If you need assistance, please send an email to forum at 4hv dot org. To ensure your email is not marked as spam, please include the phrase "4hv help" in the subject line. You can also find assistance via IRC, at irc.shadowworld.net, room #hvcomm.
Support 4hv.org!
Donate:
4hv.org is hosted on a dedicated server. Unfortunately, this server costs and we rely on the help of site members to keep 4hv.org running. Please consider donating. We will place your name on the thanks list and you'll be helping to keep 4hv.org alive and free for everyone. Members whose names appear in red bold have donated recently. Green bold denotes those who have recently donated to keep the server carbon neutral.
Special Thanks To:
Aaron Holmes
Aaron Wheeler
Adam Horden
Alan Scrimgeour
Andre
Andrew Haynes
Anonymous000
asabase
Austin Weil
barney
Barry
Bert Hickman
Bill Kukowski
Blitzorn
Brandon Paradelas
Bruce Bowling
BubeeMike
Byong Park
Cesiumsponge
Chris F.
Chris Hooper
Corey Worthington
Derek Woodroffe
Dalus
Dan Strother
Daniel Davis
Daniel Uhrenholt
datasheetarchive
Dave Billington
Dave Marshall
David F.
Dennis Rogers
drelectrix
Dr. John Gudenas
Dr. Spark
E.TexasTesla
eastvoltresearch
Eirik Taylor
Erik Dyakov
Erlend^SE
Finn Hammer
Firebug24k
GalliumMan
Gary Peterson
George Slade
GhostNull
Gordon Mcknight
Graham Armitage
Grant
GreySoul
Henry H
IamSmooth
In memory of Leo Powning
Jacob Cash
James Howells
James Pawson
Jeff Greenfield
Jeff Thomas
Jesse Frost
Jim Mitchell
jlr134
Joe Mastroianni
John Forcina
John Oberg
John Willcutt
Jon Newcomb
klugesmith
Leslie Wright
Lutz Hoffman
Mads Barnkob
Martin King
Mats Karlsson
Matt Gibson
Matthew Guidry
mbd
Michael D'Angelo
Mikkel
mileswaldron
mister_rf
Neil Foster
Nick de Smith
Nick Soroka
nicklenorp
Nik
Norman Stanley
Patrick Coleman
Paul Brodie
Paul Jordan
Paul Montgomery
Ped
Peter Krogen
Peter Terren
PhilGood
Richard Feldman
Robert Bush
Royce Bailey
Scott Fusare
Scott Newman
smiffy
Stella
Steven Busic
Steve Conner
Steve Jones
Steve Ward
Sulaiman
Thomas Coyle
Thomas A. Wallace
Thomas W
Timo
Torch
Ulf Jonsson
vasil
Vaxian
vladi mazzilli
wastehl
Weston
William Kim
William N.
William Stehl
Wesley Venis
The aforementioned have contributed financially to the continuing triumph of 4hv.org. They are deserving of my most heartfelt thanks.
Registered Member #146
Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055
Now what about the Negative rail of the bridge, as this is floating when driven with an isolated supply, should there not be some bypass capacitance to earth here as well?
I like to use a .1uF or so film cap between either buss rail (remember, it has a huge lytic that essentially couples RF as a short circuit) and the heatsink, which is grounded to mains ground.
Ive been debating putting a coupling cap between the secondary RF ground and the heatsink as well, this could potentially make the primary circuit arc-proof, but may introduce new problems. Its something i need to test out later.
And also does the same apply to the negative rail of the driver, as it too is otherwise free to float?
Thats also somewhat dependent on some conditions. The thing to think about is the capacitive coupling through the GDT, as its effectively wired to the bridge output. If left floating, there is probably a significant common mode voltage spike coupled back through the GDT. If grounded, there is likely a differential mode current flowing through the PCB ground back to the bridge rail (through the coupling cap). I dont know if either is really a problem. I guess if you ground it, there may be some benefit to connecting the ground lead closer to the gate drive section so that you dont get current spikes through the entire logic ground section, which could cause glitches as ground level of each chip is somewhat different, if only for a brief time.
I have run my control boards grounded and un-grounded, im not sure if i know which is better to be honest. With fiber optic input, there is no safety concern to leaving it un-grounded.
BTW, i recently measured one of my GDTs wound with 2 conductor shielded audio cable. It was 11 turns and had 1.17mH magnetizing L, but more importantly, it had ~100pF of capacitance between the secondary and the primary (the shield). So if I = C*dv/dt, then a CM300 switching 600V in 250nS gives current spikes of 240A. Can someone give me a reality check here? Why arent the results of this nearly as devastating as they sound like they would be? Im looking at you Richie .
Registered Member #30
Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706
I imagine it's because the current is mostly common-mode, and hence limited by the loop inductance of the gate driver board's grounding scheme, and CM300s don't switch in 250ns unless they're some special ones you stole from Area 51.
My solution was to connect the Faraday shield of the GDT to the metal case that the driver board was mounted in, the idea being to divert the current away from the driver board altogether.
Registered Member #1232
Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881
As Dr Conner said, the actual peak-current driven through the GDT's inter-winding capacitance will be smaller, and is limited by the common-mode impedance of the return path back to the power electroncs.
The path can typically be something like this... 1. HF edges (high dv/dt) are generated at mid-point of each bridge leg. 2. HF capacitively couples back through GDT to appear at output of drive IC's. 3. HF goes through drive IC's common-mode output impedance to get back to the supply rails. (Both supply rails are considered as one since they are usually well bypassed by capacitors.) 4. HF gets from supply rails onto mains wiring through inter-winding capacitance of the mains transformer powering the control/driver electronics. 5. HF gets from mains wiring back to H-bridge through bridge rectifier.
Bingo! back to the start and that is the common-mode noise circuit! This loop is familar to anyone who's tried to get a SMPSU through EMC approvals (EMI requirements for sale in the European market.) Flyback converters in particular are notorious for capacitively coupling part of the fast "flyback" edge that can be some 800V or more onto the LV output winding. Once it is there it will flow through whatever it needs to in order to get back to the high-voltage side. By connecting little ceramic Y-rated capacitors between the power side and the LV side, the designer can stop these high frequencies from leaving the confines of the power supply.
In a similar way common-mode chokes can increase the impedance presented to the unwanted common-mode noise that gets through GDT's and prevent the excessive currents that you showed could potentially flow.
As for DRSSTCs... You're probably not concerned with conducted or radiated EMI unless you are selling them commercially. So the only problem would be self-induced EMI problems within the system itself.... Firstly minimise inter-winding capacitance in the GDT. Yeah, yeah I know, it is a tradeoff between inter-winding capacitance and tight coupling - Well you can't have everything! Secondly provide a common-mode return path from the LV controller/drive side to the power side, in the form of ceramic Y-caps from the LV-side 0v rail to the DC bus negative rail on the power side. MAKE SURE YOU USE SAFETY RATED CAPACITORS And finally if common-mode noise through the GDT is still a problem you could pop a common-mode choke over the primary wires of the transformer between the drive and where they disappear into the GDT. A common-mode choke on the primary side of the GDT will kill this HF noise dead, but it will also add a little differential mode inductance that will degrade switching performance.
As i've said before, fastest switching is not always best. It can potentially cause some real EMI nightmares. So best to only switch as fast as you need to.
Registered Member #146
Joined: Sun Feb 12 2006, 04:21AM
Location: Austin Tx
Posts: 1055
MAKE SURE YOU USE SAFETY RATED CAPACITORS And finally if common-mode noise through the GDT is still a problem you could pop a common-mode choke over the primary wires of the transformer between the drive and where they disappear into the GDT. A common-mode choke on the primary side of the GDT will kill this HF noise dead, but it will also add a little differential mode inductance that will degrade switching performance.
As i've said before, fastest switching is not always best. It can potentially cause some real EMI nightmares. So best to only switch as fast as you need to.
Hi Richie,
Thanks for laying down the law again for me . But i still have a few questions to pose. Currently now ive decided that the grounding scheme that works for me is to put a 100nF film cap between the -Vbus and heatsink. Then, connect the metal box of my LV electronics (which is also PCB ground) to the heatsink, and then tie that point off to mains ground. With the use of the common mode choke on the primary of the GDT (one of those long ferrite beads was all i needed really, and doesnt seem to have really impacted the gate waveforms noticably) i have avoided creating really nasty noise on the LV PCB.
Now this differs from what you suggest slightly, and more particularly it doesnt use a Y-rated cap. Can you share maybe some experience on this? I'd have thought a nice PP film cap would be less prone to failure than a dinky ceramic, even if its rated to not catch on fire or whatever?
BTW, i tested this setup last night on my latest coil by allowing discharges (low power, just a few inches long) directly to the primary circuit! This only works because ive got the secondary ground tied with all the other grounds, as its a smaller coil. But the key, i believe, was getting the electronics to the point where its all grounded together, and the bridge is coupled to earth through the 100nF cap. Today i'll crank up the power until something goes wrong .
Registered Member #1232
Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881
> Thanks for laying down the law again for me.
There is no law, if something works then run with it. EMC is something of a black art and only complies with common reasoning up to a point
> Currently now ive decided that the grounding scheme that works for me is to put a 100nF film cap between the -Vbus and heatsink.
What you have done there is very common in SMPSUs, although 100nF is quite large. It reduces common-mode noise conducted out on the power lines, and seen on the control side for the following reason... High dv/dt at the bridge outputs get's capacitively coupled onto the heatsink through the devices thermal interface. (Multiple devices on a heat spreader over Warth sheet add up to make several 10's of pF of capacitance to the heatsink!) Once on the heatsink this common-mode current wants to get back to the DC bus -ve rail. Without any capacitor, this current usually flows to the chassis, out the earth conductor to the mains wiring. There it couples to the Live and Neutral conductors and flows back into the appliance on the L & N to the mains rectifier. Whereby it eventually ends up back at the DC bus -ve. The bigger this loop area the more noise and interference. As many Power Electronics engineers before you have discovered a decent HF cap connected right between the DC bus -ve and the heatsink keeps the return path short and low impedance, and makes for much less conducted and radiated noise in bridge circuits.
> one of those long ferrite beads was all i needed really, and doesnt seem to have really impacted the gate waveforms noticably
A single pass of wire through a ferrite sleve common-mode choke won't significantly degrade performance if the wires are passed through as a tightly twisted pair. (If you use multiple passes to increase common-mode inductance, then space them evenly around the core to reduce end-to-end capacitance. Otherwise the HF noise will jump across the common-mode choke defeating its purpose.)
> and more particularly it doesnt use a Y-rated cap. Can you share maybe some experience on this?
Y-rated capacitors are rated for direct mains connection between live and earth. They are guaranteed not to fail s/c for electrical safety reasons. Most Class 2 appliances like mobile-phone chargers, shaver PSUs etc have a Y-cap between the DC-bus -ve rail and the 0v output rail to return common-mode noise on the output side back to the power side of the transformer. In this case where the output has no safety earth, you really don't want that cap failing s/c and putting hundreds of volts of DC on the LV output side. (Incidentally the displacement current through these caps is responsible for the spark and slight shock you receive if you try to connect Class 2 appliances to earthed appliances!)
*** EDIT: *** This page has a reasonable introduction about safety rated capacitors for interference supression:
> Today i'll crank up the power until something goes wrong.
I'm sure you'll manage to break it if you try! Good shielding, good layout and attention to common-mode current paths are important in traditional Power Electronics, so I don't see why DRSSTCs should be any different.
This site is powered by e107, which is released under the GNU GPL License. All work on this site, except where otherwise noted, is licensed under a Creative Commons Attribution-ShareAlike 2.5 License. By submitting any information to this site, you agree that anything submitted will be so licensed. Please read our Disclaimer and Policies page for information on your rights and responsibilities regarding this site.
DRSSTC Grounding (Earthing)
.
. But i still have a few questions to pose. Currently now ive decided that the grounding scheme that works for me is to put a 100nF film cap between the -Vbus and heatsink. Then, connect the metal box of my LV electronics (which is also PCB ground) to the heatsink, and then tie that point off to mains ground. With the use of the common mode choke on the primary of the GDT (one of those long ferrite beads was all i needed really, and doesnt seem to have really impacted the gate waveforms noticably) i have avoided creating really nasty noise on the LV PCB.