Bleeder resistor to dissipate static
in [Design]
|
From: KJ on 7 Feb 2010 19:36 On Feb 5, 11:56 pm, "Ban" <bans...(a)web.de> wrote: > > Better to use a separate ground connection with an extra wire to the > grounded frame or another low impedance point The plate receiving the discharge is the frame. > and isolate the bolt from your > circuit. The bolt provides the electrical connection to provide the 'earth' connection to the board. Isolating it would float the board relative to an earth ground. KJ
From: KJ on 7 Feb 2010 19:40 > Since you are talking about electrostatic voltage / current, the > current is low and a resistor would bleed off the charge meaning little > or low net voltage. Nominally low...except when the discharge comes along and now you have a high transient voltage for a couple hundred ns. > I would use something between 1K and 10Meg depending on the severity > of generation. I would too...but was wondering if anyone knew of any pitfalls associated with. I wouldn't want to create an arc on the PCB as an example. > I also would add a metallic or other conductive shield between the > generator and sensitive PC boards and directly connect it to the PCB > ground system, NO other connection (it can have insulation around it if > needed). That would probably cost more than the ferrite though KJ
From: KJ on 7 Feb 2010 19:45 > A bleeder resistor won't have any place to connect, will it? Right now the large plate connects to the PCB with a post that touches the PCB to provide the electrical connection to an 'earth' ground on the board. What I was thinking was to get rid of the the PCB connection to earth and insert a resistor between where the post comes down and the earth ground net on the board. > What you > really want, is a Faraday cage around the sensitive LCD (but it may be > awkward to put metal window-screen material over the face). > > A frequently seen solution is a set of corona points (basically, tiny > little > lightning rods) next to the paper which carries the charge. Another > might > be some UV LEDs, which may create enough ionization to discharge the > paper. Radioactive strips in the old record-cleaner brushes performed > the same function. The cage and LEDs would likely be more expensive than the ferrite. The corona points would be a rework problem but might be useful for new builds. KJ
From: KJ on 7 Feb 2010 19:56 > > The reset is likely coincident with the discharge from the paper, It is. > which you can't attach a resistor to - all you can do is provide a > more controlled discharge path that does not involve the victim > circuit, That's what I'm looking for, a way to control how the static hit to the plate gets out of the system without disrupting the ground on the board. > or prevent the charge from accumulating in the first place. > There are some measures in the system to bleed off charge to the plate before it gets too large...but apparently not quite enough. > Otherwise, resistive bleeders across isolation barriers are safety > listed commodity components with values between 1/2 and 12 Meg. Pilkor > PPSR37 , PWSR37 and Phoenix HVR types come to mind. If this isn't a > safety barrier being crossed, then simply using parts of similar > construction would probably serve the purpose. > It is a safety connection. I'll keep looking for any other app notes/concerns that might be associated with separating 'frame' from 'earth' without creating other issues. > Frankly, US$0.27 does not strike me as a barrier in the cost of the > 'system'. > The 27 cents was just the upper cost limit, since that's what the ferrite that forms one possible solution costs. Actually slightly more, since there would be some minor labor as well. KJ
From: Jon Slaughter on 7 Feb 2010 23:13
KJ wrote: > How viable is it to use a bleeder resistor rather than a direct > galvanic connection to try to cut down the transient induced by > static? > > In our system we have moving paper that will generate occasional > discharges to a large metal plate which supports the mechanics. > Underneath that plate (~1 inch) there is a circuit board that has a > single connection to a post coming down from that plate. A flex > circuit is plugged into a connector which runs back up through that > plate up a tower to an LCD. Occasionally the discharge can disrupt > the LCD temporarily, basically resetting it. One thought was to try > to control this discharge by inserting some series resistance between > the post and where it would otherwise connect to the circuit board to > provide the 'earth' connection to the circuit board. Obviously this > would need to be a resistor capable of withstanding high volts and > impulse type events as well as PCB layout rules to support this. Does > anyone have any references to such a technique? > > The alternative solution is a 27 cent ferrite around the flex circuit, > and cost is a concern with the product so exotic/expensive approaches > are out. Also investigating alternatives up near the LCD as well, but > just wondering if there is a good way to dissipate the static shock as > a bit of heat rather than disrupting the ground and everything at the > LCD that is sitting up on a perch..and do it for ~$0.25 USD or less. Your still going to have issues with static buildup if you do not find a way to appropriately keep the charge from building up. This may or may not be an issue depending on the saftey requirements. It seems you simply need to properly ground the large metal plate and isolate the flex line and/or the circuit board. This can be accomplished many ways. By providing a much lower impedance path for the discharge through the plate to ground then you reduce the voltage on the flex cable. By increasing the electrical isolation of the flex cable you decrease it's susceptibility to the discharge. If you fix the source of the problem then you won't have issues with the other components. I assume the metal plate is not properly grounded? |