Keyboard Shortcuts
ctrl + shift + ? :
Show all keyboard shortcuts
ctrl + g :
Navigate to a group
ctrl + shift + f :
Find
ctrl + / :
Quick actions
esc to dismiss
Likes
Search
A homemade diode power sensor for HP meters
|
Solder-tail version of the same connector, also from Newark, also at unusual unit price.
Roy - if this is contrary to group guidelines, please accept my apology (I presumed that this would be of interest due to the very specialized nature of the parts and the prior reference to an ebay seller being defunct.) Below is a listing of the file names I retrieved using internet archive during my brief review of the thread this morning (anyone could do this, but of course it will be convenient to have them cached online). I've not uploaded them anywhere at this point - they can be retrieved using internet archive to search on each of the (defunct) URLs that were in the postings from earlier times. If someone can tell me where these files should be uploaded, I'll be happy to do that. (DOS command of the directory where I put the files retrieved) Directory of D:\848x homebrew power sensor 04/10/2025 05:22 PM <DIR> . 04/10/2025 05:22 PM <DIR> .. 04/10/2025 10:54 AM 2,658,520 08484-90019-PowerSensor.pdf 04/10/2025 11:01 AM 5,261,326 1975-10.pdf 04/10/2025 10:51 AM 4,021,004 20181117_170657.jpg 04/10/2025 10:50 AM 3,350,675 20181117_173101.jpg 04/10/2025 10:52 AM 234,284 8481schem.png 04/10/2025 10:55 AM 11,393,180 8484a.pdf 04/10/2025 10:56 AM 211,615 HP 11683A H01 ops_service.pdf 04/10/2025 11:03 AM 6,653,962 WTG_RefGuide_F1128_sm_web.pdf |
|
How about this location: David, thanks for collecting these files.? I was not able to collect the files from the now-broken references so I appreciate your efforts. Lynn C. On Thu, Apr 10, 2025 at 5:27?PM David Feldman via <wb0gaz=[email protected]> wrote: Solder-tail version of the same connector, also from Newark, also at unusual unit price. |
|
Hi Sergey,
Many institutions have recently decided that they need to know the location of their customers. For those that are ignorant about internet fraud, using Microsoft and Apple products, out-of-the-box, suffices, as they will all blab everything they know about you to anyone that asks during a web connection. Those with moderate knowledge of such things often use a variety of nonsense blockers that abort such requests for location as a matter of course... or spoof them, giving a random location. Those that use VNC, are skipped around the world until they are virtually at another location, so any information their browsers give is someone elses. Brave is particularly bad in this regard, as it dutifully blocks or spoofs just about everything asked that might identify you and your location... So, unless you give a little, most of commerce will ignore you, and you will be as lonely as the lighthouse keeper on the edge of the universe... I find I have to allow brave to give my location, for certain transactions, or they fail to work for me. -Chuck Harris On Wed, 9 Apr 2025 14:45:20 -0700 (PDT) "Sergey Kubushyn" <ksi@...> wrote: On Wed, 9 Apr 2025, Sergey Kubushyn wrote: |
¿ªÔÆÌåÓý
Reading this thread reminded me of my 431 & 432. Anyone know where I can find a copy of the 8402A power meter calibrator manual? Not at the usual suspects. John??? KK6IL On 4/10/2025 10:10 AM, David Feldman
via groups.io wrote:
I went back to the start of the thread to capture local copies of the various cited references. Most of the links are dead, however, internet archive wayback machine got me to almost all of them, so I have the files cached here now (mostly commercial reference material from HP and Boonton, along with Tobias' hand drawn schematics (I don't believe Tobias published schematics after his first few drafts, and don't know if he maintains a repository anywhere.) |
|
The idea of putting together a home made sensor got my attention.
I created a small PCB, copied the design and tried it using
thru-hole components. Some of the resistors matched the values
used by HP to the last digit...
?
I had a few 435's and not wanting to use a HP multi-pin C-091
cable I used a standard Tuchel multi pin connector found in my
junk box. Luckily, the front panel cut-out is standard C-091..
?
Using 1N5711 (5082-2800) as detector diode and a pair of U310's
+ a 2N2222, I got slightly more than 1dB per scale division.
?
Tryng to reduce the sensitivity (or gain) by reducing the emitter
resistance in the design, the sensor became unstable. I tried placing the PCB in a
small metal case which helped reducing the sensitivity to
external fields (such as power line noise emanating from my lab bench
lamps). The input resistor (15 Ohms) value also affected the function
even though I placed the detector diode right at the RF connector and
carefully kept the rectified and fitered DC output voltage away
and using small ferrites for isolation.
?
I also found out that the dymanic log-range was "a bit" limited.
?
1N5711 works for RF up to several hundred MHz
and at some gain settings, the frequency response was
almost flat but I could never achieve a correct gain while
at the same time maintaining a flat frequency response.
?
Keeping the sampler square wave inside or outside the
screened cable made no difference. At least not for the
power levels I tried.
?
It would be interesting to know exactly what kind of
detector type diode HP used in the 8484 sensor...
?
Cheers
?
Ulf
SM6GXV
?
?
? |
|
It should be possible to get good performance from a diode detector if designed carefully, even when using regular SMD parts. I've got some experience of designing detectors for use in ALC systems both at work and here at home. Many years ago I designed a couple of diode detectors to act as a poor person's alternative to the HP 8473C diode detector. I have an 8473C here but the 8473C doesn't work well below about 50 MHz.
?
?
I generally use these in the transition to linear region for levelling signal generators but I designed them to be flat down to 100 kHz. Of course, 'flat' is open to interpretation but these are about as flat as I can measure across this frequency range.
I managed to get a single diode detector to work well from 100 kHz up to 2.5 GHz.
?
However, for real world levelling of signal generators it's best to use a dual diode as this provides rejection of even order harmonics. This comes at the expense of bandwidth but I managed to make a dual diode detector that was flat from 100 kHz to 500 MHz with maybe 0.1dB droop by just over 1 GHz.
?
I still use these today and also use them for metrology as it's possible to calibrate the dual detector at 100 kHz using a Keithley 2015 DVM and then use the same cal factor right up to 500 MHz and apply a 0.1 dB correction slope from 500 MHz to 1 GHz. The single diode is essentially flat up to 2.5 GHz.
?
One design goal was to ensure the input VSWR was low and I managed to achieve a VSWR of < 1.03:1 up to 1750 MHz and 1.04:1 by 2000 MHz for the single diode detector. This helps to minimise mismatch uncertainty.?
?
If I revisited this today I think there are much better diodes available now in smaller packages.
?
?
?
? |
|
Just to clarify, the VSWR plot is taken directly into the detector. There is no attenuator ahead of the detector. It's just a precision 50 ohm termination followed by the diode. This level of performance was achieved by modelling every component in the detector and optimising it on a simulator for flattest response and lowest VSWR.
?
One issue using these detectors down in the square law region is that of thermal drift. This can be a real problem especially when handling the detector. They are fitted inside metal bodies and look similar to the 8473C in size and shape and connector type. However, just the heat from handling it can cause drift that takes a while to settle when using it in the square law region. I have used them in the square law region but only for casual research tasks because of the thermal drift issue.. I haven't tried using this detector with a chopper circuit but I do recall testing them down to about -40 dBm. I've got some results somewhere or I could retest them in 1dB steps down to -40dBm. I also haven't tried to thermally compensate the detector. |
to navigate to use esc to dismiss