High Voltage Forum
Tesla coils => Dual Resonant Solid State Tesla coils (DRSSTC) => Topic started by: Thunderstruck on February 26, 2019, 04:40:45 AM
-
Well, I just could not sit still after giving up on my last project, so I decided to have a go at building a Universal DRSSTC Driver 2.7C.
I never soldered SMD before, so I was a bit doubtful of my skills, but it turns out that it is quite easy even with an el-cheapo USB Microscope.
It took quite a few hours, I did not want to rush, still I had to take 2 components off because I put them wrong way around ( removing components without a hot air gun is not easy ).
Lag from the camera was giving me sea sickness 😁, but I got there.
Board is completed except for a few Build/IGBT specific components - I still do not know what IGBT’s to use on my coil ( I decided to convert my current TC to solid state ).
I did not even power up my board to test because I do not have an interrupter to do a complete test.
I found some plans for interrupters here:
https://github.com/profdc9/DRSSTC-PCB-Pack
Are these any good, or is there anything more suitable ?
Finally here is my board, if you something wrong, please do not hesitate to comment !
-
Damn that is a nice looking board!
For interrupters I use a standard burst mode type designed by Steve ward, and for midi I use onetesla interrupter. I haven't looked at the interrupter profdc9 has released, but if the through hole ud2.7c is anything to go by they should be top quality.
-
It looks nice, I guess one of the specific components you say is missing, that I could see, is the OCD current transformer resistor?
I am not sure if Dan ever tried building his interrupter, but it should be a good standard interrupter. There is also alternatives as the good old and simple steve ward interrupter: http://stevehv.4hv.org/drsstc_interrupter.htm and I also have the plans for a simple non-midi audio interrupter from ctc-labs: http://kaizerpowerelectronics.dk/tesla-coils/musical-sstcdrsstc-interrupter/
For midi I use loneoceans midi2 kit.
You can drive bricks up to around CM600s with these drivers, so there is plenty to choose from, depending on the size of your coil and resonant frequency :)
-
Damn that is a nice looking board!
For interrupters I use a standard burst mode type designed by Steve ward, and for midi I use onetesla interrupter. I haven't looked at the interrupter profdc9 has released, but if the through hole ud2.7c is anything to go by they should be top quality.
Hi oneKone,
Thanks, that is my first SMD build, never tried soldering that small.
I found profdc9’s github repository, that is the link I posted at the start of the thread.
He does say that it is work in progress, that is why I wasn’t sure if those plans are any good.
Everything is there, I especially like Gerber files for the pcb’s, makes everything so much easier.
-
It looks nice, I guess one of the specific components you say is missing, that I could see, is the OCD current transformer resistor?
I am not sure if Dan ever tried building his interrupter, but it should be a good standard interrupter. There is also alternatives as the good old and simple steve ward interrupter: http://stevehv.4hv.org/drsstc_interrupter.htm and I also have the plans for a simple non-midi audio interrupter from ctc-labs: http://kaizerpowerelectronics.dk/tesla-coils/musical-sstcdrsstc-interrupter/
For midi I use loneoceans midi2 kit.
You can drive bricks up to around CM600s with these drivers, so there is plenty to choose from, depending on the size of your coil and resonant frequency :)
I discovered loneoceans’s kit recently, seems like a good quality circuit, a bit on the expensive side, but not too bad. I am seriously considering just buying one.
As far as missing components are concerned, yes C33 is missing and also Slot 7 inductor ( cannot find one locally, getting one from abroad is going to be pricey because of shipping )
I cannot remember resonant frequency of my coil, I’ll have to measure it again. Complicating issue is that I cannot do it any time soon because someone filled my garage with tons of furniture, and my coil is right behind it.
It is a 4 inch secondary, 1050-ish turns of 25AWG wire, Primary is pancake coil, 600mm OD, 150mm ID. 12 inch toroid top load.
MMC based on16 x 942C20P15K Capacitors for a total of 9.3nF
Based on these numbers JavaTC gives me 400kHz resonant frequency on the secondary.
And looks like I will not have to modify my existing coil, I have a spare secondary, spare MMC, and left over 1/4 inch conductor for another primary. Just need a top load !
You can see my coil in action here:
/>
-
here's a home made version of UD2.x
doesn't look that great but it does the job and it was a fun learning experience
[ Invalid Attachment ]
-
I build both of the lineinterrupter and the standard interrupter and they both work. However, I did tweak the standard interrupter slightly since I built it to make it more similar to the line interrupter. I am fairly certain they both work. The thing I tweaked on the standard interrupter is the MIDI support to correct something to be consistent with the lineinterrupter so they should both be nearly identical now.
The lineinterrupter triggers off of the 50/60 Hz AC cycle line to best match the power factor of the bridge to the peak of the AC cycle. The idea is it allows the capacitors to charge during the cycle before allowing the interrupter to fire. I did not have a problem blowing any oscilloscopes or anything like that, but because the power I have is limited I was not able to push this hard without tripping the overload in the breaker.
The standard interrupter has four modes: standard test mode, which is basically pulsing the interrupter periodically for a particular pulse length, burst mode, which is to try to get briefer, more intense sparks, the MIDI interrupter, which can be used for playing MIDI files from a computer through USB, and the external audio sequencer, which if you play a tone like, for example, from a guitar, will generate a set of pulses with controlled length. You can see it in action here
/>
Hopefully that helps you understand the state of it. So I did build both, but the standard interrupter has a few fixes in it from the last tested version.
Dan
-
Dan, thank you very much for your help. I ordered a few PCB’s from Pcbway, I will try to build one of your interrupters. Very generous of you to give away all the files for free.
-
The standard interrupter has four modes: standard test mode, which is basically pulsing the interrupter periodically for a particular pulse length, burst mode, which is to try to get briefer, more intense sparks, the MIDI interrupter, which can be used for playing MIDI files from a computer through USB, and the external audio sequencer, which if you play a tone like, for example, from a guitar, will generate a set of pulses with controlled length. You can see it in action here
The code for the ATTiny, is that the onetesla usb midi code that you have a folder backwards in the same git? I did not find a clear description of which software was for it.
-
Yes the OneTesla code is there. You will need an AVR programmer and avrdude, which costs $5 on ebay, and the hex file is already there. There are instructions in the folder with the software.
Dan
The standard interrupter has four modes: standard test mode, which is basically pulsing the interrupter periodically for a particular pulse length, burst mode, which is to try to get briefer, more intense sparks, the MIDI interrupter, which can be used for playing MIDI files from a computer through USB, and the external audio sequencer, which if you play a tone like, for example, from a guitar, will generate a set of pulses with controlled length. You can see it in action here
The code for the ATTiny, is that the onetesla usb midi code that you have a folder backwards in the same git? I did not find a clear description of which software was for it.
-
I still did not power up my board, but I would like to check te voltages at least. Would it be safe to turn it on wthout GDT’s connected also without C33 and the slot 7 inductor installed ?
-
I still did not power up my board, but I would like to check te voltages at least. Would it be safe to turn it on wthout GDT’s connected also without C33 and the slot 7 inductor installed ?
As long as you do not have unloaded/unconnected outputs of gate drive ICs, there should be nothing wrong with powering it up to check voltages etc.
The problem with the gate drive ICs like UCC32xxx, MAX44xx etc is that they can oscillate in the MHz region when the output is left open loop and it will destroy itself from switching losses.
-
So I must connect a couple of GDT’s to it then ?
-
So I must connect a couple of GDT’s to it then ?
the UCC32xxx in a UD2.7 drives some mosfets so i don't see a problem with powering it up without a GDT connected
-
I studied loneoceans’s instructions on initial power up of the board, nowhere did he mention that GDT’s must be connected, he even mentioned that if there is no GDT’s connected, there should be no signal on the outputs.
So, I gathered enough courage to power up my board, I set my power supply to 24V DC at 200mA max.
To my relief, blue LED came on, and there was no smoke leaks anywhere, current draw 0.02 A.
So far, so good :)
While I was procrastinating whether to power it up or not, I built an enclosure for it.
PCB’s for the interrupter are on their way, most of the components are available locally. Only problem is fibre optic connector - can’t find it anywhere.
-
The fiber optic connector is from Industrial Fiber Optics. You can get it from digikey
https://www.digikey.com/products/en?keywords=IFE97
The plastic fiber is not easy to find. It might be better to order both the IFE97 and the plastic fiber from Industrial Fiber Optics directly, but it is more expensive.
If you want to attach your own fiber optic output rather than using the LED, there is a port J15 with +V, zero volts, and a digital output. You can build a simple circuit with a NPN transistor and the LED of your choice instead on a little external perfboard with literally three components (resistor, transistor, and LED).
Dan
I studied loneoceans’s instructions on initial power up of the board, nowhere did he mention that GDT’s must be connected, he even mentioned that if there is no GDT’s connected, there should be no signal on the outputs.
So, I gathered enough courage to power up my board, I set my power supply to 24V DC at 200mA max.
To my relief, blue LED came on, and there was no smoke leaks anywhere, current draw 0.02 A.
So far, so good :)
While I was procrastinating whether to power it up or not, I built an enclosure for it.
PCB’s for the interrupter are on their way, most of the components are available locally. Only problem is fibre optic connector - can’t find it anywhere.
-
If you look on eBay you should be able to find some cheap skinny spdif optical cables for audio use. The 2.2mm (I think) diameter is suitable for these POF optical links, especially short ones
-
Ok, I ordered those parts from Digikey and Ebay, few other pieces to get free shipping too :)
-
I think it would be a good time to start considering selection of IGBT for this project. I just measured frequency of my secondary coil, and it comes at 365 kHz.
This is just the coil, no top load, no streamer simulation ( FG, resistor and the scope technique )
With a 12” toroid as a top load, JavaTC shows that frequency should drop to about 260 kHz.
A well written selection guide (over at Kaizer :) ) makes me think that IXGN60N60C2D1 would be a good start. Good choice ?
-
I just connected the three indicator LED’s ( signal, OCD and power), powered up the board and to my surprise the OCD light came on as well as the power light.
Is this normal ? Board still has no Inductor, C33 or any of the jumpers installed. I am hoping that is the reason for the OCD trigger, otherwise I made a mistake assembling it.
-
I just connected the three indicator LED’s ( signal, OCD and power), powered up the board and to my surprise the OCD light came on as well as the power light.
Is this normal ? Board still has no Inductor, C33 or any of the jumpers installed. I am hoping that is the reason for the OCD trigger, otherwise I made a mistake assembling it.
It sounds perfectly normal. OCD should trip until it gets a reset signal from your interrupter.
-
Thanks Mads ! I started to get a bit nervous to tell you the truth. Now I can continue building my interrupter.
I’d like to be able to understand how this thing works in detail, but desire is just not enough sometimes, that way I would not worry about silly things.
Would UD2.7 be able to drive this:
https://www.easternvoltageresearch.com/flexibrute_halfbridge10.html
Seems like a neat setup, everything is there including the PCB - love working with PCB’s because it minimises the possibility of making a wrong connection.
Decent IGBT’s too...
-
I did not install 2 components on my board Slot-7 Phase Lead Inductor and C33 Capacitor.
Would someone please be able to tell me value of the capacitor if the board is going to be used to drive a full bridge based on IXGN60N60C2D IGBT's ?
Also the inductor is a complete mystery to me, there is no info on inductance, just Slot-7 from Coilcraft. Is there anything else I can use and what inductance value should it be ?
-
Interrupter is almost complete, Attiny85 is missing because programming is a bit more complicated than I expected.
But, it is partially functional without the Attiny, enough to check if UD2.7 is working correctly.
I connected them together using a bit of fibre optic cable, and to my horror UD did not react to the input ! Input signal LED stayed firmly off.
After a bit of scoping and tracing the signal I discovered that I did not solder one side of D13.
Note to self: Missing solder joints are bad for functionality
After I soldered the diode, signal LED came on and started flickering in sync with the interrupter.
So far I confirmed that the UD powers up ok, and receives the signal.
I need to check if the outputs for GDT’s work correctly, is it possible to do that without C33 and the inductor in place ?
-
Update:
Today I connected a couple of GDT's to the UD aaaaand I got mixed results. Good news is that half of it is working - GDT 2 works as advertised, GDT 1 - nothing !
Let's start from the beginning:
UD is supplied with 24V DC, connected to profdc9's interrupter via IF95T fiber optic receiver.
65-70kHz signal is supplied to the feedback connector.
I made 2 GDT's for testing purpose only and connected one to each output.
Board powers up fine, power LED comes on, signal LED turns on as soon as the interrupter is turned on.
( Earlier OCD LED would turn on too, it would turn off as soon as the signal is supplied to the board - just like Mads predicted, but since I fixed that solder joint on D13 it does not come on any more )
Image 1 shows output from GDT 2 which looks fine, and output from GDT 1, which is not so fine.
So, I did some signal tracing...
GDT 1 is driven by a pair of Mosfets, U1 and U2
Mosfets are driven by IC9, UCC27423
U1 - is driven by pin 7
U2 - is driven by pin 5
I compared the output signals on pins 5 and 7 on both UCC's ( IC9 and IC10), and discovered that pin 5 on IC9 outputs strange waveform ( pin 7 on IC9 outputs the same waveform as pin 7 on IC10, so probably it works fine ).
Image 2 shows output from pin 5 on IC10
Image 3 shows output from pin 5 on IC9
So, I think that I have faulty UCC27423. I do not have any spare ones, but I am more concerned with removing the existing one without a desoldering station.
Image 4 shows the offending part. At least solder joints look good....
By the way, as far as GDT waveform is concerned, do not ask me how I did that because I have no bloody idea.
I even started a separate thread to document my attempts - mostly failures to produce a decent waveform from a GDT.
-
Beautiful assembly job!
By the way, if you attach a 3.5 mm female plug to the Sound +/- input, you can try out the sound interrupter feature.
Make sure you adjust the minimum Ignore Interval and Pulse Width when you use it so that you don't try to pulse the coil too often or too long. Start with a short pulse width if you have any doubt.
Also, you can hook an earphone up to the digital + output if you want to listen to your interrupter before you use it with the coil, or you can look at it on your scope.
Also, you can solder header pins into the ISP connector if you want to use the board with a AVR programmer directly so you don't have to use a separate fixture.
Dan
Interrupter is almost complete, Attiny85 is missing because programming is a bit more complicated than I expected.
But, it is partially functional without the Attiny, enough to check if UD2.7 is working correctly.
I connected them together using a bit of fibre optic cable, and to my horror UD did not react to the input ! Input signal LED stayed firmly off.
After a bit of scoping and tracing the signal I discovered that I did not solder one side of D13.
Note to self: Missing solder joints are bad for functionality
After I soldered the diode, signal LED came on and started flickering in sync with the interrupter.
So far I confirmed that the UD powers up ok, and receives the signal.
I need to check if the outputs for GDT’s work correctly, is it possible to do that without C33 and the inductor in place ?
-
Happy to say that both outputs are working now ! It took some time to resolve, but it was worth it.
It turned out that UCC is not faulty, it was something much much more basic - poor solder joint on U2 mosfet.
Anyway as you can see below, wave forms are quite good, a bit of ringing but that is to do with GDT's I suppose.