FR4 base material resistance

[Robert Latest]

Hi all,

I'm designing a PCB for a lowish-current application and I'm wondering
how much I can rely on the FR4's isolation properties. This is a
transimpedance amplifier with the summing junction at virtual ground,
and I'm going to use guard traces. The input amp is an LT1793, DIP-8
package. I can tolerate a 10pA leakage current, which with a worst-case
input offset voltage of 2mV translates to a maximum tolerable leakage
resistance of 200 Megohms.

Will I have to use teflon standoffs, or will standard PCB do? Will a
solder stopmask make a difference? The summing junction copper area will
be small and will only connect five pads.

Thanks,
robert

 

[Rene Tschaggelar]

Robert Latest wrote:

Hi all,

I'm designing a PCB for a lowish-current application and I'm wondering
how much I can rely on the FR4's isolation properties. This is a
transimpedance amplifier with the summing junction at virtual ground,
and I'm going to use guard traces. The input amp is an LT1793, DIP-8
package. I can tolerate a 10pA leakage current, which with a worst-case
input offset voltage of 2mV translates to a maximum tolerable leakage
resistance of 200 Megohms.

Will I have to use teflon standoffs, or will standard PCB do? Will a
solder stopmask make a difference? The summing junction copper area will
be small and will only connect five pads.

The teflon standoff is only for vibrationally sensitive
parts. The isolation from not touching anywhere at all
is equally good.

Rene
--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net

 

[Pooh Bear]

Robert Latest wrote:

Hi all,

I'm designing a PCB for a lowish-current application and I'm wondering
how much I can rely on the FR4's isolation properties. This is a
transimpedance amplifier with the summing junction at virtual ground,
and I'm going to use guard traces. The input amp is an LT1793, DIP-8
package. I can tolerate a 10pA leakage current, which with a worst-case
input offset voltage of 2mV translates to a maximum tolerable leakage
resistance of 200 Megohms.

Will I have to use teflon standoffs, or will standard PCB do? Will a
solder stopmask make a difference? The summing junction copper area will
be small and will only connect five pads.

FR4's pretty good. You'll find that performance related leakage currents
will be determined by soldering residues if you don't clear these up.

You already know about guard traces - which is good.

I'm sure it'll be humidity sensitive at *very* low currents though. If you
really need to get down that low you should have the pcb cleaned after
assembly thoroughly ending up with a de-ionised water rinse and then apply a
conformal coating.

Graham

 

[Robert Latest]

On 2005-06-28, John Larkin <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote:

If it's guarded, 10 pA is easy on FR4.

The ultimate insulator is air, with the critical node being a mid-air
solder splice. You shouldn't have to resort to this... it's fA
territory.

Yeah, that's how I usually do it without thinking much about it.
But it's a pain production-wise, so I wondered if I can get away
with normal PCB insulation. Looks like I can. I have already
tried it out and it worked fine, but I didn't know if I was just
lucky that one time.

Thanks,
robert

 

[Joerg]

Hello Graham,

I'm sure it'll be humidity sensitive at *very* low currents though. If you
really need to get down that low you should have the pcb cleaned after
assembly thoroughly ending up with a de-ionised water rinse and then apply a
conformal coating.

That's exactly the point. After all, you don't want this circuit only to
function properly in the Sahara desert at high noon.

Potting the whole thing up is another option.

Regards, Joerg

http://www.analogconsultants.com

 

[Pooh Bear]

Joerg wrote:

Hello Graham,

I'm sure it'll be humidity sensitive at *very* low currents though. If you
really need to get down that low you should have the pcb cleaned after
assembly thoroughly ending up with a de-ionised water rinse and then apply a
conformal coating.

That's exactly the point. After all, you don't want this circuit only to
function properly in the Sahara desert at high noon.

Lol ! I missed the part about baking the pcb after the de-ionised water rinse for
brevity ! I expect the Sahara would do the equivalent just fine ! ;-)

Potting the whole thing up is another option.

If you like using chisels to fix it ! ;-)

Graham

 

[martin griffith]

On 28 Jun 2005 10:16:16 GMT, in sci.electronics.design Robert Latest
<boblatest@yahoo.com> wrote:

Hi all,

I'm designing a PCB for a lowish-current application and I'm wondering
how much I can rely on the FR4's isolation properties. This is a
transimpedance amplifier with the summing junction at virtual ground,
and I'm going to use guard traces. The input amp is an LT1793, DIP-8
package. I can tolerate a 10pA leakage current, which with a worst-case
input offset voltage of 2mV translates to a maximum tolerable leakage
resistance of 200 Megohms.

Will I have to use teflon standoffs, or will standard PCB do? Will a
solder stopmask make a difference? The summing junction copper area will
be small and will only connect five pads.

Thanks,
robert
just found this bit of peaseology

http://www.elecdesign.com/Articles/ArticleID/6136/6136.html
and
http://www.elecdesign.com/Files/29/6136/Figure_03.gif


martin