[TangerineSDR] Ferrous Metal connectors on Mag boards and SBCs

Scott Cowling scotty at tonks.com
Sun May 31 17:03:03 EDT 2020

Hi Dave,

No one is excluded from the design process, especially you from the 

I have sent out two schematic revisions, XA1 (April 24) and XA4 (May 4), 
to numerous people, including you.

I have received multiple suggestions from you, Dave N1HAC, Jules K2KGJ, 
John N8OBJ, and Gerry N5JXS all with valuable insights.

I have sent no updates since then, but that does not mean I am excluding 

The design has evolved (as has just about everything on TangerineSDR and 
PSWS). Most of the evolution of the magnetometer board has come as a 
result of your original design(s).

We originally had a PMOD design (my request). You pointed out the 
problem with 5V power and 3.3V signalling on the PMOD connector.

You showed, with your design, that we would not be able to shrink it to 
a PMOD form factor. Hence the larger board.

You showed that it would work at the end of 100 ft of cable. Then Jules 
showed that it could work a 400 ft.

I added the Lightning detector and DRDY signalling and packaged it as a 
Pi Hat.

Gerry and Nathaniel subsequently talked me out of the lightning 
detector, and the board went back to just a magnetometer.

I changed the addresses to allow magnetometers on both ends as an option 
(that will probably never be used), but this does not make them 
incompatible with your boards. It is simply a software feature that 
needs to be added to accommodate both designs.

Per Jules' recommendation, I added more bulk capacitance. I left the 
series inductor in place, but if it causes too much ripple on the 5V 
line, we can replace it with a zero-ohm resistor.

I will check the BOM for ferrous material in the connectors, as recommended.

As for the jumper in the shield, that was there from day 1 as an option. 
If disconnecting it causes a problem, leave it jumpered in. If 
connecting it causes a problem, jumper it out.

By the way, I have not contacted Becky from PNI on any issues, because I 
did not need any additional information from her. You asked the 
important questions and got answers, so I think we are all set for 
production quantities.

I will be sending out the latest plots and schematic out shortly for any 
additional comments, in case I missed any suggestions.

Scotty WA2DFI

On 05/30/2020 14:34, David Witten via TangerineSDR wrote:
> Jules,
> This will have to come to the attention of Scotty.  I am now completely 
> excluded from the design process for the magnetometer implementation.
> Dave
> On Sat, May 30, 2020 at 4:27 PM Julius Madey <hillfox at fairpoint.net 
> <mailto:hillfox at fairpoint.net>> wrote:
>     Dave,
>     I specifically note the change of shield material from copper alloy
>     to SS.
>     I just ordered the AMP jacks and Stewart plugs in the attached data
>     sheets.  Jack shield is plated phosphor bronze and plug shield is
>     plated copper.  I don't know if other vendors are changing shield
>     material re EMI performance (meaningful at GigE speeds but not at
>     100KHz differential i2c) and since magnetic properties of SS can
>     vary greatly depending upon specific alloy and cold working, it's
>     probably a good idea to stay away from them if at all possible if
>     the RM3100 will be mounted in close proximity to the SS part if the
>     hope is to see signals well below the microTesla range.
>     Perhaps Hyomin could provide some guidance on that ?
>     Jules
>     Delays in 1000 meters of twisted pair cable are, of course, too long
>     for 100KHz rate i2c exchange protocol, but
>     On 5/30/2020 2:21 PM, David Witten wrote:
>>     Jules, et. al.:
>>     FWIW:  The Amphenol drawing w/details for the RJ45 that I used on
>>     the prototypes is attached.  It indicates that the shield is tin
>>     coated stainless steel.  Depending on the alloy it may exhibit
>>     some degree of ferromagnetism.
>>     Dave
>>     On Sat, May 30, 2020 at 12:59 PM Julius Madey
>>     <hillfox at fairpoint.net <mailto:hillfox at fairpoint.net>> wrote:
>>         Tom,
>>         From a 5/4 email from Scotty explaining the design of his new
>>         board:
>>         "JMP6 allows us to ground one side of the shield only to
>>         minimize noise
>>         pickup and ground loops. Note that the CAT5 cable is shielded. "
>>         If I interpreted that as a 'recommendation', my bad ...
>>         Molex has at least one RJ shielded plug using brass .... mine
>>         were cheap eBay stuff ... so there appears to be no
>>         problem.....just have to order the good stuff....non ferrous
>>         metal construction is available on both RJ45 jacks and plugs.
>>         Jules - K2KGJ
>>         On 5/30/2020 11:45 AM, Tom McDermott wrote:
>>>         I have no idea where the idea that the shield should be
>>>         unconnected at the remote end came from.
>>>         In my opinion this is contrary to what probably needs to be
>>>         done.  I think Scotty put a jumper on that
>>>         line so it could be tried both ways.   Normally one floats a
>>>         shield if ground return currents from some
>>>         external path are expected.   I don't see where those would
>>>         come from in this situation.
>>>         -- Tom, N5EG
>>>         On Sat, May 30, 2020 at 8:31 AM Julius Madey via TangerineSDR
>>>         <tangerinesdr at lists.tapr.org
>>>         <mailto:tangerinesdr at lists.tapr.org>> wrote:
>>>             Dave,
>>>             I hadn't either until I saw the note from Phil. I've been
>>>             using your original small adapter board with the Quicc
>>>             connectors so haven't run into the issue yet.
>>>             Two shielded RJ45 PC through hole jacks in my parts
>>>             drawer, made by BERG, have non ferrous shields.  Couldn't
>>>             find the BERG part on line but DigiKey carries several
>>>             candidates.  One is AMP 5558342-1, the data sheet for
>>>             which shows phosphor bronze as the shield material. 
>>>             Since the shields on the RJ jacks of the three SBC's I
>>>             have are all non-ferrous, it may be that for RJ45's,
>>>             non-ferrous construction is the usual.
>>>             With regard to the shielded RJ plugs on the extension
>>>             cable, since the current recommendation seems to be to
>>>             only tie the shield to ground at the 'local' end, it's
>>>             possible to remove the shield on the plug at the remote
>>>             end; just did that with one of mine.
>>>             Jules
>>>             On 5/30/2020 10:34 AM, David Witten wrote:
>>>>             Jules,
>>>>             This is something I had not thought about.
>>>>             It does appear that Belden does sell a line of Cat 5+
>>>>             connectors that have little or no metal in them beyond
>>>>             the contacts:
>>>>             REVConnect® Connectivity System
>>>>             <https://www.belden.com/products/enterprise/copper/connectors/cat-6a-jacks>,
>>>>             but these appear to be components of a building wiring
>>>>             system, not board mountable jacks.  Further research is
>>>>             required.
>>>>             Dave
>>>             -- 
>>>             TangerineSDR mailing list
>>>             TangerineSDR at lists.tapr.org
>>>             <mailto:TangerineSDR at lists.tapr.org>
>>>             http://lists.tapr.org/mailman/listinfo/tangerinesdr_lists.tapr.org

More information about the TangerineSDR mailing list