Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

simple current source

Status
Not open for further replies.

ubergrad

New Member
I am trying to build a variable AC current source using an AC voltage supply and large valued sourcing resistors. (I know a more 'active' source is a better design, which I will also be building, in the meantime this should be an effective way to achieve what I want)

HOWEVER there seems to be some frequency dependance in my output and I can't figure out why this would be. The design employs a simple current divider with the resisitor through the top branch of the divider variable between 5 values (ie. a 5 pos. switch through 5 different resistors) of, 1MOhm, 10MOhm, 100MOhm, 1GOhm, and 10GOhm, producing an output current of 1microAmp, 100nA, 10nA, 1nA, and 0.1nA respecitively. The first three resistor values seem to work fine, however the setup breaks down on the 1Gig, and 10Gig resistors. Here, I've noticed a huge frequency dependance in my output, with for example, an input frequency of ~10Hz giving me the expected output current, however as I increase the frequency the output current increases accordingly, with, for example, 15times the expected output observed at only a few hundred Hz...

I have no idea why this would behave in this way...is there some obvious underlying circuit theory I'm totally unaware of?!?

I've racked my brains and do not understand this at all...Please let me kwow if you have any suggestions, including any references that might be helpful.

Many thanks.
 
Last edited:
I think that your problem is due to stray capacitance.
At 300hz, 1pF has a rectance of 530M ohm.
530Meg in parallel with 10Gig is 530Meg (near enough).
The stray capacitance is dominating the circuit as the frequency increases.

JimB
 
Many thanks for the quick response JimB....indeed I think you are exactly correct. I roughly ploted the resutling frequency dependancy, and the trend seems to match my device output well.

(Any suggestions on how to *reduce* stray capacitance?!?)

Again, many thanks for your help!
 
Without knowing what your "device under test" is, and what your test set-up looks like it is a bit difficult.

As a general suggestion, keep the wires short and use screening where possible.
But screening adds capacitance, maybe where you dont want it.

JimB
 
Capacitance is also a function of distance. Two conductors with a dielectric in between (even just air) forms a capacitor.

Keep your sensitive nodes (like the 1GOHM & higher) far away from other conductors. Putting them in the air with lots of gap is a common way. If it is not practical you may get some improvement with things like using teflon coated wire. Careful physical construction is paramount.

To further test if you are fallig victim to parasitic capacitance is to move your finger near the sensitive node, if your frequency characteristics change significantly, you have found your problem.
 
In high ratio divider circuits, often using resistors alone of correct ratio do not work correctly at high frequencies due to stray capacitance.

Therefore on instrument like high voltage probes a suitable capacitor is placed at the output end to "compensate" for the effect of the stray capacitance. The theory is that the capacitor together with the stray capacitance together also forms the correct divider ratio similar to that of the resistor divider.

You can experiment with capacitors of 10pF to 200pF across the output to see if it helps with your situation. The higher the ratio, the higher the value of the capacitor.
 
Are you using the same frequency all the time?

How about a capacitive divider instead of a resistive divider?
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top