SSTC Control Board / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

SSTC Control Board

Move Thread

LAN_403

JKowalski

Wed Jan 19 2011, 10:03PM

Registered Member #3595 Joined: Mon Jan 10 2011, 04:46AM
Location:
Posts: 26

Hi,

I am putting together a medium sized SSTC. I thought I would throw out my basic plans for the project and see if you guys can give me any useful feedback.

First, basics. The secondary I will use resonates at around 500 kHz without a topload. This coil will be using a phase locked loop to keep the frequency in line. Feedback is taken from a current transformer off of the secondary return path. This method has worked very well for my smaller class E SSTC.

I haven't decided on the bridge topology yet. Half bridge or full bridge?

The main point of this topic, however, is the control board. I am trying to design many features and protective circuitry into it, with emphasis on versatility so that it can be used in more than one situation.

The mode of operation can be switched from VCF to PLL. The HCT4046 is used for the PLL. The output of the 4046 is fed into an adjustable dead-time circuit, which can be modulated by a signal/DC voltage for power PWM control. I have a couple circuit ideas for this. I may try tracking down a programmable 8-bit delay line chip (some go to 1900ns... at 500kHz I need adjustable in the range 0-1000ns). Controlled by a PIC, this would give me very precise control and by sampling audio (8-bit, not gonna win awards but the coil's gonna be noisy anyways), audio modulation. Otherwise I have some discrete voltage controlled circuits to try if I decide against the chip/can't find a good relatively cheap one.

The coil will have a 4-digit 7-segment display of frequency, controlled by a dedicated PIC. (this is a separate board) This PIC will take in the buffered, cleaned up feedback signal and the 4046 output and count them over a certain time period occasionally. The count is displayed (for the 4046) and if the FB and 4046 counts are not equal (within a margin of error) then the chip signals to the control board that the PLL lock is bad. The 4046 has a lock indicator, but I am using the type I comparator not type II.

The control board PIC will handle the error signals and provide protection. I am still working out how to set up this section. It will be controlling the 4046 VCO disable signal/the gate driver EN signals, and a relay for the bridge B+ (safety measure, normally OFF. manual triggering needed to turn on. Also gives more flexibility to the control board if needed.

ERROR SIGNALS (working out which combination of errors cover all potential problems facing optimal operation + bridge protection)
- Over temp (temp sensor on heat sink) - > Disable EN/INH to pause SSTC operation (manual reset needed), light OVERTEMP indicator
- PLL lock failure - > Disable EN/INH to pause SSTC operation (manual reset needed), light PLL CANNOT LOCK indicator
- No voltage on B+ (opto maybe, means main fuse blown) - > light BAD FUSE indicator, Disable EN/INH, disable relay. Await new fuse/repairs
- Peak Overcurrent? -> open relay, disable en/inh, light overcurrent indicator

Some pics of what I have to work with so far (webcam quality sorry):

2iw29mf

The control board will be done with mainly SMT components. I will bend a small aluminum frame to house it (it's going in the box pictured along with the bridge) and shield the circuitry from interference. The power will probably be drawn from a DC jack, while the bridge has a IEC outlet of it's own, to provide a little more isolation.

Finally, the SSTC will be compatible with a variac.

Thoughts?

[Edit: Huge picture]

HV Enthusiast

Thu Jan 20 2011, 01:11AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Any particular reason you don't want to use a DRSSTC type topology? You'll get the most bang for the buck going the DRSSTC route as opposed to the standard SSTC route.

JKowalski

Thu Jan 20 2011, 01:33AM

Registered Member #3595 Joined: Mon Jan 10 2011, 04:46AM
Location:
Posts: 26

EastVoltResearch wrote ...

Any particular reason you don't want to use a DRSSTC type topology? You'll get the most bang for the buck going the DRSSTC route as opposed to the standard SSTC route.

Well, I wanted to be able to run it CW. I'd rather have a quiet, med spark length coil with good versatility.

I am going to try an interrupter at some point to alter spark characteristics, and the advantage of using the PIC is that such a feature can be loaded into the software easily. However, I do not want it to be a mandatory part of the coil's operation.

I have not had much luck finding a cheap programmable delay line. So I think I will turn to discrete methods. This is one of my ideas:
8vyfyf

The drive signal is low passed to create a triangle wave from 4.9-0.1 volts approx, which is compared to a 0-5v control signal. Basic PWM really. This pwm is put through a AND gate with the drive signal. The PWM pulse will always be centered on the falling edge of the drive signal, so therefore the AND gate cuts out the half of the PWM where the drive signal is dead and voila, 0-100% 0-5v controlled dead time. Only problem is that when frequency changes considerably the 'triangle wave' loses it's shape. For small deviations this shouldn't be a problem though.

If you have any suggestions... Or any info on circuits that might help me out.

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