Miniature solenoid

Where do you get the calibrated femtoamp from?

How long do you need the calibration current to be on? I'm thinking a linear voltage ramp into a very small capacitor might work. It doesn't need an isolation switch.

John

sm to

cant you replace the existing reed? Or rewind the coil? Rapidonline do/ did a small bistable relay, and are in the UK.

NT

In the 80's I was using the Aromat S2E-L2 relays in my designs to measure fA currents. This is not a reed relay and it was a low thermal EMF type designed for uV applications but it worked a charm measuring opamp Ib too. Careful layout is critical, as is ES shielding of the fA circuitry on the PCB.

If your need is one-off for lab use then you may want to look at the S2E-L2 as your solenoid. Stripping the case and attaching directly to the armature may get you what you need. Good luck. Art

ism to

In recent times the National/Calibration labs have realized the need for calibration at very low currents e.g. Willenberg G-D, Tauscher H N and Warnecke P 2003 A Traceable Precision Current Source for Currents Between 100aA and 10pA IEEE Trans. Instrum. Meas. 52,

though the technique goes back at least as far as:

Praglin J, Nichols W A 1960 High-speed electrometers for rocket and satellite experiments Proc. I.R.E. 48, 771-9

and I have used a crude version of this so far, but knowing accurately the value of the capacitor presents considerable difficulty. For the present application there are five or more amplifiers ( they are operated in an evacuated and temperature controlled housing measuring ion currents in a mass spectrometer) and the need is to intercalibrate them so having a separate input capacitor on each is not feasible. Any connection present during normal amplifier operation can introduce noise and extra capacity affects the stability and frequency response. Retrofitting to existing spectrometers very much limits the space available. The operation also needs to be compatible with the existing control system.

Scott.

I would use a to-99 latching relay, good ones are clean past sub-picoampere levels. Not sure where to get them these days, but they were plentiful 30 years ago.

?-)

OK. What do you use to generate the cal current?

A steerable electron beam would be interesting.

John

nism to

Mercury-wetted reeds have no more leakage than any other reed relay. If you apply enough volts you can get some leakage through the glass from ion migration - the electro-luminescence from this limited the voltage we could put across a photomultiplier-tube face-plate in the Everhart-Thornley detector on the Cambridge Instruments electron microscopes to less than one thousand volts.

-- Bill Sloman, Nijmegen

Aren't you also going to have some fun and games to avoid infringing the various patents on the reference current calibration systems - or have they just expired recently? Or is it a one off at a uni?

We typically used a very stable reference voltage, a 10^11 ohm resistor and a bunch of carefully selected reed relays. Timing was critical to get the optimum solution same applies for systematic zero offsets.

The kit chosen usually has to be able to survive being hot baked to get a hard vacuum too which is a minor detail you seem to have omitted to mention. I assume here you are talking about a gain and offset reference system for a multicollector Thermal Ionisation MassSpec.

Thanks for all the suggestions. The overriding problem is that everything leaks at these current levels (fA and aA). A few comments on some posts: John Larkin: A steerable electron beam is probably the ideal source from the point of view of testing the transient response but raises some difficulty in knowing the current intensity. A mass spectrometer provides a source of very small currents but this does not provide a relative calibration over a wide range of amplifier gains. The matter of calibration current is discussed in the references in my previous post. Bill Sloman: All relays leak and at these levels the glass envelope ( and any crud settling on the surface which is why the amplifiers are run in an evacuated enclosure) provide a limit at small currents. The mercury relay is not necessary since the relay resistance is of little significance since the current source has relatively very much higher resistance. Martin Brown: The amplifiers are removed from the system before baking. The system is a multi-collector noble-gas mass spectrometer rather than a TIMS. There is no problem with patents as the voltage ramp with capacitor generator was published very many years ago (see references in previous post). You have to allow for a factor of 1000 increase in gain relative to the 100G amplifiers used so far in mass specs.

Scott Hamilton.