Remote Power Sensing Works

[John Williams]

As some of you may know, I had remote power sensing working well on my HL1. I had created a board (Power2) that had enabled the AINx interface to the gateware so that I could have a remote board (FRSense) mounted at the output of my 50W linear amp. Power2 is no longer necessary as AIN1/2 is supported directly on HL2.

These 2 boards can be viewed on github - https://github.com/W9JSW/Hermes-Lite/tree/Amplifiers

You can see the FRSense board on my QRZ page. (note to self - time to revise the commentary)

I have now moved that over to my HL2b8 unit. Jim N2ADR had graciously added an unpopulated connector CN5 on his board for this purpose. I removed R12 and R15 and mounted the connector. I then connected the FRSense board to it. It all works! My sensing circuit is copied from Hermes so it agrees closely with that set of ADC values for power. When I key up using PowerSDR, the forward power now agrees with the spec for HL2.

I still have to work on some stability issues. The reverse sense voltage bounces around quite a bit, causing the SWR to incorrectly indicate a bad antenna connection. That may be due to improper cabling or the need for some decoupling. My boards had some interesting decoupling caps that are used on Hermes - NFM21PC224R1C3D on both ends of the cable.

For anyone to want to follow along, the boards are available on OshPark for anyone to order. ~$15 for 3 of them with free shipping. 

https://oshpark.com/shared_projects/z3pei6Ab

They could also be ordered from Elecrow in larger quantities using the gerbers that are on my github. I never built a BOM for the board, but will if anyone is interested.

Those that are using the Hardrock50 Amp may be able to tap into the forward/reverse power sensing circuit to feed this interface. I have not tried it. My amp is a early version of that amplifier design. One would need to measure the voltages being generated to insure they stay within the proper range of voltages for HL2. I think the values must be between 0 and 3.3V, but I am not sure. Maybe Jim knows the actual bounds. Of course the calibration will be different.

Note that my board does not agree with Jim's power sense values. When running Quisk or Spark, the power needs to be re-calibrated.

John - W9JSW

[Graeme Jury]

Hi John,

I forgot that you did that some time ago, I was just going to do the same thing for my radio and had already ordered the op amp so to save myself some work I will get a board from OshPark and save myself some time. Really appreciate your post John.

73, Graeme ZL2APV

[John Williams]

I will clean up the BOM in the morning.

--
You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/b773d977-2dc3-4bb5-90db-8421a579a64b%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[John Williams]

Here is the BOM.

I am still working on getting the voltages more stable. I see that HL2 has 2 caps on the AIN1/2 lines that are marked DNI. Will install some 0.1uf caps there first. Also going to try a shielded cable between the board and the filter. The one I have is open twisted and worked fine on my HL1.

John

--

[dick_...@hotmail.com]

Hi John

Thanks for the credit on the Alt Alex TX board , but it is a direct copy of the ALEX sense circuit designed by Graham Haddock KE9H. 

I noticed a significant difference in your  ground planes.  I have attached Pix of the two board layouts. I had air gaped the two ground planes with through hole diodes.  This can also be done with SMD parts and a narrower gap.  I had plenty of room on the Alt Alex board for through hole parts and a 8pin DIP.

If you put a DPDT switch on the output lines you can easily reverse the readings for any application since both sides are symmetrical.

I was considering relaying out my alt alex board with the two toroids mounted on either side and with SMA / BNC option and all SMD except for the two diodes. It is possible to mount the SMD diodes on a piece of perf board to extend their leads and place on the through hole footprint.  Note there is an edge mount BNC from Bomar that is currently $3.88 at Mouser [also available @ Digikey] Mouser # 678-361V504EFT. 

Third photo is of my version of KU6X TRS from QST August 2018 using the Bomar connector, more info, but no Pix on my page http://www.k9ivb.net/KU6X%20TRS/

If you wish to wait I will post my redesign [likely at the end of the month due to other commitments.]

Dick K9IVB

To unsubscribe from this group and stop receiving emails from it, send an email to herme...@googlegroups.com.

[John Williams]

I will be interested to see what you come up with, Dick. Had not occurred to me to separate the ground planes. Makes good sense. 

John

To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/b5ed1543-2756-4ec8-9a40-61fe73073dcf%40googlegroups.com.

[dick_...@hotmail.com]

Hi John

I have finished the redesign of the Alex circuit and have posted the results to my web page.

http://www.k9ivb.net/Directional_Coupler/

Also attached is basic documentation.  Boards are 3 for $9.45 at OSH Park.

Link on web page

Dick K9IVB

[John Williams]

Thanks Dick, Ordered some...

To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/f89a3af5-170b-4f86-85df-beb8f71d1229%40googlegroups.com.

[dick_...@hotmail.com]

[dick_...@hotmail.com]

[dick_...@hotmail.com]

I have read some discussions about how the Alex circuit works but they are not in this thread.

I did a theoretical analysis of the ALEX circuit with the following results.

I have attached a document and spread sheet to explain how I derived the above chart of numbers.  Perhaps it will be helpful in understanding and applying the circuit.

If someone would like to expand as to how the A/D's are programmed to function and how the results are interpreted, that too may be helpful.

When I get time I will put on my web page with the new PCB design http://www.k9ivb.net/Directional_Coupler/

Dick K9IVB

[Christopher KB3CS]

1mW into 50 ohms is 0.0 dBm.   :-P     ( 10*log(1) )

 - 2201 (base 3) -

On Tuesday, July 16, 2019 at 11:16:27 PM UTC-4, dick_...@hotmail.com wrote:

I have read some discussions about how the Alex circuit works but they are not in this thread.

I did a theoretical analysis of the ALEX circuit with the following results.

I have attached a document and spread sheet to explain how I derived the above chart of numbers.  Perhaps it will be helpful in understanding and applying the circuit.

If someone would like to expand as to how the A/D's are programmed to function and how the results are interpreted, that too may be helpful.

When I get time I will put on my web page with the new PCB design http://www.k9ivb.net/Directional_Coupler/

Dick K9IVB

[...]

[dick_...@hotmail.com]

Thanks for catching the typo.  It did make a difference in the 2nd & 3rd decimal place, but is less than 1LSB in a 12 bit A/D with 3V Max.

I put corrected files on my web page.

Dick K9IVB

[Steve Haynal]

Hi Dick,

I assume you are asking about the instrumentation ADC. This is a 12-bit value that ranges linearly from 0 to 4095 for 0V to the 3.3V power supply.

73,

Steve

kf7o

[dick_...@hotmail.com]

Hi Steve,

I was asking about both 12bit A/D's.  What you have just indicated is that Hermes-Lite is using +3.3V [Vdd} as the reference and not the internal 2.048V reference.  The Maxim part also appears to be bipolar technology and will load the Alex circuit differently than the TI part which is CMOS.  The operation of the 4-input device will also differ from the 8-input TI part, all of which is dependent upon how the FPGA is configured to program the A/D registers.. 

The N2ADR Filter board detector configuration is completely different from the Alex circuit and I did not try to analyze or compare. 

Meters on the output in the diagram are just to show other metering circuits will work depending upon the desired application.  When used with other hardware some modifications of the circuit [like transformer turns or the output attenuator] will need to be adjusted to match the application and the scale factor interpretation for 0 - 3V or up to 12V.  The spread sheet is handy to evaluate any circuit changes, to generate a calibration curve or to calculate the range of SWR .  I did not choose to include the bit reference to the 4096 bins in the spread sheet or any of the other calculations.

The circuit that John originally used had a phase reversal on the transformers [that was my pin numbering error] and I did correct on the new Schematic and PCB's.  The phase reversal switched the FWD and REV but did not affect performance.

Dick K9IVB