Current source

I'm feeling the need for a bench current source. Like an old Keithley 225. 100mA to 1nA would be fine. Beside the Keithley are there any other brands (models) I should look at?

George H.

Reply to
George Herold
Loading thread data ...

I'm down on Keithley lately. Sent back four of their 2100 DVMs (cheap buggy Chinese rebrands) and their 2401 sourcemeter (terrible EMI sensitivity.)

2401 when you grab the test leads:

formatting link

Crashed 2100, and its fix:

formatting link

The old stuff was great:

formatting link

--
John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

Agilent has a lovely current/voltage source (B2900A), but it's like $6k and up, IIRC.

formatting link

John L. has reported on some issues he had with his Keithley equivalent (Model 2400?).

--sp

Reply to
Spehro Pefhany

Thanks guys, I can get a Keithley 225 for ~$250 on ebay. I really don't care if it's a bit out of calibration. Mostly I just want to push some low value current around ... I'm futzing around with Si and the hall effect at the moment. Are there current sources from the same era by HP, or fluke or someone?

George H.

Reply to
George Herold

On a sunny day (Fri, 17 Jan 2014 10:53:04 -0800) it happened John Larkin wrote in :

That character set (font used) is terrible, hardly readable. For the same or less money you can make a clear font:

formatting link

Yes, more readable, mirror scale, too.

Reply to
Jan Panteltje

On a sunny day (Fri, 17 Jan 2014 14:02:58 -0500) it happened Spehro Pefhany wrote in :

1nA is not that exotic. He should build one.
Reply to
Jan Panteltje

Well, it is a software crash report. It says all you need to know: This Keithley 2100 Doesn't Work.

The Fluke is great, and made in USA.

--
John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

If you need something exotic, or if you can use some boat anchor at $150 from eBay, it's cheaper to buy.

Not everyone can throw stuff together as fast as you. ;-) And just the Digikey order for the parts would probably be $150 if you want a box and power supply.

--sp

Reply to
Spehro Pefhany

On a sunny day (Fri, 17 Jan 2014 17:08:52 -0500) it happened Spehro Pefhany wrote in :

We, he did not specify a voltage, but the simplest current source we used (linear or just about) was some voltage with a high value resistor. So 1V and 1GOhm...

100V in 100GOhm, at 1% accuracy you can then have about a volt out.

But I think there are transistor and opamps combinations that would work here.

I have been pissed with ebay lately, the guy send the wrong LED strips, and and some LEDs stopped working after a few hours.

I bought some Microchip 24LC1025 and the one I am testing now (the first one) works OK in the first 512 pages, but above 16 bit address the chip refuses to ack the i2c... (at address 168 decimal), so it is only a 24LC512.. And there was more junk. I am still looking into it, will scan all them chips for ack addresses tomorrow.

SO YMMV with eeeeeebay.

Reply to
Jan Panteltje

I made a 10nA -10mA decade current source with a voltage reference and 0.1% resistors.. well some 1%. But compliance was only 5V, for driving diodes.

I don't need hundreds of volts of compliance, 30 V would be OK. I'll just order the 225.

George H.

Reply to
George Herold

Don't be so sure about that.

Jamie

Reply to
Maynard A. Philbrook Jr.

Mine says MADE IN USA on the back.

--
John Larkin         Highland Technology, Inc 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

Maybe this is a silly question, but is it mot the case that currents in the 1 nA range are indistinguishable from electronic noise ? That is if one drives 1 nA into a circuit, would the circuit recognize it?

Reply to
dakupoto

It is not. The *average* circuit with bias on the order of 1mA won't notice it, but that doesn't mean it's not there.

One example of a "pedestrian" circuit that uses nA currents is a function generator like this,

formatting link
(oops, the diff pair output is swapped)

The capacitor charging and discharging is achieved through current mirrors, and the voltage monitored by a JFET follower. At room temperature, with the mirrors more-or-less cut off (V(Freq.) < 0.5V), leakage through the top 2N4403, 1N914 and JFET typically amounts to ~10nA, IIRC, so that the frequency spans from single digit Hz to MHz in a single range.

The smallest currents that can be resolved with human-sized equipment are on the order of femtoamperes. Measurements are very slow at that current (indeed, thousands of electrons per second), because with capacitances on the order of 2pF for an input pin (let alone if you do anything with it), the signals are hardly milivolts. Resistors, leaky paths, etc. are not major sources of noise, because they can all be made very small indeed (today's CMOS op-amps have nearly unmeasurable input leakage). What is a major source of noise is background radiation: ionized air around the active electrodes causes large random shifts in charge.

Monolithic circuits get even better. Flash memory stores information as voltage levels on an electrode embedded in silica glass; the leakage is essentially nil. On modern processes, an electrode might hold 100 excess electrons, and maybe one or two leak out after several years at rated temperature.

Obviously, such structures are even more vulnerable to radiation, since just one stray particle can zip through and generate thousands of free charges, causing momentary conductivity in SiO2, voltage offsets in Si, and causing permanent structural damage (crystal defects) that leak charge away.

Tim

--
Seven Transistor Labs 
Electrical Engineering Consultation 
 Click to see the full signature
Reply to
Tim Williams

With a cheap LMC6041 or LMC662 op-amp you can make an amplifier with input currents of around 5x10^-14A or better, in my experience. So yes it is relatively easy to recognise 1nA since it is 20000 times more than the input current of a cheap op-amp assembled into a circuit at home.

formatting link

The LMP7721 would be really nice (low offset voltage as well as low bias current) but they don't sell it in DIP and I doubt I can reliably solder a wire to a SO-8 without getting flux on the body of the package, nor put the input pin into a single socket.

Chris

Reply to
Chris Jones

On a sunny day (Fri, 17 Jan 2014 22:17:26 GMT) it happened Jan Panteltje wrote in :

My error, should read the datasheet from page 1 first, it needs A2 connected to Vdd, even though hardware address line A2 is not used.... Had it connected to ground. Chips working now.

Build a little 24LC1025 test-rig programmer for the PC parport, parport output also powers the EEPROMS:

formatting link

parport 24LC1025 signal pin 1N4148 signal pin

10 -----|
Reply to
Jan Panteltje

My old Keithley has a full-scale range of 1e-13 amps, which resolves 1e-15 amps, about 6000 electrons per second.

formatting link

I built a fA parts tester that works pretty well.

formatting link

formatting link

formatting link

I can plug in, say, a 1T resistor into the Z1 port and the DUT on the other.

The worst problem was huge leakage in the cheap Radio Shack banana posts. I had to hack the hole and add the lexan sheet to insulate them.

An ordinary cheap DVM will have a 10M input resistance and might have 100 uV resolution. So if you put it on its lowest voltage range and use it as an ammeter, it will resolve 10 pA.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

Rosin flux seems to conduct not at all. And you can always wash it off with acetone or something.

--
John Larkin                  Highland Technology Inc 
www.highlandtechnology.com   jlarkin at highlandtechnology dot com    
 Click to see the full signature
Reply to
John Larkin

The theoretical limit is something like 6 orders of magnitude less (making some assumptions about measurement time, temperature etc.).

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward" 
speff@interlog.com             Info for manufacturers: http://www.trexon.com 
 Click to see the full signature
Reply to
Spehro Pefhany

If you are only worried about picoamps then sure that is pretty easy to clean things, but for femtoamps I am not so enthusiastic.

In my experience I cannot get chips quite as clean as when they are new. I have tried. I think there might have been impurities in my solvents though I did finish with DI water from a wafer fab.

Also I couldn't use acetone as it eats the low-leakage polystyrene capacitors that I was using.

Even if there is a successful and complicated procedure that allows cleaning parts to achieve the same leakage current as brand-new parts, it would be nicer not to have to do it in the first place, which is quite feasible if you can use a DIP package and just bend up the input pin so it doesn't go into the socket, and connect the input pin up in the air, using a single turned-pin receptacle removed from a socket.

Chris

Reply to
Chris Jones

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.