NE555 interrupter and GDT problems / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

NE555 interrupter and GDT problems

first previous 4 5 6 7 next

Move Thread

LAN_403

Tonic

Sat Dec 01 2007, 09:48PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Uzzors wrote ...

The waveforms look great, now you just have to wind your GDT for 1:2:2 because currently your GDT will only give 5V of gate drive. Staircase waveforms are completely normal, and the only reason you would need to worry about them is if they exceed the mosfet's maximum gate voltage. The ground line is where the scope trace is with no voltage applied to the probe. Positive voltage will be above this line, negative below. If you switch to DC coupling on your scope you can see where the bottom part of the waveform is compared to the ground line. Low-noise resistors are of no use at these frequencies.

Thanks, that explained my few dimnesses! A while ago I did cut in half a primary wire. At first time, it worked, secondary voltage increase was significant, but when I plug dummy load to second secondary, primary waveform blurries. With only one load it works. I guess it's a matter of winding, which I'm going to solve tomorrow. If it will go according to my plans, I may have working half-bridge and maybe supply a Tesla coil, which lies in my room waiting for some power

Dr. Dark Current

Sun Dec 02 2007, 08:38AM

Registered Member #152 Joined: Sun Feb 12 2006, 03:36PM
Location: Czech Rep.
Posts: 3384

If you need a 1:2:2 GDT I suggest you to wind it 5-filar, use one winding for your primary and 2 seriesed windings for each secondary. This way you get less leakage inductance -> less ringing.

Tonic

Sun Dec 02 2007, 11:11AM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

jmartis wrote ...

If you need a 1:2:2 GDT I suggest you to wind it 5-filar, use one winding for your primary and 2 seriesed windings for each secondary. This way you get less leakage inductance -> less ringing.

In hurry I wrapped 6-filar

But, I did notice if I use two wires in paralel for primary, overshootings slightly decreases. Now I don't have any blurry waveforms with two loads, but slopes are deeper. See photos:

New GDT, smaller core and 10 turns of UTP wires:

Waveforms, 0.5 V/div, 2.2nF capacitors and 37ohm damping resistors:

Now, there are options what to do next:

1). Increase NE555's supply voltage by 4V. Datasheet says that chip can handle to 18V, so 16V shouldn't hurt. In that way the output peak voltage would be 15V, lowest voltage during on-time 10V and biggest 12V.

2). Because slopes are bigger, according to Richie's notes (

), more turns are needed or a core with higher inductance. Buying a core with higher inductance isn't a option at the moment, thus giving more turns remained. There's no space on small core, so I would use a bigger one.

3). Increase turns ratio through decreasing primary turns but I have feeling that it would cause more deeper slopes.

I think the first two options are the most reasonable, but there are next small problems. I don't have a 15V power supply and this option requires to build a regulated PSU with LM350. Time eater option. Second is very easy, but I'm not sure if I have enough UTP wire (I should have bought more..), but I'll try.

The another thing is question about driver. Is there's possibility to add another half-bridge at NE555 side? The reason I'm asking because with full-bridge the output voltage doubles and I wouldn't have to mess with 1:2:2 GDT. 1:1:1 is much easier for me. Does anyone know, how to invert NE555's signal in very simple way?

I know it's possible with TL494 (which I'm going to build next after NE555), but I want to tune fine NE555 firstly, since I hate to leave unfinished projects :)

uzzors2k

Sun Dec 02 2007, 05:41PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Try number two, it seems the easiest and most likely to work. Thanks for the link, I had never seen that article before on Richie's site.

You could use an inverting chip, or just a single transistor set up as an inverter to invert the signal. It would increase gate drive power too = better waveforms than 1:2.2.

Tonic

Sun Dec 02 2007, 06:11PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Uzzors wrote ...

Thanks for the link, I had never seen that article before on Richie's site.

Because this link was a top-secret information

For some reason Richie didn't made files public (probably lack of time). Luckily, nothing can hide from mighty Google power, which found files on Richie's webpage server

There is folder:

. There are so many useful informations in ferrite.txt, ferrite2.txt and gdt1.html (which is uncompleted). They did help a lot with my problems.

wrote ...

You could use an inverting chip, or just a single transistor set up as an inverter to invert the signal. It would increase gate drive power too = better waveforms than 1:2.2.

Would you mind explaining how to invert with single transistor? Where to plug and how it works? And do you know popular and cheap inverting chips? Not those TC44XX's or UCC's, since they're hard to get and aren't cheap :)

By the way, I did manage to IRF820 half-bridge working. I've tested with 18V and 30V supply. Works fine, as so far, without heating. And I have three question regarding my half-bridge:

1). The output of half-bridge is a square wave, right?

2). When I place HV coil on flyback transformer core, waveform gets messy and I don't see any arcs. What are possible reasons? And is it possible that electromagnetic field may cause distortions scope's measurements?

3). How I can scope primary waveform? Ground lead to NE555's or half-bridge's PSU ground line, and other probe lead to one of two ends of primary?

And for bonus, photo of whole config:

uzzors2k

Sun Dec 02 2007, 09:55PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

It's near the bottom of that page. Send the 555 output into it, and the output from the inverter into a new BD1XX stage. When I said inverter chip I meant any inverting logic chip, such as the 4049. The output from a logic chip must be boosted by a transistor totempole.

Tonic wrote ...

1). The output of half-bridge is a square wave, right?

Yes.

Tonic wrote ...

2). When I place HV coil on flyback transformer core, waveform gets messy and I don't see any arcs. What are possible reasons? And is it possible that electromagnetic field may cause distortions scope's measurements?

What are you scoping? Have you tried pulling an arc with the flyback output? It might just be very small.

Tonic wrote ...

3). How I can scope primary waveform? Ground lead to NE555's or half-bridge's PSU ground line, and other probe lead to one of two ends of primary?

Ground to one end of the primary, scope lead to the other end.

Tonic

Sun Dec 02 2007, 10:35PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Now half-bridge is working fine, but results doesn't impress after seeing arcs drawn by ZVS flyback

With 30V (peak to peak) I could stretch arc from 0.5cm to 3cm.

But it works, it's the most important thing. IRF820 were slightly warm and both of them had very similiar temperatures. Finally!

Again but.. I swapped IRF820 with IRFP250. The new ones weren't even warm, thanks to much smaller Rds(on). Next, I did plug a autotransformer to 230V/24V transformer and raised main voltage to around 300V. Arc was better, but moment later, after some testing without +30V supply, I forgot to remove oscilloscope probes from gates and sources. The capacitance of probes seems to disturb half-bridge work completely.. so, when I was removing them (after plugging +30V) I accidently pulled out wire from drain and again touched, making a big bang from +30V PSU's bigass capacitor. I'm kinda proud, my first dead transistor in half-bridge

The bang caused to open transistor all time and a lot of current was drawn from BD's and gate resistors, killing them also. I've replaced them and got it working. Only one IRFP250 is dead.

Thanks for link, seems that I might be able to understand this. I'm going to replace transistors and add half-bridge for NE555 and power section.

uzzors2k

Mon Dec 03 2007, 10:14AM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Congratulations! Did it work to run it from mains?

You should adjust the turns ratio according to voltage, just use the formula on thedatastream's website. For reference I use 30 turns at 100kHz when running my flybacks offline. At 30V that's only 15V applied to the primary, less than 1/10 of full mains power. The turns-ratio can be decreased almost as much. For low voltage supplies the Mazzilli driver is totally superior to a half-bridge, as you get pi*supply voltage applied to the core, greatly increasing the output power. From direct mains pi*supply voltage would kill most mosfets so the advantage is lost.

Tonic

Tue Dec 04 2007, 10:52AM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Not yet, I want to add few changes firstly.

At this moment, I have no idea what resistors use for transistor inverter. The transistor I have is BC337 (NPN 45V 800mA). I know that resistor between +12V and collector is supposed to limit current flowing through collector-emitter. I gave there 1kohm (12V/1000ohm=12mA). What does a resistor in series with base, I don't have much idea. Probably limits current draw from input, but NE555 should give much current as BC337 wants, so I skipped this resistor. In that way I don't get proper inverted squarewave between 1kohm resistor and collector, it's a rather flat line with small spikes there where are switches. Like as something is shorting or too big resistance. What values would you suggest?

Also, I have two ideas to get two signals for full-bridge of BD's, below you have schematics. Is there's practical different between them in terms of outputs?

scheme nr.1 :

scheme nr.2 :

Steve Conner

Tue Dec 04 2007, 11:32AM

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

The base resistor is needed otherwise too much current will flow from the 555 and destroy the transistor, and maybe the 555 too. 10k is a usual value. If this makes the switching too slow, don't make the resistor smaller: put a few hundred pF capacitor across it instead. Or choose a different transistor with less charge storage.

first previous 4 5 6 7 next

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