Current Feedback Circuit / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Current Feedback Circuit

1 2 next

Move Thread

LAN_403

Graham Armitage

Wed Aug 29 2012, 12:30AM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

I hope this is an easy one for those more experienced. The coil was running fine, but after stopping to change the modulator, the coil stopped working (no major component failure I know of). Have removed all ICs from the controller board and started debugging. The problem as I see it is that pin1 of the 74HC14 (U1) is at 1.8v (because of the clamping diodes CR5/6.

As a result when performing a test of the controller board (all chips installed) using a signal generator at 5v and 100kHz the two gate drivers are just locked - one on one off because of this 1.8v DC offset that is being carried through the U1 output. The U1 output shows the 100Khz signal with the 1.8v offset.

Have replaced U1, CR5,6 and C1 and no difference. Totally stumped...

Thanks in advance.

Ben Solon

Wed Aug 29 2012, 12:53AM

Registered Member #3900 Joined: Thu May 19 2011, 08:28PM
Location:
Posts: 600

Can we see the entire circuit? The beauty of a schmitt trigger is that you need to two two levels in order to change the state of the output. If there's 100khz output from the inverter then it implies that the output is going 0-vcc at 100khz. Is yours going 1.8-vcc+1.8? If you havnt tried to pull down the pin to drain charge(if somehow that offset is due to charge being held), then I would check your vcc and ground lines. Make sure that the u1 ground really is 0v rather than 1.8.

Graham Armitage

Wed Aug 29 2012, 01:13AM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

It's essentially Steve Ward's SSTC circuit, but I replaced the antenna feedback with the MiniBrute current transformer feedback circuit. Has been running perfectly for some time now so I know the schematic is correct.

When I run the test, the 100kHz output from the 74HC14 is from 1.8v to 5v (Vcc). When the chips are removed there is a constant 1.8v at pin 1. If I short out the DC blocking cap C1, it drops to 0v. The ground and Vcc are on U1 have been checked and are correct.

zzz_julian_zzz

Wed Aug 29 2012, 01:40AM

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

Look beneath it, there must be a unproperly soldered pin or anything... How about the modulator? did you just changed it? Are you sure, that the modulator is properly working? Are you using a D Flipflop? or just the SSTC 5 schem. In my experience, I have never made the circuit work using the CT from EVR, instead, I make my own CT..

Graham Armitage

Wed Aug 29 2012, 01:53AM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

I have been using a separate modulator and that is working fine. Actually have a couple of different ones with variable duty cycle and frequencies. For testing, I peg the gate drivers (pin 3) to 5v to force them on, thereby excluding the modulator altogether. Trying to isolate the problem. The CT is the CST206-1A from DigiKey and has been working fine running on the secondary current. So for now don't want to change that.

What I don't understand is what components are supposed pull pin 1 to 0v as a steady state? If the clamping diodes present 1.8v at pin 1, how is it supposed to fluctuate between 0 and 5v and not 1.8v-5v (with a 0v-5v signal)?

Ben Solon

Wed Aug 29 2012, 02:26AM

Registered Member #3900 Joined: Thu May 19 2011, 08:28PM
Location:
Posts: 600

Check your diodes. The thing is that there are no electronics to stabilize that node because there shouldn't be anything to stabilize. Run a sine wave of a decent amplitude through the input to a heck if any diodes are faulty. That diode config is intended to clamp the voltage to ground and 4.7 referenced to that ground. The the fact that its no clamping to ground means that there a bad diode or some connection issue.

BUT. The cap c1 is a dc blocking cap. Therefore there should be no dc component after that point unless imposed by yet another error. Everything I said above should hold no use to you because frankly it shouldn't matter after the blocking cap. Except a slight change in amplitude of coarse...

Graham Armitage

Wed Aug 29 2012, 03:21AM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

That was what I thought too - so I replaced both clamping diodes and still the same result. if I bypass the blocking capacitor the Pin 1 node is pulled to 0v, implying that the problem is after the cap (C1) - leaving a resistor and two diodes to be the culprit.

If I create a circuit (and I actually did this with just two clamping diodes and a 5v source), I measure the same 1.8v at the junction of the two diodes. Again, not understanding how the clamping configuration creates 0v at that node?

By the way - thanks very much for the suggestions and help with this. I know I will kick myself when I figure it out.

Graham Armitage

Wed Aug 29 2012, 03:42AM

Registered Member #6038 Joined: Mon Aug 06 2012, 11:31AM
Location: Salado, TX
Posts: 248

I took your advice and ran a high amplitude sign wave through the input. The first image is the wave form at the node between the two clamping diodes. It is clearly clamping, with a peak to peak of about 6v, instead of the 5v I would have expected.

the second screen is the output, which is converted nicely to a square wave as we would hope and expect, but notice the slight DC offset. That just locks the gate drivers on or off.
(scope set at 2v/div)

ConKbot of Doom

Wed Aug 29 2012, 06:07AM

Registered Member #509 Joined: Sat Feb 10 2007, 07:02AM
Location:
Posts: 329

Since you have a DC block cap, the circuit isnt pulled to anything, youre seeing a combination of the input leakage of the logic gate (some of the older logic can have quite a bit of leakage, usually pulling up, but the HC/HCT/ACT ones are all pretty good) and the leakage of the diodes forming a divider. If you set your meter for DC miliamps you should be able to measure the leakage current from pin 1 to ground, and select an appropriate pull down resistor accordingly. Or just whack a 20k -30k pull down on there and see what voltage you get. Dont get too low on the pulldown unless you have to, as you'll be making a divider with the 560 ohm series resistor.

Steve Conner

Wed Aug 29 2012, 09:02AM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

No matter what happens at the input to the logic gate, the output should be either 0 or 5. 1.8 is not a valid logic level.

So, the ground pin of U1 is probably disconnected. Either that or the next thing downstream has blown its input stage and is leaking lots of current and fighting the output.

1 2 next

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