Suggested Mod of T2 for Efficiency Improvement

DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Suggested Mod of T2 for Efficiency Improvement

Post by DL2MAN »

Hey guys,

One thing that annoyed me for a while now, was the heating up T2, which suggested that we´re losing energy:
BEFORE_Labelled.JPG
BEFORE_Labelled.JPG (55.59 KiB) Viewed 23534 times
In a Rig that is all about efficiency, that is obviously something unwanted.

The reason for the Loss is: While development I thought of it ONLY as a transformer, while it is also an inductor from RF to GND with very low Inductance. This means: Part of the RF is leaking to GND through T2 and at the same time core goes into saturation (additional core loss due to high magnetic flux) and we have some loss here. This is shown in the pic above as it clearly heats up.

The original Idea to solve this, was to change the Ratio of the transformers (Like 1:18) but this would have required a firmware change (beacuse of different FWD/RVS Voltages arriving on AtMega). Because of memory shortage we could not make it selectable, and since there will be probably a lot of people that don´t care, we would have needed to provide 2 different firmware Versions so the PWR/SWR/Efficiency measurement would still be accurate in all cases. This causes not only confusion for the customer, it also increases the amount of work for Guido, so we wanted to avoid that under all circumstances.

We could also not change the amount of turns on T1 and keep the ratio 1:7, as 2 or 3 turns primary would mess with output power. But I thought while keeping the same Ratio, I could increase the amount of turns on T2.... So instead of 1:7 Turns, I changed it to 3:21 Turns (same Ratio, more inductance, T1 unaffected).
This type of SWR Measurement we´re using, usually requires 2 equal Transformers, but it turns out, it also works with 2 different Transformers, with just the same Ratio.

Before:
T2_BEFORE.JPG
T2_BEFORE.JPG (138.86 KiB) Viewed 23534 times
AFTER
T2_AFTER.JPG
T2_AFTER.JPG (125.76 KiB) Viewed 23534 times
T2 Winding Detail: 3 Turns Primary : 21 Turns Secondary
I used 0,36mm Wire, as 0,4mm Wire would end up too tight:
T2_NEW.JPG
T2_NEW.JPG (91.37 KiB) Viewed 23534 times
And the result was roudabout 3% more Efficiency on all Bands, while Power and Efficiency Displayed vs REAL (external measurement) got closer together. This means: 3% more efficiency and more accurate Display readings.

And finally: The Thermal Shot after the T2 Mod:
Thermal_AFTER.JPG
Thermal_AFTER.JPG (52.21 KiB) Viewed 23534 times
T2 stays absolutely cool now !

Have fun !

73 Manuel; DL2MAN
va3rr
 

Posts: 54
Joined: 30 Dec 2021, 21:39
Location: FN25eg
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by va3rr »

Wonderful out-of-the box thinking! I would have suggested using a binocular core like a BN43-202, as it provides far more inductance with 7 turns than a FT37-43 does. But this allows us to continue to use our existing RF boards!

Well done!
Slava Ukraini!
g7jur
 

Posts: 35
Joined: 18 Feb 2022, 22:49

Re: Suggested Mod of T2 for Efficiency Improvement

Post by g7jur »

Hi.
I have done the T2 mod, but now showing 8 watts when outputting 5 watts.
Should I mod T1 as well ?.
Philip G7JUR
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL2MAN »

No you should absolutely not modify T1.
Please post a closeup picture of your Transformers.

73 Manuel DL2MAN
Iw2fpe
 

Posts: 12
Joined: 20 Mar 2022, 17:06

Re: Suggested Mod of T2 for Efficiency Improvement

Post by Iw2fpe »

Hi
I modified T2 as described, now I have a swr and power ratio faithful to the output rosmeter, better than the original T2 and more stable. The efficiency is still low (*Rxxxx* B2), I have to work on it, but the change was certainly effective. I used 0.4 wire, tight but it can be done.
Thanks!

73 IW2FPE
hb9hem
 

Posts: 36
Joined: 07 Jan 2022, 04:39

Re: Suggested Mod of T2 for Efficiency Improvement

Post by hb9hem »

DL2MAN wrote: 13 May 2022, 15:19 I used 0,36mm Wire, as 0,4mm Wire would end up too tight:
0.4mm wire is all I have around, so I tried. The efficiency is better, but the SWR into a 50R dummy load is between 1.07 (80m) and 1.02 (20m) now. Too much parasitic capacitance, I guess? In any case, 0.4mm is indeed too thick.

And one thing to watch out for (strange I never noticed) is the winding direction, clockwise vs. counterclockwise, so to speak. The footprint is not symmetrical.
DL1GKC
 

Posts: 6
Joined: 28 Apr 2022, 07:52

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL1GKC »

I still have T2 with its original 1:7 windings. I never got the ref/SWR display working correctly after finishing the build. I first want to sort this out before applying the mod.

On my rig, reflected power is always displaying 0 watts and SWR is 1:1.0 even if the real SWR is >1:3.0. PWR indication is working correctly. What could be the reason?
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL2MAN »

DL1GKC wrote: 22 May 2022, 05:30 (...)
On my rig, reflected power is always displaying 0 watts and SWR is 1:1.0 even if the real SWR is >1:3.0. PWR indication is working correctly. What could be the reason?
This clearly indicates a problem in the path of one of the 2 Transformers. Usually it´s cold solder joint on one of them (or missing contact because of enamel coating), it maybe one is wound in opposite direction as the other. (I did not think this was possible, as the Footprint allows only one direction without crossing leads) and in more unlikely cases: Bad solder joint around some factory assembled stuff.
Focus your Investigation to continuity in this section:
SWR.JPG
SWR.JPG (22.24 KiB) Viewed 22807 times
However: Since you´re going to re-solder it anyway, I would do the mod, and if you´re lucky it was T2.

73 Manuel; DL2MAN
sp9dlk
 

Posts: 1
Joined: 22 May 2022, 19:34

Re: Suggested Mod of T2 for Efficiency Improvement

Post by sp9dlk »

Manuel you used 0.36 winding wire. will the 0.31 wire be good or too thin? 73 sp9dlk
AntonR
 

Posts: 20
Joined: 23 May 2022, 10:44

Re: Suggested Mod of T2 for Efficiency Improvement

Post by AntonR »

Hi Manuel,

Where can we find 0,36mm winding wire?

Bye,Anton
User avatar
DH2LAB
 

Posts: 51
Joined: 08 Mar 2022, 01:39
Location: Kiel
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DH2LAB »

AntonR wrote: 23 May 2022, 10:51
Where can we find 0,36mm winding wire?


for example
https://www.ebay.de/itm/264039597632
73 de Matt, DH2LAB
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL2MAN »

sp9dlk wrote: 22 May 2022, 19:38 Manuel you used 0.36 winding wire. will the 0.31 wire be good or too thin? 73 sp9dlk
Yes, that can be used, too. You can also use 0,4mm wire, as someone pointed out. It is absolutely not critical. Use, whatever you have on hand and fits on there.

73 Manuel; DL2MAN
wa2t
 

Posts: 180
Joined: 01 Jan 2022, 16:30

Re: Suggested Mod of T2 for Efficiency Improvement

Post by wa2t »

Hi Manuel,

Just wondering, was 2T/14T not viable for some reason? Just thinking that this turns count would (more) easily fit the core using 0.4mm wire.

73,

Robert, WA2T
DL3MHT
 

Posts: 21
Joined: 14 Mar 2022, 09:46

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL3MHT »

Brilliant solution Manuel. Thanks! :idea:
I'll go for it, hoping it improves my ~75% average efficiency :)
DL1GKC
 

Posts: 6
Joined: 28 Apr 2022, 07:52

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL1GKC »

*Rxxxx* recently published a document dealing with the efficiency issue in their kits. They recommend a different approach for T1/T2, by using a thicker 0,7 - 0,9mm wire as primary that must be placed in the center. Has anybody tried this approach?

I have modified my T2 to the 3/21 winding (0,35mm wire) but I do not see an increase of the displayed efficiency (average is 64%). At least the reflected power / SWR measurements are now working accurately in my kit (*Rxxxx* B3).
wa2t
 

Posts: 180
Joined: 01 Jan 2022, 16:30

Re: Suggested Mod of T2 for Efficiency Improvement

Post by wa2t »

wa2t wrote: 24 May 2022, 13:50 Hi Manuel,

Just wondering, was 2T/14T not viable for some reason? Just thinking that this turns count would (more) easily fit the core using 0.4mm wire.

73,

Robert, WA2T
??
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL2MAN »

DL1GKC wrote: 04 Jun 2022, 16:29 *Rxxxx* recently published a document dealing with the efficiency issue in their kits. They recommend a different approach for T1/T2, by using a thicker 0,7 - 0,9mm wire as primary that must be placed in the center. Has anybody tried this approach?

I have modified my T2 to the 3/21 winding (0,35mm wire) but I do not see an increase of the displayed efficiency (average is 64%). At least the reflected power / SWR measurements are now working accurately in my kit (*Rxxxx* B3).
I have seen that *Rxxxx* Document, and I disagree with the value of it, as it´s mostly repeating my Instructions and adding some good practice, so it´s kind of pointless.
However, there´s some real disagreements from my side:
I disagree with Chapter 2 of Step 1: take special care for C11/C14/C19/C32 which have viases in their SMT pads and cover the pads with solder / make sure you realise proper solder joints. That is soldered by Factory, and should not be of your concern. What *Rxxxx* probably want to say is: Putting Vias in Pads is considered bad practice, and that´s true. However: I´ve been doing that for ease of routing and I have not seen a single Kit where this was creating a problem. Look at the QCX Kits: Hans does the very same thing for years now (over 10000 sold QCX) without issues.

I especially disagree with "Step 3" which is connecting 2 ground planes with a zero Ohms Resistor. Those
are already bridged through the other 3 Layers and plenty of VIAs. I don´t see the Point of this mod.

Sunnys Kits are above 80% Efficiency without Step 4, I never did it the way, it is being described in the *Rxxxx* Document. I always use the copper enamel wire, that is provided with the kit and make it as short as possible. Not through the center of the Toroid.

Another Issue of this *Rxxxx* Document is: It claims to have improved Efficiency, but the Tables with measurements to back that up, are still missing in Document Rev V1.0 (3rd run/4th run).

The provided Inductances can be misleading as they are most likely not measured at Target frequency per Band, but at much lower frequencies in the KHz Range, which is common with LCR Meters. That´s just to keep in mind, when you try to comprehend.

While I appreciate *Rxxxx* attempt to give the customer some guidance, I do not agree to all of their findings. This is not meant to be "bashing" or "ranting". It´s just my personal opinion based on technical reasons and experience.

My resumee of the *Rxxxx* Document is (ignoring the mentioned issues above):
If the Kit is factory assembled with the right parts (especially NP0/C0G Caps on RF Board), and you build it according to my Instructions (https://www.youtube.com/watch?v=VVGBapoCUls), and Tune the Filer Section according to my Video (https://www.youtube.com/watch?v=qN7hHsvcNMM), there should be no Issues to achieve 80% or more efficiency. So: If *Rxxxx* fixed the capacitor issue from the first batches, I´m positive you can achieve Efficiency if you tune it. Jumping through extra hoops not neccessary in my Opinion.

73 Manuel; DL2MAN
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL2MAN »

wa2t wrote: 24 May 2022, 13:50 Hi Manuel,

Just wondering, was 2T/14T not viable for some reason? Just thinking that this turns count would (more) easily fit the core using 0.4mm wire.

73,

Robert, WA2T
Hi Robert,

I thought of 2:14 Turns, but I wanted to cure the issue, not mitigate it.
There are claims, that 3:21 is possible with 0,4mm.
Just try it.

73 Manuel; DL2MAN
va3rr
 

Posts: 54
Joined: 30 Dec 2021, 21:39
Location: FN25eg
Contact:

Re: Suggested Mod of T2 for Efficiency Improvement

Post by va3rr »

wa2t wrote: 24 May 2022, 13:50 Just wondering, was 2T/14T not viable for some reason? Just thinking that this turns count would (more) easily fit the core using 0.4mm wire.
Robert - it looks like Owen Duffy did the maths on the 2T/14T and it checks out ok.

He also has other blog articles about the Tr(u)SDX

https://owenduffy.net/blog/?p=25315
Slava Ukraini!
wa2t
 

Posts: 180
Joined: 01 Jan 2022, 16:30

Re: Suggested Mod of T2 for Efficiency Improvement

Post by wa2t »

va3rr wrote: 10 Jun 2022, 13:43
wa2t wrote: 24 May 2022, 13:50 Just wondering, was 2T/14T not viable for some reason? Just thinking that this turns count would (more) easily fit the core using 0.4mm wire.
Robert - it looks like Owen Duffy did the maths on the 2T/14T and it checks out ok.

He also has other blog articles about the Tr(u)SDX

https://owenduffy.net/blog/?p=25315
Excellent, thank you!
DL3MHT
 

Posts: 21
Joined: 14 Mar 2022, 09:46

Re: Suggested Mod of T2 for Efficiency Improvement

Post by DL3MHT »

One more stupid question:
Winding directions for the two coils in order to avoid a 180° phase shift.

What is the correct way to do it?

Looking at one side of the ferrite are the two windings wound in the same direction or is one wound clockwise and the other counterclockwise?

In the first picture it looks like they have the same direction. In the second picture it looks like opposite winding directions...

Thank you,
Tobias
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: Suggested Mod of T2 for Efficiency Improvement

Post by Ohwenzelph »

The wires look like they are wound in the same direction in both, to me.
N4XTL
 

Posts: 24
Joined: 08 Jul 2022, 01:44

Re: Suggested Mod of T2 for Efficiency Improvement

Post by N4XTL »

I tried the T2 mod on one of my radios using 26 AWG magnet wire. Efficiency as measured on the (tr)uSDX actually went down about 1%. The toroid is very packed full of wire, so perhaps that is part of the reason. I have some 28 gauge magnet wire on order, but overall efficiency on that unit is very good, so I'm not too worried about it.

I have an FT37-75 toroid on hand. Any thoughts on building the 7:1 transformer with it instead of the 43 type? The calculated inductance with 7 turns on the 75 type would be 100 uH, versus 20 uH for the 43 type with 7 turns. Quite a nice improvement, and I could wind it using the normal wire and number of turns...

Thanks
N4XTL
 

Posts: 24
Joined: 08 Jul 2022, 01:44

Re: Suggested Mod of T2 for Efficiency Improvement

Post by N4XTL »

Answering my own question, after looking at specs, I believe the 75 type toroid would not provide sufficient performance at higher frequencies. The stock FT37-43 re-wound with 28 AWG wire seems to be the best option.
R2AZR
 

Posts: 1
Joined: 20 Aug 2022, 18:06

Re: Suggested Mod of T2 for Efficiency Improvement

Post by R2AZR »

I registered on the forum to ask a question about this, but a search fortunately yielded me this thread, which answered it. However now I've got a suggestion.

See, the situation is as follows: I bought the uptech aliexpress version of the (tr)uSDX, and the sheet of paper that they provide lists T2 as 3:21. However, Manuel's video still has it as 1:7, as does the schematic and iBOM. further, the way it's assembled (with just a jumper through the hole in the toroid) is rather uncommon, so I was confused whether I should do 1:7 or 3:21, and if the latter, how to do it. Fortunately, the post explains all of that perfectly well, but I think it would be beneficial to add a note to the build page mentioning the change (probably a couple of words and a link to this post would suffice), so that anyone who has a similar question like me would be led straight to the answer.

Thanks!
Post Reply

Users browsing this forum: No registered users and 3 guests