Joan (of Arc) Small uP Controlled Battery Powered DRSSTC / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Joan (of Arc) Small uP Controlled Battery Powered DRSSTC

Move Thread

LAN_403

Extreme Electronics

Fri May 29 2009, 09:14PM

Registered Member #74 Joined: Thu Feb 09 2006, 09:17AM
Location: Nottingham UK
Posts: 99

Joan (of ARC)

Joan is a portable uP Controlled DRSSTC similar to my Sprite but with many improvements.
These include:-
Microprocessor controlled.
Full Bridge Driver
Primary Current feedback
Computer Control via RS232.
Over Current Detect

Primary: 4T 5mm copper pipe.
Secondary: 200mm tall, 1500T enameled wire former 55mm diameter. wound with 0.125mm Wire
Topload: 40mm x 400mm Toroid.
Frez :~180Khz

Joan (of Arc)
Joan

Joan runs from a pack of 10 NMH AA cells giving a 12VDC supply. This is fed into a small 75W inverter to give 350V DC bus voltage to the Bridge.

The Coil is controlled by a PIC 16F690 microcontroller. This controls:- Burst Length, Burst Repetition rate and Number of Bursts.
There are 5 Banks of these settings (of Burst Length, Burst Repetition rate and Number of Bursts.
) which are selectable from two push buttons on the case. The two push buttons for controlling the coil, one is a "setting bank" button which allows the selection of the bank as above and one is a GO button.
The settings in each bank can be programmed and stored from the RS232 interface.
The Bank selection is displayed on a flashing LED on the top of the case.

This Controller also allows the Coil to be fully controllable and programmable via RS232 from a PC (see opto-isolator below).

The controller also has the ability to implement Over Current Detect (OCD) by using the micro's onboard comparator to take a coltrol voltage (set from a preset) and compare the output from a CT

Microprocessor and TC control
Micro

The coil is controlled by a PIC 16F690

This allows the usual DRSSTC functions (including the interrupter) to be achieved in only 4 Chips.

The two chips in the top right hand corner were a pair of microchip drivers that would not supply enough current and are now replaced by a pair of IXYDD414 drivers on the small board bottom right.

The board takes its input from two push buttons for control and from a CT to give OCD.

The variable is to set the max OCD current level via the onboard comparator. At present this is disabled due to noise issues.

There is a 6 way header at the bottom of the board to allow for in circuit programming of the microprocessor.

Circuit Diagram is HERE

12V-400V Inverter
Inv

12V is not enough for driving a decent spark !

So the 12v from the battery pack is inverted to 350VDC by a SG3525 PWM chip driving a pair of IRF460 MOSFETS in push pull.
The output from these goes to a single 25mm Toroidal ferrite.
This Ferrite ring has 3t-0-3t for the primaries and 125turns on a secondary. The output is then rectified and stored in a 400uF capacitor under the bridge. The whole inverter runs at 50Khz and is voltage controlled to save battery life. When the capacitor is charging the inverter provides around 75-90W of power.
On the top of the case is an LED that indicates when the 400uF cap is fully charged. (350V at present)

Bridge
Bridge

Bridge and GDT

The bridge is made up of 4 x STGW30NC120HD 1200V 30A (135A pulsed) IGBT's. I am of course pushing these up to 200A with very short burst lengths.

You can also see on of the pair of CT's to the left, One gives a phase signal to the driver and the other one is for over current detect.

The main reservoir capacitor is under the bridge and is bypassed with a 0.1uF @1Kv.

The drive is via a gate drive transformer which has one 12T primary and four 18T secondary's this ensures that I get a solid +/-15v gate drive which is clamped by 15V zenners on each IGBT gate.

Primary Cap, Ct's and Feed
The Primary is fed via a ~0.4uF capacitor made from an assortment of smaller caps, this allow me to tune the coil by adding or removing capacitance.

The feed to the primary is via a set of double sided PCB "feeders" which reduces the impedance.

Two 100:1 CT's give Phase and OCD information to the main board. The OCD is fed via an integrator to the comparitor and A2D the decay of the integrtor allows the uP to get an a2d conversion afte each burst to display Max current.

Inside
Complete Inside

This is how the whole lot fits together. into a single case 300mm long, 200mm wide and 55mm deep.

There is now the addition of a charging socket on one side of the case so I can charge the coil without taking the batteries out.

Opto Isolators

Tesla Coil End

This box is powered from Joan itself and is a plug in extra.

It allows a degree of isolation (200mm worth) from the tesla coil itself so I can safely attach a laptop to program or control Joan

On the transmit side there is a single transistor that takes the PIC's output and drives a high speed transmitter LED.

The receive side takes the signal from the opto-diode and amplifies it with a single HEF4049. This is then fed directly into the PIC micro controller.

PC End

This is the PC-Side of the opto-isolator. As the (cheap) RS232-USB dongle I use only supplies 5V, I need to have a small battery to give the current for the transmit LED. this is achieved with a single transistor.

On the receive side, to ensure the isolator is compatible with "real" RS232 the signal from the opto-diode goes via a HEF4049 amplifier and a MAX232 chip to give +/-12V to the PC.

Both units together will give a throughput of 19.2Kbd isolated to greater than 80Kv.

Sparks

Spark

Joan will produce 12"+ arcs to air using a high repetition rate (1Khz) over 10 bursts. Unfortunately at this rate the capacitor needs to be re-charged completely which can take 2-3 seconds.

Using variations of different burst lengths, repetition rates and number of repetitions smaller discharges can be maintained with faster recharge rates.

more spark pictures here

Pc Software in development to give musical drive, ideally from single channel midi files. I was looking at using Flash, just for the irony, but flash dosent have any direct control over serial ports.

Circuit Diagrams
Main uP Board (to do)

Inverter Circuit

Serial Commands
The current command descriptions are here.

Each command is a single (capital) letter followed by 4,2 or no hex digits.

A Screen dump is here

Eventual Aim
To create a uP based DRSSTC driver board and software capable of running many different designs of DRSSTC's with full control, in real time from a PC.

Still to Do

PC / Linux Software development to allow music and other effects.

A2D / Comparitor Over Current Detect (OCD) - BAD! noise issues at present.

Improve inverter to give 100W plus.

Circuit Diagrams

Derek

Bennem

Sat May 30 2009, 07:27PM

Registered Member #154 Joined: Sun Feb 12 2006, 04:28PM
Location: Westmidlands, UK
Posts: 260

I saw this little coil functioning today at the Derby Teslathon,
its an impressive little coil!!
Derek was able to set different bursts etc from his laptop
giving 12 inch spark to air, and only from 10 X AA type rechargable batteries!

I want one!....lol

Dalus

Sat May 30 2009, 09:28PM

Registered Member #639 Joined: Wed Apr 11 2007, 09:09PM
Location: The Netherlands, Herkenbosch
Posts: 512

Seen it on his website. Great small coil and nice that you are able to implement uC control near such a coil :)

Reminds me I still want to attend a teslathon, sometime . . . .

Robert2

Sun May 31 2009, 06:11AM

Registered Member #1773 Joined: Tue Oct 21 2008, 06:56PM
Location: Poland
Posts: 93

Hello, You can give me circuit for 12->350V inverter ?

Extreme Electronics

Sun Jun 07 2009, 10:20PM

Registered Member #74 Joined: Thu Feb 09 2006, 09:17AM
Location: Nottingham UK
Posts: 99

Good News, Ive found the problem that has been stopping my OCD circuit from working, a solder 'hair' between tracks. The OCD now functions, but It apparently is telling me that I'm getting >200A pulses from my IGBT's so something is still not right.

I'm in the process of documenting the circuit diagram (as it is now), progress is slow as Ive started to use Eagle, last time I used this software it was called boardmaker and ran in DOS 5, so I have some learning to do..

The First Circuit diagram, The Inverter, is here Inverter Circuit

Derek

Extreme Electronics

Thu Jun 11 2009, 02:29PM

Registered Member #74 Joined: Thu Feb 09 2006, 09:17AM
Location: Nottingham UK
Posts: 99

Update,
The main board circuit diagram is here

Top Post Updated..

Much of the diagram is based on Steve Wards New DRSSTC Driver, simplified, microprocessorized and then re-complicated

..

The OCD problems are mostly solved, One issue remains. At high powers the OCD trips well before the set comparitor voltage, Scoping the comparitor voltage shows some spikes, so Ill be adding a load of caps around the pot and see if that improves.

Solving The OCD issues ment that I can now get burst by burst Max Primary Current or Bus Voltage readings in real time ( but in hex at the moment). I did want to take measurements mid burst, but unfortunatly the A2D is not fast enough on such a small HF coil, so I have changed the OCD circuit by adding a diode so that there is a slower decay of the OCD voltage accross the filter cap. This Allows me to sample the voltage on the cap after the end of the burst with a fair degree of accuracy, Without slowing down the responce of the OCD comparitor.

(To DO .. Add Battery voltage monitoring with the spare A2D input..)

I Do need to convert the readings in to a human readable form (which takes a 16bit x16bit division and conversion to BCD) which I can now do, as this will now be after the time critical in-burst period.

Serial Control
Tty1

The screen shot above is the output from the tesla coil in a terminal emulator

The first block is a Status display
BurstL (Burst Length)= Hex Value = Value in uS
RepR (Repitition Rate) Hex Value = Value in Hz
MaxB=00 of 10 (Burst count , Set to do 10 bursts)
ClrDel (is the delay after a burst that the latch is cleared) roughly 10=~1uS
BSel is the currently selected settings bank.

The bottom block is the output from running the coil.

'Go' Indicates that a you have asked for a burst of MaxB bursts with a length of BurstL at RepR
the hex numbers at then the A2D output of the current divide by ~328Decimal to get a current reading (about 150A in this case.)
As the Hex current readings are output directly from the A2D interrupt they do destroy the formatting of the output at the moment.

'Go Once' is the same as above, but with only one burst. Hence only the one current reading.

The current command descriptions are here.

Each command is a single (capital) letter followed by 4,2 or no hex digits.

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