Boonton 41-4B power sensor repair advice

[xrunner]

Need advice to repair a Boonton 41-4B power sensor. Pics attached (yes the year is wrong I was setting the clock on the microscope and made a mistake).

I've had issues with this power sensor for the past year, in that it's VERY erratic. It might be measuring the internal cal signal for example (0.0 dBm) and if you barely touch the sensor the reading goes whacko (technical term). I have a different sensor and nothing like that happens while using it.

I've inspected the inside before, but I never found anything wrong. This time I decided to go further and start using a microscope. I think I found the issue - a failed wire connection to what appears to be a 100 ohm precision resistor (it measure about 100.2 ohms. There are two of them that are the same part). These are the two red components in the pics. Since they are essentially in parallel coming off the center rf input, it must be designed to provide a 50 ohm impedance to the source.

What can I use as a substitute 100 ohm resistance (I would replace both of them)? I have SMD 1206 / 100 ohm parts on-hand that would be about perfect for the size needed. It appears from initial research that SMD resistors are good into the GHz range.

Please advise - Thanks.

[Kosmic]

Ouch 

Those 2 100\$\Omega\$ resistor in parallel form the 50\$\Omega\$ termination on the input of the probe. It's going to be really hard to replace them with something equivalent to exceed 1GHz with acceptable S11.

The problem with SMD components is going to be fixing them to the aluminum chassis. Personally, I would try to find some good (low inductance) through hole 100\$\Omega\$ resistors.

In general 2 x 100\$\Omega\$ is better than 1 x 50\$\Omega\$.

Good luck!

[jonpaul]

Use a selected 100.0 Ohm 1% 1/8 or 1/4W axial MF or CF.

Avoid spyril cut resistors.

Even a carbon comp 100 Ohm 1/8 W can work if selected.

At 1 GHz the axial small watts are OK.

Jon

[Kosmic]

Use a selected 100.0 Ohm 1% 1/8 or 1/4W axial MF or CF.

Avoid spyril cut resistors.

Even a carbon comp 100 Ohm 1/8 W can work if selected.

At 1 GHz the axial small watts are OK.

Jon

The inductance on Metal Film or Carbon Film is going to be too high I think. Carbon Composite would be a lot better.

[edpalmer42]

Have you tried to fix the resistor?  Put the lead back in place and touch the connection with your soldering iron.  If the central point measures 50 ohms to ground, you're done.

Ed

[xrunner]

Ouch 

Yea really. Thanks for the schematic, wasn't able to find it earlier. Bad news follows ...

Have you tried to fix the resistor?  Put the lead back in place and touch the connection with your soldering iron.  If the central point measures 50 ohms to ground, you're done.

Ed

I was going to attempt to but when I touched the end of it with my soldering iron the end of the resistor just fell off. In any case there was hardly any sliver plating left to solder to anyway. This is old parts I'm dealing with.

First thanks to everyone for their inputs I was studying the options and before that I was going to scab in a cheap 50 ohm resistor in place of the broken red one, just to see if the sensor would behave at a low frequency - i.e. without the erratic way it was acting. But the party is over for now.

When I heated up the place where the two red resistors and two diodes meet (center pin of the N connector), the solder indeed melted. However, even being careful all the parts moved a bit, and one of the tan colored 160 ohm resistors broke, even though it was at the other end of a diode. Wow really?  It's that delicate? 

So that puts an end to things for now. I will put all the parts in a bag, and perhaps get back to it at a future date. Fortunately I have another Boonton sensor (41-4A) that works, even though it doesn't have the frequency rating of this one, it's OK for my needs.

Any closing advice accepted of course. If you have suggestion as how to proceed at this point, please advise

[edpalmer42]

Many years ago, I bought a dead 41-4A sensor.  The problem turned out to be a cracked capacitor.  Once I had it working again, I decided to see what mods I could do since it was a spare sensor.  I removed the original diodes and replaced them with a couple of random Schottky small-signal diodes to hopefully survive accidental overloads.  I also increased the main isolation capacitor to improve the lower frequency response.  My probe used a capacitor on the input side of the 100 ohm resistors so that DC wouldn't fry them.

At 1 GHz, my modified probe is within about 0.2 dB of my R&S SME03 generator from 0 dBm down to about -50 dBm.  The low frequency reading doesn't move from 1 MHz down to the generator's low end of 5 KHz as compared to its normal -3 dB frequency of 200 KHz.  This is without any calibration.  I went looking for my notes, but I can't find them so I can't give you part numbers or values.

Bottom line is that these things aren't rocket science.  Remember that even the official probes have to have their calibration constants entered into the meter.  As long as you aren't dealing with a requirement for absolute accuracy or crazy high frequencies, just have fun!

Ed

[xrunner]

Bottom line is that these things aren't rocket science.  Remember that even the official probes have to have their calibration constants entered into the meter.  As long as you aren't dealing with a requirement for absolute accuracy or crazy high frequencies, just have fun!

Ed

Well the Boonton schematic actually says -

"When making your own probe, you may use ..."

So OK, you inspired me to continue. I'll try to re-make the probe using the original diodes. I'll post whatever progress I make in this thread. I'll start by removing the diodes, 100, and 160 ohm resistors and clean up the area, and substitute new resistors. I'll have to order the 160 ohm parts.

[Kosmic]

Well, everything is possible if your not necessarily interested in accuracy and higher GHz range.

This guy made his own probe 

[TimFox]

Use a selected 100.0 Ohm 1% 1/8 or 1/4W axial MF or CF.

Avoid spyril cut resistors.

Even a carbon comp 100 Ohm 1/8 W can work if selected.

At 1 GHz the axial small watts are OK.

Jon

The inductance on Metal Film or Carbon Film is going to be too high I think. Carbon Composite would be a lot better.

The inductance of low-resistance film resistors is often overestimated, but rarely specified.
Carbon composition resistors suffer from internal capacitance problems at high resistance values.
One should be able to use cermet or "thick film" resistors (with reasonable resistivity) that aren't helixed (spirals are planar geometric objects).

[edpalmer42]

Here's a web page that talks about various RF probes.

https://web.archive.org/web/20110303221109/http://members.shaw.ca/novotill/RfProbeSensorMvUw/index.htm

Note that unlike most probes, ours have the terminating resistors exposed to DC that might be coming from the DUT.  Even if they didn't burn out, thermal cycling could stress the solder joint and cause eventual failure.  I thought that my probe had the resistors protected from DC, but it turns out that I was wrong.  The same failure mode also exists with repeated exposure to high RF levels.

I've also attached a French document that talks about repairing a Boonton probe.

Ed

[xrunner]

Well, everything is possible if your not necessarily interested in accuracy and higher GHz range.

Yea under 1 GHz is pretty much going to cover my needs. 

This guy made his own probe 

I did see that. Seems to work reasonably well.

Here's a web page that talks about various RF probes.

https://web.archive.org/web/20110303221109/http://members.shaw.ca/novotill/RfProbeSensorMvUw/index.htm

Note that unlike most probes, ours have the terminating resistors exposed to DC that might be coming from the DUT.  Even if they didn't burn out, thermal cycling could stress the solder joint and cause eventual failure.  I thought that my probe had the resistors protected from DC, but it turns out that I was wrong.  The same failure mode also exists with repeated exposure to high RF levels.

I'm sure the sensor I'm dealing with was abused, because I know the ham very well I got it from. He gave me a Boonton Model 42A and the sensor I'm trying to repair. He got them at a hamfest but didn't understand how to use it to test his handheld transceiver, which puts out either about 1 W or about 5 W. I'm pretty sure he zapped the sensor with of of those two power levels because he didn't understand he would need an attenuator, but for how long I'm not sure. 

I've also attached a French document that talks about repairing a Boonton probe.

Ed

Oh nice thanks.

By the way, does anyone know where I can get the connectors for the Boonton sensor cable (and preferably the correct cable wire also). I want to put new connectors on the cable or build a new one, because the one I have isn't in great shape. I can't manage to find the parts anywhere. I know here it comes ... they aren't available any longer. 

[edpalmer42]

Looks like the connectors are available on the standard auction site, but only at the cost of one or two body parts.      If you have a strong stomach, search for < Amphenol microphone connector "2 pin" >.  I think that's the right connector, but you should do some further checking.

If you have to replace the cable itself, I think I read somewhere that it's very special.  Due to the very low DC values that will be present at low RF levels, the cable has to be low microphonic, low triboelectric, shielded and yet physically rugged and flexible.  IOW, pure magic!

[Kosmic]

By the way, does anyone know where I can get the connectors for the Boonton sensor cable (and preferably the correct cable wire also). I want to put new connectors on the cable or build a new one, because the one I have isn't in great shape. I can't manage to find the parts anywhere. I know here it comes ... they aren't available any longer. 

The connector is Amphenol 80-MC2M. They are not made anymore.



As for the cable it look pretty standard to me (shielded with 2 conductors. see picture below). I made a new one using Canare L-2T2S Audio Cable and could not see any differences in term of performance.

[xrunner]

Looks like the connectors are available on the standard auction site, but only at the cost of one or two body parts.      If you have a strong stomach, search for < Amphenol microphone connector "2 pin" >.  I think that's the right connector, but you should do some further checking.

Ah thanks I found them on Fleabay - "AMPHENOL 80-MC2M 2 PIN OFF-SET MICROPHONE PLUG FOR COLLINS HEATHKIT BROWNING NOS"

  $39 to $50 for ONE? They can shove it where the sun don't shine!

If you have to replace the cable itself, I think I read somewhere that it's very special.  Due to the very low DC values that will be present at low RF levels, the cable has to be low microphonic, low triboelectric, shielded and yet physically rugged and flexible.  IOW, pure magic!

You mean I can't use the roll of old Radio Shack speaker wire in my junk box? 

The connector is Amphenol 80-MC2M. They are not made anymore.

That's what I was afraid of, and I'm also afraid of NOS prices. 

As for the cable it look pretty standard to me (shielded with 2 conductors. see picture below). I made a new one using Canare L-2T2S Audio Cable and could not see any differences in term of performance.

I could just unsolder the connector pins, clip back the cable a few inches past where it's been flexing for 40 years, and make it all back up again. I did see that brass ferrule that is crimped on the braid. I doubt I can use it again. I'll have to figure out to get a good ground from the shield to the connector housing. Do the two signal wires also have their own shield and are these shields tied to the main outside shield?

[Kosmic]

Do the two signal wires also have their own shield and are these shields tied to the main outside shield?

I retract what I said earlier about the cable. It appear to be a little bit more complex than what I thought initially. The 2 conductors are also shielded and there's some kind of black isolator in between the second shield and the conductor wire.

[alm]

Low noise cables sometimes use a graphite coating between layers to reduce friction and hence triboelectric effects. This is not an insulator and if anything slightly conductive. It serves as lubraction. For example this is discussed in this thread on the Keithley / Tek forum or this picture.

[TimFox]

Another approach in low-noise cables was using a layer of conductive (carbon-loaded) plastic for the same purpose.

[Kosmic]

Low noise cables sometimes use a graphite coating between layers to reduce friction and hence triboelectric effects. This is not an insulator and if anything slightly conductive. It serves as lubraction. For example this is discussed in this thread on the Keithley / Tek forum or this picture.

I'm all aware of the carbon layer used sometimes to reduce triboelectric effect.

Nevertheless, In this case, the black coating is non-conductive and seem to isolate the center conductor from the second shield.

[edpalmer42]

Low noise cables sometimes use a graphite coating between layers to reduce friction and hence triboelectric effects. This is not an insulator and if anything slightly conductive. It serves as lubraction. For example this is discussed in this thread on the Keithley / Tek forum or this picture.

I'm all aware of the carbon layer used sometimes to reduce triboelectric effect.

Nevertheless, In this case, the black coating is non-conductive and seem to isolate the center conductor from the second shield.

But if you were looking for a replacement cable, one with such a carbon layer might be helpful, both for shielding and reducing triboelectric effects.  A while ago I repaired an XLR-equipped microphone cable with just such a layer.  That might be a starting point for a cable search - rugged, flexible, shielded, low triboelectric.  Hmmmm.

Ed

[Kosmic]

Like I said, I used Canare L-2T2S in the past. Was working fine and did not notice any differences, even at -60 dBm.

I remember looking around and never found a double shielded cable like that.

[xrunner]

I removed all the components from the center pin. I took the opportunity to remove the input hardware and clean them up.

When the solder is melted on the top on the center pin, all the components touching it slide around. This because the chip capacitor acts as an out-of-control force pushing in various directions as it will making it nearly impossible to work on individual components in that area. In order to hold things down, I made up a little assembly to press the chip capacitor down while the solder is melted, so I can solder two diodes and two SMD resistors in accurate positions.

I have a little bit of material coming that I want to use in concert with the SMD resistors, so I'll get to that soon.

[xrunner]

Here's my plan. I got some 0.1 mm thick copper sheet and cut off a strip with a width equal to a 1206 SMD part (0.06 "). I'm making a connection point with "wings" for the 1206 100 ohm resistors that doesn't require soldering directly on the pedestal. Trying to get all the components soldered in the right place without disturbing other components already there is too difficult with the parts I'm using. With this I can tack the SMD resistors on to the wings without melting the center area.

So I first soldered the copper wing with chip capacitor on top while being pinned down. I checked the capacitance afterwards to make sure it was correct. Now the center area does not again have to be completely melted. I can tack on the SMD resistors on the wings and then tack the diodes on top of the chip capacitor quickly, without the whole area melting.

Anyway that's my solution, others may come up with different ideas based on the physical construction of the components being used.

[xrunner]

Today I installed the two 100 ohm 1206 size resistors using more copper strips. They have a little bend in them so the assembly has a just a little flex in it.

[edpalmer42]

Do you have access to a VNA?

It would be interesting to see what the impedance of your rebuilt probe looks like compared to an official one.  I think that's the purpose of the offset washers that the resistor leads were clamped under.  By turning the washer you could tweak the termination.

Ed