"<2fA resolution (200pA range)"
That is a bold claim especially on normal BNC cables.
Someone probably needs to buy one and perform a tear down.
Fairly 'standard' transimpedance amplifier design with careful layout that's gone through a couple of iterations, lots of slots in PCB's etc..
The front-end and the bias generator are isolated with a transformer and opto-couplers.
We are updating our website at the moment (as well as manuals/datasheets/etc...) , so the picture is a little out-of-date, most of our products now have a label on-top and all have stainless steel front & rear panels. Here's a picture courtesy of our US distributor, Saelig.
Fairly 'standard' transimpedance amplifier design with careful layout that's gone through a couple of iterations, lots of slots in PCB's etc..
Perhaps you can answer some questions?
1. BNC instead of triax? Why did you decide not to use them?
2. What was the noise level on the lower range?
3. Any pictures inside?
Fairly 'standard' transimpedance amplifier design with careful layout that's gone through a couple of iterations, lots of slots in PCB's etc..
Perhaps you can answer some questions?
1. BNC instead of triax? Why did you decide not to use them?
2. What was the noise level on the lower range?
3. Any pictures inside?
BNC's tend to be used on picoammeters (i.e. Keithley 6485), generally there is no requirement for a guard system - the inner and outer parts of the BNC are typically at the same potential; it can be thought of as essentially zero ohms.
Also cost is a major consideration - BNC's are low-cost and fairly readily available - triax connectors are very expensive (40-100x more), more difficult to obtain, etc.. Two triax connectors would equal the retail price of the SPA100. I bought some recently for my Keithley 237 SMU, eye-wateringly expensive and that was without the cable.
Noise. On the lower range will depend on sample rate (2/10Hz) a little and sample (rolling) averaging (1x to 64x) - however it tends to be dwarfed by aliased mains hum pick-up. We have another SPA100 batch finishing production this week, so will post a few screen captures of the o/c noise if you'd like.
Sorry no pix inside.
Fairly 'standard' transimpedance amplifier design with careful layout that's gone through a couple of iterations, lots of slots in PCB's etc..
Perhaps you can answer some questions?
1. BNC instead of triax? Why did you decide not to use them?
There is no need to use a triax connector for a transimpedance amplifier. The reason is simple, the potential difference between the ground and the input terminals is practically zero (to be precise equal to the amplifier offset). And this offset doesn't generate tangible leakage current.
But warning in noise that can be generated by not a very high quality insulator.
Fairly 'standard' transimpedance amplifier design with careful layout that's gone through a couple of iterations, lots of slots in PCB's etc..
Perhaps you can answer some questions?
1. BNC instead of triax? Why did you decide not to use them?
There is no need to use a triax connector for a transimpedance amplifier. The reason is simple, the potential difference between the ground and the input terminals is practically zero (to be precise equal to the amplifier offset). And this offset doesn't generate tangible leakage current.
But warning in noise that can be generated by not a very high quality insulator.
We haven't noticed any noise we can trace back to insulator materials. The insulated 50R BNC's are made from white polypropylene (so without black/carbon fillers).
On the lowest current range you can create a small transient shift in the reading by changing the bias generator output from say -40V to +40V, but then waving your hand 3 feet away does the same when the nothing is plugged in.
but then waving your hand 3 feet away does the same when the nothing is plugged in.
This is the simplest way to test whether the electrometer input is still alive.
The dielectric noise will be lower than the noise of your amplifier. Therefore, it is not noticeable.
P.S.
Unfortunately, I am not able to test this amplifier.
Its parameters and design look very balanced.
But I don't believe in 1E-15A noise free, only if with a very "evil" filter, with a long settling time.
I think you used a 10GΩ feedback resistor for the low range?
Thanks.
My interest in the resistance of the Rf is only in the context of a rough estimate of the possible noise floor of the amplifier.
We haven't noticed any noise we can trace back to insulator materials. The insulated 50R BNC's are made from white polypropylene (so without black/carbon fillers).
Measuring current generation after physical impact (connection) and noise. Pay attention to how long it takes to settle (More that 2 hours). Also You can view noise of this triax connector.
Case and inner ring(guard ring) are connected to ground, central pin connected to electrometer input. Direct connection without using a cable. The input connector of the electrometer served as a measuring chamber.
In addition to the graph, I also attach the raw data in the archive.
Interesting. What picoammeter/electrometer are you using, and what did the test setup look like?
Your data is approx -0.28fA to -0.04fA (40aA!) - what makes you think this is purely noise from the triax insulator and not from elsewhere?
Interesting. What picoammeter/electrometer are you using, and what did the test setup look like?
For this measurement I'm use my V7-45 electrometer. Self noise of this electrometer in attach.
Also measurement setup looks like at the photo in the attachment.
In these photos, I not connect the ground. As I quickly reassemble it to show how it looked.
Your data is approx -0.28fA to -0.04fA (40aA!) - what makes you think this is purely noise from the triax insulator and not from elsewhere?
I think the above description and the attached graphs and photographs describe why I am confident that I have measured the noise and parasitic current of a standard triaxial connector with PTFE insulator.
Great. Looks like a beast of instrument. I'm assuming these are only available used in FSR countries?
I'm assuming the V7-45 is a two box unit. Any pix of the insides??
bvi = does this connect to the head?
kop = control/computer?
vb1chod analog vb1y = analog output?
bvi = does this connect to the head?
this connector for connect external measurement block (yes, head)
kop = control/computer?
- this is GPIB IEEE488.1
vb1chod analog vb1y = analog output?
Yes.
Love those vintage eastern block instruments. Performance looks amazing - did they build an updated unit or did it fissile (pun intended) with the end of the cold war?
The updated device was called B7-57 / 2, produced after the collapse of the USSR. Its production was piece and the device was unobtanium.
I'm restore one. Found it in a very poor condition and unfortunately not complete. Some photos "as received" in attach.
So do you have different heads for current and voltage
No, one head can measure currents, voltages and charges.
The 'dust' at the corners of the PCB is from opening the enclosure and is a mixture of the blacking of the taptite screws and the aluminium extrusion - it isn't there when assembled.
Worth tapping the empty-box on a desk before reassembling, and brushing the dust off the PCB.