My first DRSSTC / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

My first DRSSTC

1 2 3 next

Move Thread

LAN_403

futurist

Thu Jan 05 2017, 07:07PM

Registered Member #61400 Joined: Sun Jan 01 2017, 01:01PM
Location:
Posts: 33

Hi everyone,

I'm new here and I come from Croatia

I wanted to have a Tesla coil since I saw one at the tehnical museum as a kid, so a few months ago I decided to try to make a DRRSTC

Now after reading most of what I could find about DRSSTCs on the Internet I need some help finishing the coil

I was stupid enough to start with the UD2.7C driver so this turned to a quite difficult project for me, I have little experience in electronics (studiying chemistry) so excuse me if I ask stupid questions :)

After receiving UD27C PCB from loneoceans I went to source all the other components - for the IGBT I decided to try the CM200DY-24NF
which was the cheapest IGBT brick I could find (62 $ with shipping from aliexpress), so this led to a halfbridge medium sized coil with 240 V AC voltage doubler.

The design I have choosen is something between loneoceans DRSSTC 2, Steve Ward's DRSSTC 1 and Kaizer's DRSSTC I

After 3 months worth of weekends I finally put most parts together, these are the specs:

Topload
130 x 620 mm from flexible aluminum tube with 2x 400 mm diameter stainless steel serving plates

Secondary coil
160 mm diameter on PVC drain pipe, ~2200 turns (~605 mm long), 0.25 mm double enamelled copper wire
2x polyester resin coated (epoxy is around 10x more expensive here), turned pretty bad with lots of bubbles but ok for my first try

Primary coil
8 mm diameter soft copper tubing (1 mm thick), 257 mm inner diameter, ~9 turns with 10 mm spacing (too large inner diameter?)
3x 3 mm plywood supports glued together with PVA, used metal screws to secure them

Electronics

Input 240 V AC with the voltage doubler
KBPC5010 1000 V 50 A bridge rectifier with 2x EPCOS 4700 uF 400 V caps
All bridge connections made with 15 x 3 mm copper bus bar

MMC air-cooled 12x 942C20P15K-F capacitors with 10 Mohm 1/4 W balancing resistors for 0.45 uF at 4000 V DC

Halfbridge CM200DY-24NF sitting on air-cooled 1.5 kg copper plate, IGBT snubber 1 uF 1200 V + 0.15 uF 2000 V DC parallel on the bus capacitors
Old CPU cooler fit nicely under the copper plate allowing IGBT to be mounted horizontally

0.10 uF 2000 V DC capacitor from negative rail to grounded copper heatsink
10 kohm 20 W resistor is connected accros the output of the inverter
Connections bridge to MMC and to primary are made with the 16 mm^2 welding cable

Universal driver 2.7C with 24 V switching PSU, 2x 1:~625 CTs wound on high permeability core
13 turns 1:1:1 GDT N87 core with ferrite bead, gate resistors 4.7 ohm 1/2 W, gates and emitters shorted with 1.5KE33CA TVS (Are the TVS ok? It was the only one I could buy locally)

I have calculated the power rating for the gate resistor to be 4.7 W - it seems high and I have just 1/2 W resistors available, can this cause problems?
formula: P = 1350 * 10^-9 C * 50 V * 70000 s^-1

Bus precharger and fan speed controller (LM35-DZ temp. sensors) are made with Arduino nano 328A, 2x 30 A relays and 150 ohm (charge) + 1 kohm (discharge) 10 W resistors.

The interrupter was made by my friend - it's Arduino nano 328A with modified oneTesla interrupter code for playing MIDI directly from USB + 10 m ST-ST 125/62.5 um optic cable

I made a mistake with the primary radius while making it so the coupling coeffitient is 0.128 vs recomended 0.136
JavaTC output is below

Are my specs ok and is this going to work?

Any advice is appreciated

zzz_julian_zzz

Fri Jan 06 2017, 05:31AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

Hi and Welcome..

For a first timer, I can say that your work is impressive :) my opinion is that it is ready to be tested and fired up.

Here are some expectation from your end before operating:

1. to have idea on SSTC and DRSSTC works
2. to ensure the UD2.7C is working properly using interrupter function gen and oscilloscope
3. to ensure you have checked the GDT phasing on the bridge, via scope (check V max, waveform shape, frequency, phase lead)
4. to slowly ramp the AC input voltage on your first run (test first only Â¼ of the max voltage) â€“ VARIAC will help
5. to monitor the Primary Coil current thru winded CTs(with burden) â€“ using oscilloscope
6. to monitor the Primary frequency to be close as what you have simulated, check the wave shape â€“ remove notching later for higher power transfer
7. to activate the OCD Protection leveled to a not so high current limit
8. to ensure your interrupter is working ok - verified via scope, start to low BPS and from 0 us On time
9. to use break out point
10. to install ground strike rail

here are some answers to your Q:

1. Prim Inner diameter is ok, coupling should be your consideration (0.128 is fine) â€“ just start with low V input at first run to check whether the two coils flashes over, then ramp slowly
2. TVS 33V are ok â€“ I donâ€™t use these actually
3. 1 to Â½ watt gate resistors are ok â€“ what is the resistance value?

My Q:
1. Why Primary Cap is so high? 0.45 uF â€“ but as long as you are using OCD protection and you are monitoring the current via oscilloscope on your first run, then its ok.. Donâ€™t play too long ON Times in your interrupter as I believe your coil has very low zurge impedance.
2. What ON time you plan to use?

Hydron

Fri Jan 06 2017, 11:27AM

Registered Member #30656 Joined: Tue Jul 30 2013, 02:40AM
Location: UK
Posts: 208

As I mentioned to futurist earlier on IRC, this is a very impressive start to coiling!

.45uF does sound a little big, though it will help getting decent primary currents using a half bridge design. The DC rating of 4kV is maybe a little low, i would suggest limiting the current to ~600A to avoid going over the DC rating.

If the primary is large enough you may be able to reconfigure the MMC to 4 parallel, 3 series for .2uF at 6kV DC. Would need to move the primary tap further out, and if you run out of turns the resulting frequency will be higher than before, but this may not be a bad thing if upper pole tuning works well for your coil, and would increase the coupling a bit too.

futurist

Fri Jan 06 2017, 12:52PM

Registered Member #61400 Joined: Sun Jan 01 2017, 01:01PM
Location:
Posts: 33

For the primary capacitor I considered using 9x 942C for 0.15 uF 9 kV or 12x 942C for 0.45 uF 4 kV.
I choose the 0.45 uF variation because it is used here

and here

JavaTC says 7.3888 turns for 0.20 uF primary, I have almost 9 turns so I can try.

I'll install the strike rail, make primary tap and then borrow an osciloscope and a function generator. I couldn't find 1:1 transformer but I have 1:2 ~500 VA. Can I use it with variac?
Without an osciloscope I could only verify that the UVLO is working, the interrupter turns the signal LED on so that part also works. For the signal LED I installed a HE green LED, but at 100 uS ontime it's not quite bright, is that normal?

I marked the GDT secondary leads and connected them antiphase, but I'll doublecheck. I stripped the CAT5 cable with a lighter to avoid scoring the wire.

The interrupter on time is from 10 to 100 uS, maybe higher for the MIDI but for now I won't be using that.

5. to monitor the Primary Coil current thru winded CTs(with burden) â€“ using oscilloscope

Do I have to make a separate CT or I can use existing one and how? Can you explain shortly?

loneoceans

Sat Jan 07 2017, 01:24AM

Registered Member #4098 Joined: Fri Sept 16 2011, 09:26PM
Location:
Posts: 236

Futurist, good work on the coil. Here are some comments in no particular order:

1) Your capacitor bank is only rated 1000VAC, and in fact less due to the high frequencies weâ€™re running the coils at. This will certainly lead to premature MMC failure quickly at any sort of high current rating. Your primary has about a 5.1 ohm impedance. The CM200NF bridge should be able to do 500A but at that rating youâ€™re seeing a 2.5kApk which is far over the RMS AC rating. So you'll need to adjust the current limit for the maximum capacitor voltage. Iâ€™d recommend at the very least dropping to a 200nF 1.5kVAC configuration or have 4 or 5 of the CDE caps in series and making more strings of them. Itâ€™s more expensive at first, but itâ€™ll save you the trouble of dealing with popped MMCs later.

2) The signal LED corresponds to the actual ON signal. So 100us on-time at 200Hz Is just 2% duty cycle, so your LED will appear to be 2% as bright as itâ€™s supposed to be. I.e. itâ€™s normal for it to look dim. You can either use a brighter LED or change the series resistor (add another SMD one on top) to increase the LED brightness.

3) Iâ€™d recommend buying a proper current probe like an Ion Physics one or a Pearson on ebay if you can find a good deal. Otherwise the easiest is to just make another CT like the one youâ€™ve made, and terminate the ends of the secondary with some known resistor. You can probe across the resistor and to get a measurement of the current (I = V/R).

4) Youâ€™ll find that at low power, a roughly equal primary and secondary frequency leads to good results. At high power, itâ€™s usually good to detune the primary such that it is about 5 to 10% lower in frequency than the secondary. When the spark grows it adds capacitance and drops the secondary frequency.

5) If youâ€™re going to probe gate drive and inverter outputs to tune phase lead on the UD2.7C (as outlined in my DRSSTC 3 page here

, Iâ€™d recommend using not only a variac to ramp up power to the bus, but also itâ€™s important to keep the bus isolated for your safety and to avoid ground loops. You can use an isolation transformer for that. Alternatively, a high voltage DC power supply is good as well. Typically I tune for phase lead without the secondary in place.

6) The 1.5ke33ca TVSes are fine. You can go even lower like a 1.5ke25. 1/2W gate resistors should work fine, but why not just use 4 of them to get a 2W resistor :).

The other feedback from Hydron and Julian are great as well. Good luck!

zzz_julian_zzz

Sat Jan 07 2017, 09:16AM

Registered Member #3964 Joined: Thu Jun 23 2011, 03:23AM
Location: Valenzuela City
Posts: 332

For the caps - ok I see.. but just be careful when operating the coil, higher caps tends to create larger currents as the LC is low impedance.

For steve wards' coil, he's using very tight coupling ~0.2 - which is pretty high.. but in my test, the highest I can achieve is around 0.16 only per JAVATC, more than that and racing sparks appears.. steve's coupling could be the reason why he gets decent primary ring up (both time, shape, and magnitude) precisely tuned for his taste of operation.. I think if coupling is lower in his configuration, then there will be too much current in the primary so the user will tend to decrease the ON time, which leads to insufficient transfer. My point is that it should be balanced properly.

if you can try lowering your Pri Cap to 0.2uf then, it will be safer.. Steve's operation for his DRSSTC 1 is a bit extreme and needs experience and skills like his for it to work properly or else, it will fail catastrophically.

REGARDING THE TRANSFORMER

I assume you are referring to Isolation transformer. then yes, you could use 1:2, BUT just be aware of the voltage rating. In my practice, I am not using isolation transformer for my VARIAC >> DRSSTC bridge, instead, I use it on my scope. I don't experience any issue with it, PLUS, my coil is not limited to the ISO trans VA rating.

REGARDING THE LED

I assume the signal is from interrupter. then yes, 100 us will not make the LED bright enough.. What I do here to make it bright is just use a diode in series with 1 a uf capacitor which is then connected in parallel to LED - so that the LED will have a bit longer time/ enough energy to reach its full brightness. I am able to do this because I have current buffer (TC4427) on my signal - this IC charges the cap..

REGARDING GDT PHASE AND O-SCOPE

yes - double check everything before firing up. Just to share, I didn't have an oscilloscope too on my first DRSSTCs, what I did is download a Zelscope software on my laptop and use the AUDIO input as my channel (with some Over Voltage protections and resistive voltage dividers on my probe .,) this is how I checked the phasing of my GDTs - of course with some 10 khz generator (square wave from a 555 timer) which serves as my feedback signal.. though, I don't recommend you to do this.

REGARDING CURRENT MONITORING

yes- wind a dedicated CT for current monitoring.. you cannot use the scoping method I used above since that is limited to <20khz signals.

This CT is where you connect your oscilloscope in order to see the waveform, magnitude, behavior of the primary circuit current. It will tell you at least if there are erratic/abnormal operations in there not only in your firsts trials but also for further tuning. It will allow you to see whether there are notching happening in the current.
Of course this CT should be burdened properly for it to convert CURRENT to VOLTAGE thru Ohms law.
Installation of this CT is same on where you put your OCD ct and FB ct...

Hope this helps..

regards,
Julian

EDIT: Hear Loneoceans :), he's one of the best here..

Hydron

Sat Jan 07 2017, 10:55AM

Registered Member #30656 Joined: Tue Jul 30 2013, 02:40AM
Location: UK
Posts: 208

If you can find a 0.01V/A to 0.1 V/A Pearson "Current Monitor" or similar (other brands are "Ion Physics" and "Stangenes") for a good price then that is the best way to measure primary current with a scope. They don't come up cheaply often though, so at first I'd suggest using a cascaded CT like you did for feedback and OCD. Two 33:1 cascaded CTs with a 10R burden resistor should give approx 0.01V/A with very good frequency response (for this application the only disadvantage vs a commercial CT is less noise immunity - I made a thread about it a while ago here

).

As for measuring the bridge with an isolation transformer, i would strongly recommend against isolating the scope rather than the coil unless you know exactly what you're doing. You'll have better measurement accuracy and safety isolating the bridge, and even a small isolation transformer can be used if you keep the coil power low.

futurist

Sat Jan 07 2017, 01:08PM

Registered Member #61400 Joined: Sun Jan 01 2017, 01:01PM
Location:
Posts: 33

Thank you guys for the nice feedback, now I have more than enough things to do on the coil :)

About MMC
I don't have enough turns for the 0.15 uF and the other solution could be 6 strings of 4 capacitors for 0.225 uF at 8000 V DC
But that is another 70 â‚¬, so I'll reconfigure the MMC to 4 strings of 3 capacitors for 0.2 uF at 6000 V DC and hope they won't blow up
I never thought of the primary capacitor and coupling like that but now it seems completely logical

I'll update the thread when I finish everything

Uspring

Sat Jan 07 2017, 03:43PM

Registered Member #3988 Joined: Thu Jul 07 2011, 03:25PM
Location:
Posts: 711

Reconfiguring the MMC bank as you describe won't help. You will increase the total specified max voltage from 4kV to 6kV, but lowering the total capacitance from 0.45uF to 0.2uF implies a higher voltage on the bank by a factor of 2.25 for a given bridge current. That is larger than the 6kV to 4kV ratio.

futurist

Sat Jan 07 2017, 05:40PM

Registered Member #61400 Joined: Sun Jan 01 2017, 01:01PM
Location:
Posts: 33

What should I do?

1 2 3 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob