 just in case there is more to blow. It's rather expensive here (2$ for one) and the shop is quite far and I can't drive yet.
My resources are scarce and help locally is impossible to find since I'm just in highschool or rather the electrical engineers I know seems to have no knowledge in this field. So asking here is my best bet.
The first try I did created a spark naturally but due to the wrong cap I used, it burned down. So I shall assume that my GDT is doing fine as of now. The current sparks I have been getting recently are like a Van De Graaff Generator. Only can be seen if drawn to a metal stick or sort, rather minute.
Just to reassure are these caps correct?
Post by: Mads Barnkob on June 24, 2017, 12:35:38 PM
The capacitors are great for DC blocking on the primary circuit. They will be able to work with the currents in this circuit, atleast for shorter run times in the minutes range, not like we do much more than that anyway.
It is great to hear that your are venturing into unknown fields to electrical engineers, just tells how little basic electronics they actually do :) You are on a great journey and I wish I started as young as you did!
Post by: CCX191 on June 25, 2017, 10:55:25 AM
I understand that the wire should be physically close together and wound tightly to reduce leakage. And that screened circumferential type of winding seems to perform better.
From what I understand, tri-filar is by twisting all the wires together and then winded in the core. So if it's screened, is it just basically a tri-filar that's been somewhat insulated with perhaps heatshrink? Is it even called a tri-filar.. there's like 5 wires..
I have some enamelled 0.5mm wire should I use that instead or stick with Ethernet wire.
I think I'm still a little lost on this whole GDT thing :(
I'm currently using this core http://my.element14.com/epcos/b64290l0048x830/ferrite-core-cylindrical-34mm/dp/2355017?CMP=i-55c5-00001621 (http://my.element14.com/epcos/b64290l0048x830/ferrite-core-cylindrical-34mm/dp/2355017?CMP=i-55c5-00001621)
Regarding my scope, I think I will try it again and take some pictures. I'm not too familiar with this scope yet though, or rather use one. I guess I will just play around with it. Do I connect both of the probes at each end of the wire or how?
Edit: This is what I get in my scope connected one end to the negative terminal and one end to the GDT from the primary
And the other end of the primary and the negative terminal
And this is what I get connecting it to the gate and source
Post by: CCX191 on June 30, 2017, 11:11:13 AM
My capacitor would still hold it's charge though so I assume it's not even switching or power is even consumed. The MOSFETs are new same goes to the diodes and resistors which I have tested all except for the blocking capacitor. Perhaps the blocking capacitor has failed? I can only get hold of my DMM this Sunday to test it. I I'm actually trying to get it to work by this weekend supposedly so I can present it in school but I guess that's not going to happen.
Or.. is there a way to make this to a half bridge with the irfp460?
Post by: Mads Barnkob on July 01, 2017, 08:58:07 AM
Regarding using the oscilloscope, you should at least watch this, and also some of the other introductions to oscilloscope videoes that Dave has on his channel: />
The square wave going into the primary of the GDT is the first shot? That looks okay.
The second picture shows a duty cycle over 50% and thus can not be a correct signal to either drive a GDT with or get out from the secondary side of it.
The third picture clearly shows that something is not right in the driver -> GDT -> gate setup, you should see a square wave here, maybe with some minor oscillations, peaks or roundings of the edges, but it should be a square wave.
I have, when I first started doing electronics, had to give up and on a circuit and start all over from scratch, because you might have multiply errors that keep interlocking each other, so even if you fix something you can not see an improvement. You have to just work your way out from left to right. Feedback -> driver -> GDT -> gate drive -> inverter output and make one thing work before moving on.
You can make a half bridge, that is no problem, check out the SSTC1 on my website.
Post by: nabzim on July 01, 2017, 09:49:13 PM
I have some enamelled 0.5mm wire should I use that instead or stick with Ethernet wire.
When I was trying different GDT winding methods, a couple times I tried I using magnet-wire. The insulation is so fragile, that just the process of winding it on the core actually scraped some insulation off the wire, and I ended up getting arcing between wires, and they were little purple sparks. So there must have been a high voltage induced in the GDT windings I think. So from that point on I vowed to always use wire with real insulation, and I just spent the money to buy a small roll of cat5 from home depot. It was 24awg solid core,
Which works good, and I even recently bought another roll but with stranded wire, so I can use that for all sorts of stuff too. (I highly recommend the Ethernet wire. It gives you 4 different pairs of wire and they're already in twisted-pairs AND they're already all color-coded differently for you) it is worth the small investment, then you will also have that wire to use for routing signals and/or power for anything else you want.
Post by: futurist on July 02, 2017, 12:55:20 AM
Strip it by hand after heating with lighter, to avoid scoring the wire
Post by: CCX191 on July 02, 2017, 06:09:10 PM
I redid everything and converted it to a half-bridge to save my time so it's the Kaizer SSTC I now !
I have seem to maybe actually point out one of the issue. I have several UCCs lying around.. One set from 2 different sellers from China and a set from element14 but one is from the US and one is available locally.. all of the 21s and 22s are different.
Is there a possibility of compatibility issue with them? At most of the combination.. the 22 will get extremely hot while the 21 stays cool. At the GDT it does not resemble a square wave at all but like the third picture of the scope from my previous post. At certain combination nothing will be produced. A little step to progression again :D
Or, could the 74HC14 be faulty?
Oh, how would I be able to test the 74HC14 though? I've grounded the unused input pins which are 5, 9, 11, 13?
Post by: Mads Barnkob on July 03, 2017, 09:41:02 AM
Cheap ICs are often not fake, but could be low quality rejects from factories. If they are fake, it is often "same" kind of IC but with lower ratings, even if you had some fake with only 3A peak power output, it would "work" for this driver. My own experience with UCCs failing/not working is that I made a mistake at the GDT so that they are not loaded correctly or the input/enable was wrong.
74HC14 testing: feed it with a sine wave and see that a inverted(or normal if you have two inverters connected in series) square wave is outputted from when the rising edge of the sine wave passes the threshold voltage of the 74HC14 input.
Post by: CCX191 on July 03, 2017, 03:47:02 PM
I guess those cheap UCCs are probably defects. They seem to not work properly at all.
The heating up seems to have some relations with the 74HC14 if that is working properly or not
For the 74HC14 I have replaced it and have gotten slightly better results.
This is what I get on the bridge gate and source
You can see the legs of all the dead Fets of the bottom right :P
It's slightly taking shape of a square wave.. but I'm probably sure I'm not quite there yet( or else usually at other times there will be just lines all over)
Post by: CCX191 on July 07, 2017, 07:50:21 PM
My GDT produces a nice square wave now and now it is running at a half Bridge configuration.
Here's a picture During testing.
[ Invalid Attachment ]
It's running pretty well.. except that I've run into this issue where there will be sparks at the tip of my antenna at times. Anything that should be done?
Post by: oneKone on July 08, 2017, 01:21:14 AM
Great news!
My GDT produces a nice square wave now and now it is running at a half Bridge configuration.
Here's a picture During testing.
It's running pretty well.. except that I've run into this issue where there will be sparks at the tip of my antenna at times. Anything that should be done?
I've been quietly following and it's awesome to see improvement!
As for the antenna try a longer breakout point, and maybe move the antenna a bit further. I've had that problem before when I made a small sstc with a largish topload.
Post by: Mads Barnkob on July 08, 2017, 07:40:22 AM
I agree with oneKone, move the antenna further away from centre of the coil and shorten it to be further away from the topload. I had mine so that there was visible corona on the tip of the antenna when running full power.
Post by: CCX191 on July 08, 2017, 08:00:20 AM
The cap I was using back then: goo.gl/wL9snH
Current caps I'm using: goo.gl/3NLiJi
Since the cap is rated at only 300v now. At what limit should my variac go?
Regarding the antenna I think the right term I meant was the visible corona. Should I be concerned over this? I can slightly move it away for sure.
Now i'm planning to just get my audio modulator to work. I see that user E-Schmok has a video on using a Toslink mini jack here: goo.gl/xwiL9w which I'm not entirely sure how to get it to work. If not I will just use a 3.5mm jack or so. Any safety precautions I should be aware of? I have a 6N 137 opto isolator but i'm not entirely sure if it's applicable or how to wire it.
Post by: nabzim on July 08, 2017, 09:27:10 AM
It's running pretty well.. except that I've run into this issue where there will be sparks at the tip of my antenna at times. Anything that should be done?
that was also happening on my coil (I can now see its very common, from Mads' and oneKone's replies) . It never seemed to do any harm, but I still did few things like: i curled the tip of the antenna into a loop to make it nice and round, then covered the whole thing in kapton tape because it was just bare 12 awg copper wire. but there was still corona through the kapton tape, so ultimately, I ended up just adjusting the distance of the antenna until the sparks didn't occur anymore. but that also reduced the actual output of the coil a little bit also, so you pretty much just take the good with the bad
Post by: CCX191 on July 08, 2017, 11:36:33 AM
Post by: oneKone on September 30, 2017, 02:39:37 AM
Have you made any more updates/progress with this coil?
Post by: CCX191 on September 30, 2017, 04:45:46 AM
Sadly no :-\ I'm having exams now till November and then I will finally be out of high school and I can continue to work on it and some other projects I have been planning to work on. As of now I will just float around here :P
Post by: oneKone on October 06, 2017, 01:54:02 AM
Goodluck with the exams.