Thank you David. Will keep these issues in mind.
Karin
_____
From: David Eckhardt [mailto:davearea51a@gmail.com]
Sent: Wednesday, October 28, 2020 12:57 PM
To: Karin Johnson
Cc: Don Kirk; Rfi List
Subject: Re: [RFI] Request For Comment (2nd Post)
Karin, it sounds like you know what to do and have the equipment to measure
results. Just a caution from EMC/RFI experience: Many active, SMPS's and
converters, do not tolerate complex loads or sources. The Corcom will
present a complex source impedance to the whole unit. Be sure all the
active 'converters' can handle that source impedance. Installation of the
filter may also 'confuse' the PFC, if it is active, as well. You commented
it is a simple inductor which may not be a problem, but might not be as
effective looking into a complex source. Mostly, just thinking out loud.
Dave - WØLEV
On Tue, Oct 27, 2020 at 11:28 PM Karin Johnson <karinann@tampabay.rr.com>
wrote:
I received a reply from Don Kirk, WD8DSB and Dave Eckhardt W0LEV.
First off a big thank you for both replies. I do have a commercial
Power line filter, which I did not mention in the Power Point documents
because I wasn't trying to propose solutions just yet. My main point
For the documents was to post my initial observations about the wiring
installation.
Now having said that. The power line filter I have is a Corcom model
16FC10. It has some pretty impressive attenuation data associated with
It. I have also confirmed the attenuation data, by actually measuring
It with my HP8753C vector network analyzer. The measurements were taken
In a 50 system using two DX engineering 1 to 1 baluns. I have calibrated
The system with the baluns so I can measure both common mode and
differential mode attenuation. It agrees very closely with the data sheet
For the Corcom 16FC10 line filter.
Now to the toroids shown in the documents. As far as I can tell these
toroids are not the typical ferrite toroids you might find in RFI
applications. They appear to be "TAPE WOUND" type toroids. I haven't found
a lot of information regarding using Tape Wound toroids for HF RFI/EMI. In
fact most of the information I have found only addresses the use of such
toroids with very low frequency (ie: less than 100 kHz) applications. I
don't know if I can get hold of a filter board outside of the unit installed
in the outside compressor, but if I can I plan to measure its
characteristics. Heck, even if I can't get another filter board, I could
plan on just shutting down the entire system, pull the safety
Interlocks for the 240V and remove the filter and toroids and measure what
is installed. I know that is going to be a bit drastic, and I am definitely
not looking forward to that task.
Now with respect to saturation of toroids. I have some partial good news
with respect to the AC current being drawn by the Power Inverter board and
the Copeland 3 phase compressor. I have an energy monitoring system on my
home made by a company in Cambridge, MA called SENSE. I have an individual
Current transformer on the dedicated supply line feeding the outside
compressor unit. The SENSE system tells me that the maximum current at 240
volts is only about 6.25 amps. The system has been running in cooling mode
only so far in Florida here. I haven't put the system into heating mode.
It is a Heat Pump. I suspect that in heating mode the current draw might be
higher but I don't know what specific value might be expected. This is
important because the Corcom power line filter I have has a current rating
of 16 amps at 240 Volt AC. Obviously if a ferrite type toroid were to be
also inserted into the 240 power source wiring, then I would have to know if
the ferrite toroid would have to be "doubled up" (ie: using multiple stacked
Cores). I do have a couple of Fair Rite snap on cores of type 31 mix. They
are the biggest that Fair Rite makes.
Current Probes
Yes I have a couple of Tektronix current probe kits that I can use to
actually put on the 240V supply line and the thermostat wiring. I also have
Access to a set of the HP CLOSE FIELD probes. Eventually I will get to
doing work necessary to probe with test equipment. The only reason I have
Yet to do this is several. 1 - Dragging my Tek7000 series scope outside is
a bit of a chore. 2 - It is still quite warm in Florida, and working outside
In the heat right now, is, well uncomfortable. I am in no particular rush
to get this solved immediately. I prefer to take a slow and measured
approach to solving these things. 3 - The installer contractor is going to
replace the unshielded thermostat wiring with shielded wiring at no cost to
me.
So I want to wait a bit to see how the shielded wiring affects things,
With regard to filtering every wire. Yes that should be the plan. I do
have a major addition planned with respect to adding another shielded
enclosure to the power and control wiring inputs to the compressor. It
would be installed and bonded to the chassis of the outside unit. I would
Remove and relocate the Rheem filter and toroids to the extra shielded
enclosure, plus add the Corcom filter to the 240V supply leads. If it turns
out that the thermostat wiring is a major contributor then either massive
filtering on the thermostat wiring or replace the wiring with a Fiber Optic
link for the RS-485 data on the Thermostat wiring. By the way the
thermostat wiring is not the typical five or eight conductor wiring, but is
only four wires. Two for 24VAC and Common coming from the inside Air
handler, and two for the balanced RS-485 data communicatins from the
Thermostat. It is a Rheem ECONET communicating type thermostat.
Well that is bit long, but I thin I've covered most of my proposed
solutions. Of course the respect to solving RFI/EMI issues nothing is cast
in stone and there may have to be a few more "goodies" added to the
filtering effort. This effort will take a bit of time as mentioned above
But I'm retired and have a lot of time to play with this.
Regards to all,
Karin Anne Johnson P.E. K3UU
Palm Harbor, FL
________________________________________
From: Don Kirk [mailto:wd8dsb@gmail.com]
Sent: Tuesday, October 27, 2020 11:24 AM
To: Karin Johnson
Subject: Re: [RFI] Request For Comment (2nd Post)
HI Karin,
If it were me I would spend a little money on a commercial power line filter
like the Delta Electronics 30DPGS5C and see if that puts a dent in the RFI
(if anything just for a simple test). It only costs $35.31 if purchased
from Mouser plus shipping. I sometimes have to use a commercial power line
filter in conjunction with a common mode choke per the K9YC design to
totally eliminate RFI from equipment that has inverters, etc. but sometimes
one or the other by itself is adequate. If the combination of them does not
kill the RFI I would then use a 1 or 3 inch single turn loop sense antenna
connected to a receiver to see where the signal is leaking out of the
enclosure (like along the thermostat line, etc.).
I understand you have extensive background in power electronics, but
expecting the manufacturer or installer to change things besides installing
a commercially available filter internally is asking a lot and very time
consuming.
I have run into numerous cases where power electronics like variable speed
drives specifically say they only meet EMI regulations when used with an
appropriate power line filter and they even list what filter should be used,
but system integrators rarely design in the appropriate power line filter
(if any), and this leads to the emission of RFI out into our environment. I
ran into this with an apartment complex kitchen ventilation fan variable
speed drive as well as with variable speed drives in multiple injection
molding machines located a great distance from my QTH. Simple Delta
Electronics filter like what's mentioned above except designed for 3 phase
solved the injection molding machine RFI and the power line filter specified
by the manufacturer of the apartment complex variable speed drive eliminated
that RFI. I understand these cases are not exactly what you are dealing
with since the power electronics is part of the total design in your system,
but using a dedicated commercially available power line filter is certainly
a reasonable approach even if just to prove a point to the manufacturer of
the RFI generating gear.
Just some thoughts from my end.
73,
Don (wd8dsb)
On Tue, Oct 27, 2020 at 9:52 AM Karin Johnson <karinann@tampabay.rr.com>
wrote:
Hello Group:
This is second try to post. First one never showed on list(sent on 10/21).
I've posted in the past about RFI from a new RHEEM A/C system. I am posting
two links to Google Drive, which are power point documents. One of them I
Posted previously, the RFI Spectrum Analyzer photos. The second document I
Just created with some pictures of the board layouts in the outside A/C
Compressor unit.
I would very much like to get some feedback from the list about the issues I
have regarding the wiring layout shown in the pictures. I know I can't
Encompass all of the possible issues shown in the pictures but I have tried
To list a few.
The main reason for posting is that I don't have any knowledgeable
individuals, or other ham folks who might understand some of the issues
Involved in controlling RFI/EMI, that live locally who I can communicate
with. This list has been helpful in the past And I hope will be helpful in
the course of solving my problem with this Unit. I am working now with the
installer company and hope to have better Cooperation with them.
Here are the two links to Google Drive:
1 - CompressorElectronics Link
https://drive.google.com/file/d/18zwBme-avmMDgkQabI9mOjPdprkhQ8aQ/view?usp=s
haring
2 - RFISpectrum Link
https://drive.google.com/file/d/1ku9NwkELhui2IsTUHYg3vfHuPsLvWwpC/view?usp=s
haring
Regards,
Karin Anne Johnson P.E. K3UU
Palm Harbor,FL
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
--
Dave - WØLEV
Just Let Darwin Work
_______________________________________________
RFI mailing list
RFI@contesting.com
http://lists.contesting.com/mailman/listinfo/rfi
|