Ensemble_Isolation_Transformers LPF & Inductors

Below:- Rectifying low output on 20m for the 40/34/20m RXTX. And the 30/20/17 version.
               RX2, poor sensitivity on 10m https://groups.yahoo.com/neo/groups/softrock40/conversations/messages/79302


January 2014:- All Softrocks supplied from this date have modified boards, no modification required. But check the dates in case you have an older kit.

This page has been checked by Tony, KB9YIG, the Softrock kit producer and supplier. 

Gerhard PA1GVE discovered a mistake in the PCB layout of the Ensemble range. This is his post starting the thread. http://groups.yahoo.com/group/softrock40/message/66538 

After much discussion the following has been determined.

If the local oscillator isolation transformer is not connected properly some of the Si570 fundamental and harmonics will be radiated from the USB and other cables.

If the antenna isolation transformer is not connected properly then the outer of the antenna coax will be prone to common mode currents causing noise pickup on receive and local emissions on transmit.

The principle is that the grounded side of the isolation transformers should go to the same end of the primary and secondary. The two wires that come from the same hole in the binocular core.

This only affects the Ensemble. In all cases the schematics are correct. July 2012 Tony intends to alter the PCB when practical, at this time all Ensembles need altering.

For a good explanation of what we are trying to achieve read here, "pin 1" problems http://audiosystemsgroup.com/RFI-Ham.pdf

It is recommended that all Ensembles are modified.
January 2014:- All supplied from this date have modified boards, do not cross any wires.  Check the dates below, if in doubt ask Tony or the Softrock group.

Ensemble RX and Ensemble RXII:- T1 and T2 need one winding crossed. Date 6/15/10 printed on the board. Corrected from board date 6/20/13.
24 August 2013:- The T1/T2 layout issue has now been corrected with the circuit boards to be supplied in the RX Ensemble II kit from this date forward.  The date on the new board is 6/20/13.

Ensemble RXTX T6 needs one winding crossed.
2nd January 2014:- "A recent change to the RXTX Ensemble board has corrected the connections at T6.  It is no longer appropriate to cross either the primary lead pair or secondary lead pair of T6 to have the antenna return connectionand the circuit ground connection of T6 coming out of the same hole of the T6 binocular core.

A new RXTX circuit board will be included in each RXTX Ensemble kit as RXTX Ensemble kit orders are filled starting today.  The date printed on the new RXTX Ensemble circuit board is 12/8/13.
Tony KB9YIG"

VHF Ensemble T1 needs one winding crossed. Corrected from board date 9/8/12.
November 12 2012. "The board layout error on the 6m/4m/2m RX Ensemble board where either the primary or secondary leads of T1 needed to be crossed for best performance has now been corrected with the latest 6m/4m/2m RX Ensemble circuit board.  The new circuit board, now supplied in 6m/4m/2m RX Ensemble kits, has the date 9/8/12 in the silkscreen printing on the top surface of the board.

This shows a transformer with one winding crossed.
This was done on an installed transformer. After removing both wires (actually the primary) I soldered them from the top.
Either winding may be crossed.
So for an Ensemble that is already assembled the one that looks easiest to change may be altered.

Further Explanations.
Examination of the Ensemble receiver PCBs  shows the grounding point of the two windings of the LO isolating transformer, T1, are diagonally apart, not adjacent. This also applies to the RXTX antenna transformer, T6. 
On the Ensemble RX antenna input transformer, T2, the primary ground is the shell of the plug. The schematic shows the secondary is grounded to RF by C37. So one of the windings needs to be fitted crossed.
The RXTX LO transformer,T1, is correct. 

 Here the two secondary wires have been removed. Replace them so they are now crossed as shown. Do not touch the other (primary) winding. Either the primary or the secondary may be changed, whichever is the easiest for your situation.

Note the "ANT GND" option should normally be left disconnected. A link here defeats the purpose of using an isolating transformer.


Paul, 9H1FQ, persevered with this problem that has affected some of the Ensembles built to this configuration and found the LPF to blame.
Tony has altered the design slightly. It was found the HF cutoff was too low. See Robby's pages  http://www.wb5rvz.org/ensemble_rxtx/ and set the band option. The revision is on RF/IO.

My tests seem to put the blame on the T37-6 cores, I measured an AL_Value of 3.8 compared with the 3.0 specified.
The original design seems correct but the tolerance of the core material has caused the inductors to be too high. I measured 1uH rather than the specified 0.8uH.
BUT note that the calculator assumes 100% coverage of the core. A higher inductance will be produced if a gap is left as the assembly instructions describe.
To correct for this on assembled Ensembles I suggest just reducing the number of turns on L2 and L3. Try removing one or two turns and check that one watt output is easily achieved on 20m. 
 This should make the LPF conform to the original design.

I am wondering if this might affect other LPFs. I found a 30/20/17 metre build had significantly lower output on 17m,  removing 2 turns from L2 and L3 greatly improved things.
I found that the inductance was originally greater than specified presumably the core material has a high AL value.

Note that the available output is greatest on the low frequencies and reduces on the higher bands.
A 160m Ensemble may well give over 2 watts but a 10m Ensemble may struggle to make 1 watt.


My Ensembles have TX crossed, RX straight. I confirm that this works with all versions of PSDR and Rocky.  CHECKED!!!
But it is always the responsibility of the user to check operation, understanding what effect these  links have.



https://groups.yahoo.com/neo/groups/softrock40/conversations/topics/69627 See source    *******************
Not carried through:-

----- Original Message -----
Subject: Re: [softrock40] Peaberry using PowerSDR2.4.4 I/Q Reverse on TX only

> The following two paragraphs from from
> http://www.wb5rvz.com/sdr/ensemble/07_opamp.htm
> "In both the RX op amp and TX QSE sections, the jumper links on the
> I/Q input and outputs should ideally be fitted crossed over....


Unfortunately in a list like this confusion reigns!
We get posts from those that understand and from those that think they understand.
At the time Robby wrote that, when the Ensemble was introduced, there was a LOT of confusion, some were convinced their soundcard
L/R was reversed from the normal standard. Perhaps it needs re-writing.

My Ensembles have TX crossed, RX straight.
I confirm that this works with all versions of PSDR and Rocky. I've just looked at Rocky, it is set Left/Right = Q/I which I presume
is the way it works. I have used PSK TX but not recently......
Right, seeing other comments.
Rocky 3.6 Correct.
Rocky 3.7 Correct.
PSDR has always been correct for all versions, from 1.9 to 2.4.
Correct means that it TXs on the same frequency as RX.

> provides compatibility with all versions of PowerSDR. For Rocky and
> Winrad it is necessary to use the 'swap IQ' functions in their set up.
> If you jumper them straight across fine for Rocky and Winrad, but you
> will not be able to use any version of PSDR. (Author wired for Rocky)"

In this configuration RX/TX will not be correct for PSDR or any software except those that have separate RX and TX switches.

> "Some builders (author included) have encountered issues in this setup
> and siscovered that the only the TX leads should be crossed, leaving
> the RX leads uncrossed. Be prepared to change these should you
> encounter this issue."

Robby got it right in the end. But some were not convinced.

> I own a SoftRock RXTX Ensemble and wiring them as per the first
> paragraph, with both crossed, left me in a situation where Rocky was
> impossible to set up.

Making the change in the second paragraph allows
> use on Rocky, but PowerSDR was impossible.

TX crossed? Now this I do not understand. This is the correct way.

> I have tested many different applications with the Peaberry and would
> love a design that didn't require i/Q reversal. Unfortunately, this
> doesn't seem possible when two very popular applications appear in
> direct conflict with each other and are inflexible with their
> configuration.

David, if this is really true then surely there would have been something said in the 5 years I have been with this group?
I repeat, once the IQ is setup for PSDR then just the software switches on ALL other software is all that is required.

> To everyone who transmits with both PowerSDR and Rocky on a RXTX
> Ensemble: What versions are you using? How are your jumpers set up?

Let's hope we get some other input.....
Which seems just as confusing! I just do not understand how any extra crossovers are needed.

73 Alan G4ZFQ


Radiation from the RX Ensemble II

https://groups.yahoo.com/neo/groups/softrock40/conversations/topics/69627 See above