Bird 50dB sampler power limit

[ZL2BKC]

Last week I received a package from Jim W6PQL so somtime in the future I will need to measure 1KW @ 144MHz with a level of confidence

I don't have a Bird 1KW element to "pin" my wattmeter with, and I am struggling to justify in my mind buying something I will only need a few times.
However I do have a Bird 50dB 25-1000MHz sampler - but the catch is it's rated at 500W

Given 1KW elements (and higher) are available for the Bird, I wonder Why the 500W Limit??

I have previously measured the sample level thru my Bird 43 at 50dB on the network analyzer and it was close to 50dB across the full range, so the plan is to feed the output (1KW in -> 10mW out) into a 10dB attenuator to HP 8482A and HP Power Meter. In fact this will be more accurate than a Bird Element (which seems to be +/- 5% or more)
Plan B is to use a calibrated length of RG213 to reduce the level to 500W (or maybe 250W for my 100C element) - stringing together all of my field day coax to get this loss means doesn't sit well in my mind.

As an asside my dummy load is only rated at 500W as well, but I have several reports the oil filled Bird dummy loads are good for 1KW or more - at least for short durations.

Has anyone managed to use a Bird Sampler at a higher power limit? Any technical reason why they can't work??

73
Wayne ZL2BKC

[VK2OMD]

Last week I received a package from Jim W6PQL so somtime in the future I will need to measure 1KW @ 144MHz with a level of confidence

I don't have a Bird 1KW element to "pin" my wattmeter with, and I am struggling to justify in my mind buying something I will only need a few times.
However I do have a Bird 50dB 25-1000MHz sampler - but the catch is it's rated at 500W

Given 1KW elements (and higher) are available for the Bird, I wonder Why the 500W Limit??

I have previously measured the sample level thru my Bird 43 at 50dB on the network analyzer and it was close to 50dB across the full range, so the plan is to feed the output (1KW in -> 10mW out) into a 10dB attenuator to HP 8482A and HP Power Meter. In fact this will be more accurate than a Bird Element (which seems to be +/- 5% or more)
Plan B is to use a calibrated length of RG213 to reduce the level to 500W (or maybe 250W for my 100C element) - stringing together all of my field day coax to get this loss means doesn't sit well in my mind.

As an asside my dummy load is only rated at 500W as well, but I have several reports the oil filled Bird dummy loads are good for 1KW or more - at least for short durations.

Has anyone managed to use a Bird Sampler at a higher power limit? Any technical reason why they can't work??

73
Wayne ZL2BKC

I don't know how THAT sampler works, but a common technique is to make a current transformer and load it with a resistor / attentuator to give a controlled source impedance and power ratio into a known load. In that type of construction, typically the resistor or attenuator limits the maximum continuous power rating.

To build a sampler of that type is not a big variation of a normal Bird 43 VHF+ sampler, a bit simpler since it does not have a detector.

Owen

[ZL2BKC]

Thanks Owen

If it had an in built attenuator then I would expect some upper power limit which may explain the 500W limit. Not game enough to open it up to find out.

I have opened up faulty elements for repair by sitting in paint thinners overnight, but in this case the BNC would sit in the sovent which will probably react with the dielectric.

--
Wayne ZL2BKC

[VK2OMD]

Wayne, this is a simple sampler I made for HF.



It is rated for 2kW and has a coupling factor of -37dB, flat to within 0.1dB over HF. The shortcut of the design is that the source impedanc is not 50 ohms, but closer to 20. 50 ohms is achievable, but increases the demands on the resistor dissipation.

I expect the Bird couple may well have an equivalent source impedance of 50 ohms.

Owen

[ZL2BKC]

Wayne, this is a simple sampler I made for HF.
I expect the Bird couple may well have an equivalent source impedance of 50 ohms.

Owen,

Nice sampler. Not sure how well it will perform at VHF/UHF frequencies, but I may just steal that idea and give it a try.

I had an idea how to determine if there was any in line attenuation by sweeping the port with the network analyzer. Interesting results:

Sampler output port measured with VWNA

The smith chart shows it's nowhere near 50 ohms, and the plot skirts the outside perimeter indicating it is a complete open circuit.
At low frequencies it appears like a 270pF capacitor (including DC as confirmed by my multimeter), so it looks to be completly passive inside. Probably nothing more than an open wire loop and a capacitor to equalize the performace. To find out more I would need to model it completly.

There may be a diode limiter in circuit (which may explain some of the capacitance), but my diode check function on the multimeter shows nothing.

One would almost think it was broken, but a quick double check shows it actually has 50dB coupling in the Bird 43.

73,
Wayne ZL2BKC

[VK2OMD]

...
Nice sampler. Not sure how well it will perform at VHF/UHF frequencies, but I may just steal that idea and give it a try.

I had an idea how to determine if there was any in line attenuation by sweeping the port with the network analyzer. Interesting results:

Bird 50dB Sampler.png

The smith chart shows it's nowhere near 50 ohms, and the plot skirts the outside perimeter indicating it is a complete open circuit.
At low frequencies it appears like a 270pF capacitor (including DC as confirmed by my multimeter), so it looks to be completly passive inside. Probably nothing more than an open wire loop and a capacitor to equalize the performace. To find out more I would need to model it completly.

There may be a diode limiter in circuit (which may explain some of the capacitance), but my diode check function on the multimeter shows nothing.

One would almost think it was broken, but a quick double check shows it actually has 50dB coupling in the Bird 43.

73,
Wayne ZL2BKC

Another approach for samplers is a voltage divider. These are very popular in the ham 'literature' (QST etc), though IMHO they are much more challenging in terms of broadband calibration.

What you are seeing looking into the coupler port (270pF) may be the lower capacitor of a divider. If it was a simple two component divider of 50dB, the 'small capacitor' would be 0.85pF... quite feasible as a broadband divider for frequencies above about 30MHz. It might be as simple / crude as that! That 0.85pF would limit the upper frequency (for low insertion VSWR and frequency independent response) to something a bit less than 1000MHz. Still, it is cheap to produce!

BTW, the sampler I showed (which will not be good at UHF as implemented) is more fully described at http://vk1od.net/module/RfCurrentSampler/index.htm . I didn't sweep it above 30MHz, but I expect it would roll off. There are some things that could be done to improve high frequency response, but the design was intended for HF... and it works fine for HF.

Calibrated directional couplers are another possible solution to your measurement challenge.

Owen

[ZL2BKC]

Thanks Owen.

Your suggested capacitor voltage divider approach is probably how the Bird element is designed based on what I have measured thru the BNC port. The 0.85pF capacitor would no doubt be a metal disc at the bottom of the element to achieve the capacitance off the thru line section, and it will be divided down using a capacitor.
I say metal disk as there is no change in coupling no matter what angle the element is oriented.

If it's that simple then the voltage breakdown of the capacitors would set the upper power limit. The 0.85pf section would be withstanding the highest voltage thus setting the upper power limit. The 270pf class cap would not be stressed voltage or power wise.

I think it's worthy of a try at 1KW when I complete the amplifier and I will report back the findings. Bird products have a pretty good safety margin built in anyway - I suspect the power rating is more appropriate at the 1GHz end of the freq range, and at 150MHz there is slightly more headroom.

73,
Wayne

[VK2OMD]

Thanks Owen.

Your suggested capacitor voltage divider approach is probably how the Bird element is designed based on what I have measured thru the BNC port. The 0.85pF capacitor would no doubt be a metal disc at the bottom of the element to achieve the capacitance off the thru line section, and it will be divided down using a capacitor.
I say metal disk as there is no change in coupling no matter what angle the element is oriented.

If it's that simple then the voltage breakdown of the capacitors would set the upper power limit. The 0.85pf section would be withstanding the highest voltage thus setting the upper power limit. The 270pf class cap would not be stressed voltage or power wise.

I think it's worthy of a try at 1KW when I complete the amplifier and I will report back the findings. Bird products have a pretty good safety margin built in anyway - I suspect the power rating is more appropriate at the 1GHz end of the freq range, and at 150MHz there is slightly more headroom.

73,
Wayne

Your reasoning seems pretty sound Wayne. No warranty offered, it may be worth the risk.

It might be worth Googling for "xxx repair" or the like where xxx is the exact Bird part number... someone may have pulled one apart and have some hints or even photographs.

Cheers
Owen