Author Topic: Active probe?  (Read 1344 times)

0 Members and 1 Guest are viewing this topic.

Online awallin

  • Frequent Contributor
  • **
  • Posts: 657
Active probe?
« on: January 02, 2020, 07:28:43 pm »
Are there any good ~1 GHz active probe open designs out there?

would simply a high-speed FET opamp work? comments, ideas?
The +/-2.5V supplies would be nice to generate from 5V USB - any ideas for a small low-noise dc-2-dc converter for that? (followed by low-noise LDOs)

 
The following users thanked this post: vk6zgo

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 10963
  • Country: us
  • DavidH
Re: Active probe?
« Reply #1 on: January 02, 2020, 10:50:46 pm »
An FET operational amplifier will work however transient response will be compromised by feedback to the inputs even if configured as a voltage follower.  I think you would actually get better performance using a medium input resistance and current feedback amplifier or even better, an operational transconductance amplifier like the OPA860 however package and circuit layout limitations will limit ultimate performance.

I think a better way is the classic JFET source follower input stage preceded by a x5 compensated divider and then bipolar emitter follower to drive a double terminated 50 ohm input yielding a total x10 attenuation.  Without any gain stage, this makes best use of the parasitic elements in printed circuit board construction.

There are some more specialized amplifier ICs which someone else may suggest but I am not familiar with them.
 

Offline RCinFLA

  • Contributor
  • Posts: 7
  • Country: us
Re: Active probe?
« Reply #2 on: February 15, 2020, 09:21:04 pm »
The main objective of an active scope probe is to reduce the 11 pF to 15 pF shunt capacitance of a normal divide by 10 passive scope probe.  Secondary objective would be to avoid  the 10x signal loss.  In order to get the capacitance down you have to have the amplifier at the probe tip end of the scope cable.

Expensive active probes have only couple tenths of pF shunt capacitance.  This is less than most common chip resistor parasitic shunt capacitance.

There are some low cost FET followers on ebay but they are capacitively coupled. 

For response to D.C. you pretty much have to use op amp.  For an op amp solution look for the highest slew rate op amps.  Keep the feedback resistors values low to prevent effects of op amp neg input node stray capacitance and layout needs to avoid nearby ground stray capacitance at the op amp - input node.  Too high feedback resistor with too much neg input node capacitance can make the op amp unstable.

The op amp output drive current must support the low value feedback resistor in parallel with 50 ohm load impedance output to scope.  Scope input needs 50 ohm termination so cable is terminated properly.  If scope does not have 50 ohm input option a BNC-T with 50 ohm terminator load will do.
 
The following users thanked this post: thm_w

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 19427
  • Country: nl
    • NCT Developments
Re: Active probe?
« Reply #3 on: February 16, 2020, 03:54:23 pm »
The parasitics will likely kill any hope of getting some kind of bandwidth. The schematic in the first post isn't going to work at all. It starts with an 100k / 15pf RC filter.
There is a long thread about making your own (differential) 1GHz probes: https://www.eevblog.com/forum/testgear/gt-1-ghz-diy-differential-probes/ which ended up with a neat product  8)
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 10963
  • Country: us
  • DavidH
Re: Active probe?
« Reply #4 on: February 16, 2020, 04:53:46 pm »
For response to D.C. you pretty much have to use op amp.

JFETs work just fine down to DC and a JFET source follower has many advantages over a series feedback amplifier.  Up to 500 MHz, the common active probe design in the past used two JFETs and two RF bipolar transistors.
 

Offline Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9641
  • Country: my
  • reassessing directives...
Re: Active probe?
« Reply #5 on: February 16, 2020, 06:28:37 pm »
The schematic in the first post isn't going to work at all. It starts with an 100k / 15pf RC filter.
where do you get the 15pF figure? BAV99 is 2pF (maybe 4pF for both) OPA858 is negligible. but yeah, talking about copycatting from bad schematics (pin 2 BAV99 should go to +ve power rail)
It's extremely difficult to start life.. one features of nature.. physical laws are mathematical theory of great beauty... You may wonder Why? our knowledge shows that nature is so constructed. We simply have to accept it. One could describe the situation by saying that... (Paul Dirac)
 

Offline MasterTech

  • Frequent Contributor
  • **
  • Posts: 924
  • Country: 00
Re: Active probe?
« Reply #6 on: February 16, 2020, 07:14:27 pm »
These old JFET probes (from Tektronix, HP, Philips...) not only have a low bias input but also a very low input C, which is what makes them really useful in HF measurements at mostly any point in the circuit.
However I'm amazed at the high input C of the hyperexpensive tektronix IsoVu (starting at 41dB$ according to the X-chapters), 20pf at 1MOhm with no input divider and 0.5V range, there goes all the bandwidth



« Last Edit: February 16, 2020, 07:26:19 pm by MasterTech »
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 19427
  • Country: nl
    • NCT Developments
Re: Active probe?
« Reply #7 on: February 16, 2020, 07:45:59 pm »
The schematic in the first post isn't going to work at all. It starts with an 100k / 15pf RC filter.
where do you get the 15pF figure? BAV99 is 2pF (maybe 4pF for both) OPA858 is negligible. but yeah, talking about copycatting from bad schematics (pin 2 BAV99 should go to +ve power rail)
The pads of the board and other capacitances add up quickly so 15pf is kind of a worst case number. And no, the BAV99 are drawn correctly if it was for a low impedance clamp circuit. Dumping excess current into supply rails isn't a good idea. But in this circuit the uneven leakage current through the diodes could introduce an assymetric loading of the input resistor. Dual diodes are often 2 dies in one package which in theory should come from the same wafer but there are no guarantees. With such a high input series resistor you don't really need an external clamping diode. The ESD diodes in the opamp are sufficient.
« Last Edit: February 16, 2020, 07:49:13 pm by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline SaabFAN

  • Frequent Contributor
  • **
  • Posts: 702
  • Country: de
Re: Active probe?
« Reply #8 on: February 18, 2020, 08:04:58 pm »
What about this thing:https://github.com/dkramnik/Eagle-Public/tree/master/instrumentation-active-probe-elektor ?

I think it's been around for ages and the ones sold on ebay for 20€ (based on this circuit) provide decent results.

Offline Warhawk

  • Frequent Contributor
  • **
  • Posts: 522
  • Country: 00
    • Personal resume
Re: Active probe?
« Reply #9 on: February 21, 2020, 05:18:26 pm »
What you need is this http://welecw2000a.sourceforge.net/docs/Hardware/Aktiver_Tastkopf_mit_OPA659.pdf
The author is active here on the forum too.

I built three pieces and now I am doing actially a power supply with this IC http://www.ti.com/product/LM27762
I'll piggy back it to the board and supply it from USB.

Close enough.

Best regards, Jiri

Offline SaabFAN

  • Frequent Contributor
  • **
  • Posts: 702
  • Country: de
Re: Active probe?
« Reply #10 on: February 23, 2020, 10:07:14 am »
What you need is this http://welecw2000a.sourceforge.net/docs/Hardware/Aktiver_Tastkopf_mit_OPA659.pdf
The author is active here on the forum too.

I built three pieces and now I am doing actially a power supply with this IC http://www.ti.com/product/LM27762
I'll piggy back it to the board and supply it from USB.

Close enough.

Best regards, Jiri

I would suggest using one or two LiPo batteries and recharge them via USB.
Much one less cable that is being dragged along the bench when trying to measure something :)


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf