High-value resistors and leakage

Look at the voltage specs on the components. They're often overly optimistic.

Is this for a product? If so, look up "creepage" and "air clearance".

PTH resistors?

It is, which is why I want to collect all the info I can not just for effects in a one-time circuit, but what might cause problems down the road.

--
www.wescottdesign.com

"Tim Wescott is an eWanker"

** ROTFL !!! 10Gohm is a high resistance, 10 Mohm is totally hum drum.

** Do tell...........

** Get real, for a start. ** Assuming you have about 500 volts DC, best use a leaded component designed for high voltage.

A Philips VR25 or VR37 is one option.

If you MUST use SMD, then go for a series string of at least 5 resistors (

1206 size) that total 10M.

The voltage ratings given to ordinary film resistors ( 250, 350 volts etc) are largely fictional and have nothing to do with long term stability and reliability at such voltages.

The rule here is: " Divide and Conquer ".

.... Phil

For a couple of hundred volts, I'd say don't worry too much. I presume you have a DMV that can handle a couple of hundred volts? Just hack one of them and see how they handle it. Why reinvent the same old wheel? Especially when you only have to hold "a percentage point or two?" When I was in the military, we routinely repaired electronics that handled thousands of volts, and there was nothing exotic about them at all, other than the "DANGER HIGH VOLTAGE" signs all over the shop. :-)

Have Fun! Rich

AC or DC? Is the voltage present 24/7 or just occasionally? Any transients to allow for?

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Multiple series resistors. I've seen multiple resistors on font ends of opamps measuring voltage in the range of 700vdc. A big med manufacturer, no failures. It also minimizes the Voltage coefficient on the resistors. Or you can get a nice Cadock resistor and plug it in there.

The thought of a Mill slot under the resistor comes to mind, or go thru hole for the upper arm of the divider.

Also keep the Upper arm and the lower sense resistor in the same thermal location.

Cheers

Use a flatso and put conformal coating on the loop components after assembly.

Even 'coffee breath' can cause a value change or failure mode at high values.

I would use an axial leaded resistor for the large valued "upper" resistor in the divider. Almost a must for handling the voltage standoff requirements reliably.

--

Michael Karas
Carousel Design Solutions
http://www.carousel-design.com

"Naomi Price"

** Flatso = one of these ???

** 10Mohms is not that high.

..... Phil

DC, low likelihood of transients. I'm mostly concerned with knowing how much I need to worry about fingerprints, "no clean" flux turning conductive, whether I need to conformal coat or not, etc.

It's looking like a through-hole part on the high side is a popular suggestion, I'll see how that flies with the production folks.

--

www.wescottdesign.com

10M is no problem for, say, an 0805 at hundreds of volts. If you want accuracy, use thinfilms, which will require several in series, since they are hard to find in high values.

If a couple per cent accuracy is good enough, a couple of regular thickfilm 1% resistors will do. Check their data sheets and make sure the voltage coefficient is OK.

John

Right, so look up the two terms and related specs. There is a lot of information on the net (unfortunately, not the specs themselves) that appears to be pretty good. It would be a good discussion but I don't see any takers.

It's not obvious stuff, and I'm certainly no expert, just having dipped my toe into these waters, but it certainly looks to me like component manufacturers stretch the truth, a lot. ;-)

I wouldn't worry too much about fingerprints etc., but I have seen 1% resistors drift 30+% with HV dc applied. Vishay/BC (formerly Philips) VR25/HVR25 resistors would have been my choice years ago, but these guys might be okay for a modern design:

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

Flatso means 'flat Silicon Oxide' IIRC. They are metal film on SiO substrate.

You can get them coated or uncoated, but without encapsulation, coated is the only way to keep the value.

Yes, 10M Ohms is not that high, and a radial leaded HV (VR25) package would probably be just fine.

The IEC spec forces low-to-the-board placement, so you may have to explain to them that there are cases where the EIC spec can be foregone, and mount it slightly elevated.

I once measured 30kV using a motor-driven rotary capacitor.

After all the real equation is....

I = d/dt(C*V).

If V is DC, this reduces to I = V*(dC/dt) ;-) ...Jim Thompson

[On the Road, in New York]

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.

--
http://www.caddock.com/Online_catalog/Mrktg_Lit/TypeUSVD_HVD.pdf

I usually use several 1206 resistors in series. The number of resistors depend on the rated voltage of the resistors.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

We use 1206 melf resistors MMB 0207 professional range. These have a pulse rating of 1kV and 35W. We use 6 in seires on 600V and have never had a failed resistor.