HF Probe build help needed

I'm building a probe for a Boonton 92EA meter. I found a schematic and pictures here.

formatting link

Page down to the SMD PCB and see the pictures and a schematic just above pictures. See the parts used under, THE UPDATED REV.B BOONTON PROBE CONTAINS: I have duplicated the parts and made a PCB.

(wherein, the problems may lie)

Here is a picture of the disassembled probe showing the board.

formatting link

The back side has a ground plane the extends from output connector to the 1000pf capacitors that connect to ground, but no further. I soldered the PCB top ground and bottom foils to the case through the holes in the case. (ok, the copper pipe) I'll install heatshrink to insulate and beautify! Here's a picture with tip, ground lead and heatshrink.

formatting link

My probe is flat to somewhere between 1 MHz and 2 MHz. Worst case on one range was 2% low, most ranges were with less than 1% of my scope reading. (yes, I converted Vpp to RMS) Yes, I know the scope is not a great reference, but if it's close to my scope, it is close enough to calibrate. (I hope) Some of the higher frequency loss may be in my measurement method. I'll take another stab at it tonight. I just realized my scope has an internal 50 ohm termination. My new plan is to install a BNC T on the scope input. Connect the sig gen to one side of the T and My probe to the other side of the T. At this time I have a frequency generator that will go to 10Mhz. :-( But, I'll get 10MHz working and find someone with better equipment to test further.

The Boonton Probe is supposed to work to 1.2GHz. Manual suggests a 3% error 100MHz to 1GHz and 7% 1.0GHz to 1.2GHz I don't know if I can duplicate those spec's on my kitchen table, but with your help I want to make the probe work to as high a frequency as possible. I want to make it good enough that it would be possible to send a probe and meter somewhere for calibration. As you can see I have wires to reverse the polarity going to the output connector If needed I can redo the PCB to make the crossover like Boonton did and connect the output connector directly to the PCB. Better, I may just reverse the orientation of the BAT62. Originally I was going hardwire the cable to the Probe and thought, I'll just connect the cable properly and eliminate the crossover. Plans changed when I decided cutting a Boonton cable in half made it to short. (I wanted two probes from one cable)

So, looking at the pictures: Do you see any problems with my PCB layout to cause HF loss? Do you see any problems with the probe tip or housing? The complete probe is 1/2 inch OD and about 6 inches long.

Thanks, Mikek PS. Hope I didn't leave anything out for Phil to bitch about! :-) Naw, I probably did.

Reply to
amdx
Loading thread data ...

Changing orientation won't work, I'd need to turn it upside down.

Reply to
amdx

As I said, I would change my measurement method and retest. Now it tests flat to 10 MHz which is the highest frequency I can generate. I started using the meter on the frequency generator as my reference, rather than the scope after checking that they had general agreement. At 10kHz the 92EA reads 2% to 5% low depending on the range, 1mv to 3V in 8 steps. At 10MHz the 92EA 0% to 1.2% low. Again all this is compared to the unknown meter on the HP651B.

If anyone is interested in the measurement method that failed and the one that worked, look here.

formatting link

Mikek

Reply to
amdx

Good work!

Reply to
Robert Baer

Ya, thanks, I guess I was writing to myself. Maybe I should just write notes to myself. It does help give me time to think where the problems may lie. I still have questions about foil layout. The RF section is pretty short, a cap, the diode is the RF AC section, then a resistor to the filter cap is a DC ripple section. I eliminated the ground plane from that section. I'd like to make it accurate to 1.2GHz and function to 8 GHz like the Boonton model. (I don't know how I would know though) Is there a better foil width? Is a thinner width better at high frequency? Was I right to eliminate the ground plane from the AC section. The original probe does that, but I don't know about the SMD version. Mikek

Reply to
amdx

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.