Gate drive signal / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Gate drive signal

first previous 2 3 4 5 next last

Move Thread

LAN_403

TwirlyWhirly555

Mon Oct 28 2013, 08:15AM

Registered Member #4104 Joined: Fri Sept 23 2011, 06:54PM
Location: Uk .
Posts: 122

I would suggest you rewind the GDT so that all 3 windings are over each other and trying a DC source such as battery

IMG_0084[1] by TwirlyWhirly555, on Flickr

This is my setup using the same drive IC's as you driving a IGBT brick similar to you , Supply is 12 Volts DC from a SMPS . Gate drive IC's have 50uF / 0.1uF caps across their supplies .

NewFile0 by TwirlyWhirly555, on Flickr

Hydron

Mon Oct 28 2013, 09:53AM

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

Are those big 2.7 ohm resistors wirewound? If they are then they'll add inductance (not good), and would be the first thing I'd change (use carbon or metal film).
I'd also look at using something like the pairs from a CAT5 cable to wind the transformer, using one wire from each pair as a primary and the other as the drive to a gate. For my DRSSTC I had some big cores so i just wound the whole cable through, sheath and all. The close proximity between the primary and secondary windings increases coupling and is normally a good thing for a gate drive transformer. Obviously you need to parallel the multiple primaries up with the right phasing. Twisting the wires to and from the transformer will also reduce stray inductance.

Edit: The above picture looks like a good solution, I just suggested CAT5 because it worked well for me, and I needed all four pairs.

Second edit: I highly doubt the silpads underneath the IGBT brick are necessary. Bricks have internal ceramic insulators and are designed for direct mounting to a heatsink (with thermal paste).

Dr. Dark Current

Mon Oct 28 2013, 10:40AM

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

Sorry for not reading the whole thread, but 1 uF on the supply rails of the drivers is NOT enough. Add at least a 100 uF electrolytic cap. I'm pretty sure the oscillations will disappear.

Ben Solon

Mon Oct 28 2013, 11:13AM

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

Dr. Dark Current wrote ...

Sorry for not reading the whole thread, but 1 uF on the supply rails of the drivers is NOT enough. Add at least a 100 uF electrolytic cap. I'm pretty sure the oscillations will disappear.

I agree. A low esr cap around 1uf is needed right next to the ic, but much more is needed on the bus.

If the supply is clean and the logic signals to the driver ic's doesn't contain any sine wave component of that frequency, but you still have that signal when floating the outputs, then the drive ic's themselves are faulty. Is it convenient to swap out new drivers straight from the anti-stat package? Even better, if there are any pin for pin equivilant ic's from another manufacturer.

furnace

Mon Oct 28 2013, 12:42PM

Registered Member #4992 Joined: Wed May 23 2012, 03:57PM
Location:
Posts: 108

I'm receiving my new ics tomorrow than I wil try all the above steps that were mentioned. I actually tried winding a large gdt increasing capacitance and a few other things and I also came to the salutation that the ics had the fault because after a lot of effort I could not get the signal to even change just a little bit.
Wil definitely post a pic tomorrow when I receive my new ics and thanks for everyone's input in this.

Proud Mary

Mon Oct 28 2013, 02:39PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

Poor supply regulation is a common source of instability. Where a supply is overloaded, the supply voltage will be pulled down with every pulse, which is then transmitted into every part of the circuit being supplied. If that voltage pulse should then be amplified by any part of the circuit dependent on that same supply, oscillation may result.

What is to be done, as was famously asked?

Decoupling: Take a 1000 ÂµF or larger electrolytic and wire it in parallel with a 100 nF and a 10nF, keeping wiring short. These two smaller values will encourage higher frequencies to go to Ground, shorter wavelengths that might otherwise have been impeded by all those tightly packed foil windings and chemical goop inside the electrolytic can. Strap this capacitor assembly across the supply right up close to the devices.

That's it.

furnace

Mon Oct 28 2013, 06:34PM

Registered Member #4992 Joined: Wed May 23 2012, 03:57PM
Location:
Posts: 108

My 15v regulated supply runs with 2x 470uf 400v electrolytic caps in series which is connected with 0.33uf , 0.1uf plus 0.22uf 500v on both ends of the com and ground side. These were the largest electrolytics i could find, but i will definitly try to get larger caps to see if this will help and i will definitly try what twirlywhirly555 suggested becuase i know this also made a big difference on my inverters coupling transformer when all the wires are winded on top of each other.

Proud Mary

Mon Oct 28 2013, 07:10PM

Registered Member #543 Joined: Tue Feb 20 2007, 04:26PM
Location: UK
Posts: 4992

furnace wrote ...

My 15v regulated supply runs with 2x 470uf 400v electrolytic caps in series which is connected with 0.33uf , 0.1uf plus 0.22uf 500v on both ends of the com and ground side. These were the largest electrolytics i could find, but i will definitly try to get larger caps to see if this will help and i will definitly try what twirlywhirly555 suggested becuase i know this also made a big difference on my inverters coupling transformer when all the wires are winded on top of each other.

Put a scope probe on the 15V line when your circuit is running so we can see what's going on there.

By the way, why have you connected the two 470 ÂµF 400V caps in series? If they are across the 15V line and Earth, they would smooth much better in parallel.

Shrad

Mon Oct 28 2013, 09:15PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

2x470ÂµF in series will give you 235ÂµF with an horrible esr

at least use a couple 1000ÂµF 25V caps from an old PSU so you have something half decent... .22ÂµF is not the smallest you should put... 100nF and 10nF as stated would be better

in that case this is not event the fact that you have a serie of caps of decreasing values which makes decoupling work, but the fact that your caps are decoupling at frequencies you see in parasitic oscillations

usually 1000ÂµF is there for smoothing DC, and 100nF and 10nF span the higher frequencies and gets them to the ground because they are low impedance to the ground at these frequencies

if you still have a ringing, try to measure the frequency by deducing it from the time divisions of your scope and find the corresponding C or RC value to add a decoupling at that frequency (it behaves like a RC low pass filter)

furnace

Tue Oct 29 2013, 04:38PM

Registered Member #4992 Joined: Wed May 23 2012, 03:57PM
Location:
Posts: 108

This is the schematic im using. I replaced the electrolytic caps with 2200uf 35v and the last 39000uf are also replaced with the .22uf 500v. Ive connected everything and got a nice clean dc wave and also a clean pulse wave from my tc chips when left floating. When i connect the the gdt the signal looks alot better than before, but when i start my brick the same problem is stil there and when i rewind the gdt so that all the wires are running on top of each other the signal gets alot worse.Another problem i have on the positive rail of this 15v supply is the line connecting the caps to the possitive rail, at the end it says iverter ground but when connected it blows my regulators so i left it floating.

1383064727 4992 FT158270 Inverter Power Supplies

first previous 2 3 4 5 next last

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