"Jasen Betts" bravely wrote to "All" (05 Mar 04  20:41:20)
 --- on the heady topic of "{at}%^{at}#$%^ veroboard"

 JB> Hi All.

 JB> I was tring to build a prototype that involved 1M resitors
 JB> but I couldn;'t get it to behave sensibly

 JB> I measured 400K between the tracks on the stripboard.
 JB> dunno if the 30C heat and 100% humidity had anything to do with that.
 JB> so I stripped the components off the board and built the whole thing
 JB> free-standing between two wires...  seems to work now.

Might be some invisible surface contaminant. Try cleaning it with 99%
alcohol and see if it has any effect. I know that some types of flux
residue will conduct, obviously acid types are one of these.

My cousin asked me once "isn't flux supposed to be non-conductive" but
even electronics rated non-acid fluxes may result in increased
conductivity because they can absorb water out of the air. The flux
can also hide tiny solder balls or hairline bridges which may be so
small as to be nearly invisible. So if your prototype requires megohm
impedances you would do well to always clean off all flux on a pcb.

Aside from the flux there may also be triboelectric charges on a pcb
which can influence conductivity. However, these tend to be a DC
effect which becomes annoying when measuring very tiny voltages and
currents in the microvolt and nanoampere range respectively. This
comes about when rubber insulation rubs against a conductive metal and
causes charges to be displaced and accumulated. This will of course
adversely affect a high impedance DC circuit node.

There are still other effects which will upset a high impedance point
for example simple static electricity. The solution for this is to
build a conductive grounded guard ring around the high impedance node
as is often seen in opamp data handbooks.

 Mike
 ****

... I worked hard to attach the electrodes to it.
--- Blue Wave/DOS v2.30