Author Topic: Active probes  (Read 15937 times)

0 Members and 1 Guest are viewing this topic.

Offline SafarTopic starter

  • Regular Contributor
  • *
  • Posts: 119
  • Country: ru
Active probes
« on: May 12, 2017, 10:35:27 pm »
A question on active probes. I have a Pico oscilloscope and look for active probe for HF "non-invasive" measurements. Now I'm looking for used probes from Tektronix, Agilent etc.

Do I understand correctly that question is only in power supply and 50 Om termination? And the digital interface in these probes is used only to store the calibration values.

Is there a model (not new) with BNC connector that is not capable with other oscilloscopes?

Thanks
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3634
  • Country: us
  • If you want more money, be more valuable.
Re: Active probes
« Reply #1 on: May 12, 2017, 10:40:56 pm »
Tektronix has remote power adapters for some of their active probes - can't remember the models exactly. What bandwidth are you needing?

Single ended, differential, max voltage?
Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 

Offline swingbyte1

  • Contributor
  • Posts: 17
Re: Active probes
« Reply #2 on: May 12, 2017, 11:05:03 pm »
A question on active probes. I have a Pico oscilloscope and look for active probe for HF "non-invasive" measurements. Now I'm looking for used probes from Tektronix, Agilent etc.

Do I understand correctly that question is only in power supply and 50 Om termination? And the digital interface in these probes is used only to store the calibration values.

Is there a model (not new) with BNC connector that is not capable with other oscilloscopes?

Thanks
Look at the tek1103 series probe power supply for information on the way to connect with other Scopes. There is a diy project around here

Sent from my Nexus 5 using Tapatalk

 

Offline SafarTopic starter

  • Regular Contributor
  • *
  • Posts: 119
  • Country: ru
Re: Active probes
« Reply #3 on: May 13, 2017, 10:27:12 pm »
OK, thanks all. I think that Tek P6243 or P6245 will be an good option. They need -5v and +5v power (and compensation in P6245). Plan to add cable from internal board and connect it to a separate unit via mini-Din connector

 

Offline MrW0lf

  • Frequent Contributor
  • **
  • Posts: 922
  • Country: ee
    • lab!fyi
Re: Active probes
« Reply #4 on: May 13, 2017, 10:45:12 pm »
« Last Edit: May 14, 2017, 09:19:08 pm by MrW0lf »
 
The following users thanked this post: saturation, bitseeker

Offline exe

  • Supporter
  • ****
  • Posts: 2559
  • Country: nl
  • self-educated hobbyist
Re: Active probes
« Reply #5 on: May 14, 2017, 08:30:07 am »
$15 for new 1.5 GHz probe :)

Looks cool, just I'm not sure about input impedance. How is it calculated? I see the gate capacitance is > 1pF which is already bad-enough at 1GHz. Did I miss something?
 
The following users thanked this post: commongrounder

Offline MrW0lf

  • Frequent Contributor
  • **
  • Posts: 922
  • Country: ee
    • lab!fyi
Re: Active probes
« Reply #6 on: May 14, 2017, 11:20:31 am »
Looks cool, just I'm not sure about input impedance. How is it calculated? I see the gate capacitance is > 1pF which is already bad-enough at 1GHz. Did I miss something?

No idea, just noticed, bought just in case along with some other stuff - but so far been too lazy to solder cable to it. But can find quite a lot info on the net, seems "accepted design", for example:
https://elektrotanya.com/files/forum/2009/10/e04a036.pdf
 
The following users thanked this post: exe

Offline abraxa

  • Frequent Contributor
  • **
  • Posts: 377
  • Country: de
  • Sigrok associate
Re: Active probes
« Reply #7 on: May 14, 2017, 06:31:59 pm »
It's a great find for sure, however personally I'd be a bit reluctant to purchase the one you linked to on ebay. Three reasons: first, he sells them as 1.5GHz probes when the BF998 is only rated for 1GHz (I doubt they incorporate a 50% BW margin). Second, he changed the design by using different passives - e.g. 5x2M instead of 10M (R1), 51R instead of 47R (R4), 5k1 for both R2 and R3 instead of 4k7 and 6k8, respectively. Capacitors are unknown, so who knows what he uses there, but he also replaced the input capacitor C1 by three capacitors in series. Third, C1 is supposed to be a PCB cap, not a discrete component. Hence, he turned a non-existant component into three existing ones.

I'm not against changing the values as such, they could of course yield better performance. However, the 3x?nF and 5x2M changes make me wonder if he had to "make do" because of component availability/cost in the Ukraine. Not only do they both add significant trace length on the input side (and thus increase stub length and impedance mismatch when probing 50R transmission lines) but also increase the non-linear impedance seen by the signal.

Also, what about the ground return path? The original Elektor design has a direct return where this has... none?

Overall, I'd try out the original design for sure but stay away from the design sold by the ebay auction seen above.
« Last Edit: May 14, 2017, 06:34:09 pm by abraxa »
 
The following users thanked this post: exe

Offline voltsandjolts

  • Supporter
  • ****
  • Posts: 2281
  • Country: gb
Re: Active probes
« Reply #8 on: May 14, 2017, 06:40:59 pm »
 
The following users thanked this post: MarkL

Offline joeqsmith

  • Super Contributor
  • ***
  • Posts: 11630
  • Country: us
Re: Active probes
« Reply #9 on: May 14, 2017, 06:43:04 pm »
RP wrote an article on a DC coupled probe as well that was dirt cheap to make. I made a few but they are long gone.  The BW may have been 50 to 100MHz?  All basic parts.  If interested, I can see if I can find the article.

Offline MrW0lf

  • Frequent Contributor
  • **
  • Posts: 922
  • Country: ee
    • lab!fyi
Re: Active probes
« Reply #10 on: May 14, 2017, 07:11:25 pm »
Overall, I'd try out the original design for sure but stay away from the design sold by the ebay auction seen above.

Damn now might have to actually solder cable to it for finding out :) Paid no attention to details, was cheap and was ordering SWR bridge anyway from same dude (because it was also cheap). If turns out crap... what you gonna do - cant always win with these ultra cheap momentary mood purchases...
BTW, about return path,  he says on one of pics:
"Do not forget to solder the ground wire somewhere on the ground surface of the probe..."

Edit: Fired it up with temporary connections - electrically works. Found some info in russian space:
Long Word doc, mostly in russian, with lots of schematics and this:


Rough translation of important stuff:
- input capacitance approx 0.7pF formed by C1-C3
- input resistance 10Mohm
- gate shift voltage is created by resistive divider R2-R3 and supplied thru R1=10Mohm
- input capacitance of BF998 is approx 2.1pF, because of that probe voltage will be 3x lower than actual voltage on DUT
- R4 must be matched to used cable impedance
- probe works from 100kHz to 1GHz, with max error of 2.5dB
- on >1GHz frequencies error significantly rises
- transistor gate and C1-C3 must not be placed on grounded plane or in shielded enclosure to avoid forming capacitance
- input wires max 10cm in length
- C1 feed wire must be at distance from PCB (ground plane?) and enclosure (shield)


« Last Edit: May 14, 2017, 09:23:18 pm by MrW0lf »
 
The following users thanked this post: saturation, abraxa

Offline serggio

  • Regular Contributor
  • *
  • Posts: 149
  • Country: ru
Re: Active probes
« Reply #11 on: May 14, 2017, 11:07:27 pm »
Interesting, how much goods him need to sell for purchase right soldering tools for SMD components  :D Probably in this design need focus to 0603 or 0402 capacitors at least for eliminate components/PCB capacitance .. Isoprophil alcohol can help remove colophony too.... Disaster  :palm:
 
« Last Edit: May 14, 2017, 11:11:31 pm by serggio »
 

Offline SafarTopic starter

  • Regular Contributor
  • *
  • Posts: 119
  • Country: ru
Re: Active probes
« Reply #12 on: May 15, 2017, 12:56:00 am »
Interesting, how much goods him need to sell for purchase right soldering tools for SMD components  :D
Gas oven is enough... And too much tin that also add capacitance :)

I seen this project early but think that Tek do much better. But bit pricey of course.
 

Offline MrW0lf

  • Frequent Contributor
  • **
  • Posts: 922
  • Country: ee
    • lab!fyi
Re: Active probes
« Reply #13 on: May 15, 2017, 08:38:15 am »
Naturally this is crude contraption and of course should go for Tek if money no issue. But I hear some even use Rigol scopes with amplitudes being thrown around up to 5dB between different ack modes with same input... So it not that critical when little money involved. Think I like this little soldering disaster and it was good purchase. T34 tank was also extremely crude but blasted many engineering perfections away. Crude slavic contraptions rock :popcorn:
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Active probes
« Reply #14 on: May 17, 2017, 05:54:32 am »
I would be absolutely shocked if that "1.5GHz" probe is really good for even 500MHz. Anyone want to buy one and evaluate it?
 

Offline Dubbie

  • Supporter
  • ****
  • Posts: 1114
  • Country: nz
Re: Active probes
« Reply #15 on: May 17, 2017, 07:30:56 am »
I bought one to have a play with. If it's any good I'll make a little enclosure for it. I'd be happy if it does 500Mhz
 
The following users thanked this post: MrW0lf

Offline MrW0lf

  • Frequent Contributor
  • **
  • Posts: 922
  • Country: ee
    • lab!fyi
Re: Active probes
« Reply #16 on: May 17, 2017, 07:43:17 am »
Naturally I have unit also, but absolutely no time right now to assemble & test properly, totally preoccupied with designing Arduino-PLC based home electrical system. Just ordered 0.5km of cabling :o Sorry in advance if it was s*it find on my part :( However I do find also that even if its 500MHz - ok for money already. So fingers crossed and good luck!
 

Offline abraxa

  • Frequent Contributor
  • **
  • Posts: 377
  • Country: de
  • Sigrok associate
Re: Active probes
« Reply #17 on: May 17, 2017, 10:31:04 am »
Naturally I have unit also, but absolutely no time right now to assemble & test properly, totally preoccupied with designing Arduino-PLC based home electrical system. Just ordered 0.5km of cabling :o Sorry in advance if it was s*it find on my part :( However I do find also that even if its 500MHz - ok for money already. So fingers crossed and good luck!
Nothing to be sorry for! It pointed me to the Elektor design, which I didn't know about before.
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Active probes
« Reply #18 on: May 17, 2017, 06:59:17 pm »
The probe interface also usually includes an offset signal from the oscilloscope to the probe to set the input offset.  The digital interface if present identifies the probe type and any calibration data.

If you are just dealing with HF frequencies up to 30 MHz, then something like Bob Pease's active probe is easy enough to build and has even lower input capacitance.
 

Offline exe

  • Supporter
  • ****
  • Posts: 2559
  • Country: nl
  • self-educated hobbyist
Re: Active probes
« Reply #19 on: May 17, 2017, 09:24:53 pm »
If you are just dealing with HF frequencies up to 30 MHz, then something like Bob Pease's active probe is easy enough to build and has even lower input capacitance.

Unprotected direct JFET input? How durable is this? I thought they a very susceptible for ESD...
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Active probes
« Reply #20 on: May 17, 2017, 10:04:19 pm »
I've never owned an active probe myself, but I've been under the impression that high performance active probes are notoriously fragile. Any sort of protection ends up compromising performance.
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 3012
  • Country: gb
Re: Active probes
« Reply #21 on: May 17, 2017, 10:35:58 pm »
Quote
If you are just dealing with HF frequencies up to 30 MHz, then something like Bob Pease's active probe is easy enough to build and has even lower input capacitance.

That looks like a neat circuit but I'd expect that it won't prove to be unconditionally stable. I think it could self oscillate up in the MHz region when probing certain (passive) circuits. I think it needs refining if you want to make something that should prove fully stable.




« Last Edit: May 17, 2017, 11:19:31 pm by G0HZU »
 

Online David Hess

  • Super Contributor
  • ***
  • Posts: 16545
  • Country: us
  • DavidH
Re: Active probes
« Reply #22 on: May 18, 2017, 02:20:20 am »
Unprotected direct JFET input? How durable is this? I thought they a very susceptible for ESD...

An ESD protection diode (it only takes one to protect a JFET) would at least double the input capacitance.

That looks like a neat circuit but I'd expect that it won't prove to be unconditionally stable. I think it could self oscillate up in the MHz region when probing certain (passive) circuits. I think it needs refining if you want to make something that should prove fully stable.

The 51 ohm source resistance improves stability.
 

Offline G0HZU

  • Super Contributor
  • ***
  • Posts: 3012
  • Country: gb
Re: Active probes
« Reply #23 on: May 18, 2017, 07:55:03 am »
I think that it's still going to be unstable with certain circuits and needs some refinement. I simulated this circuit and it isn't fully stable.
 

Offline rx8pilot

  • Super Contributor
  • ***
  • Posts: 3634
  • Country: us
  • If you want more money, be more valuable.
Re: Active probes
« Reply #24 on: May 18, 2017, 05:18:43 pm »
I think that it's still going to be unstable with certain circuits and needs some refinement. I simulated this circuit and it isn't fully stable.


Wishing I had your skills and software....   :-+
Factory400 - the worlds smallest factory. https://www.youtube.com/c/Factory400
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf