Topics

Stacking two FT50-2 cores for T1 in 80m QCX? #t1 #qcx #80m

Serge, ON4AA
 

Hi, I intend on buying and building an 80m version QCX before, in 11 months time, import taxes are possibly raised on importing goods from the UK into the EU. Anyhow, lurking around, I noticed the issue with winding T1 for 80 and 60m.

This made me wonder if winding T1 could be simplified by stacking (and glueing) two T50-2 cores?

Doing so would double the cross sectional area of the core. T1 would therefore only require half the number of turns for each winding.

Has this ever been considered, Hans? Would this work?

Hans Summers
 

Hi ON4AA

QRP Labs kits are not sent from UK so Brexit has no relevance. They are sent from Turkey. In any event, import taxes are not normally required. 

There is not really any issue with 60m or 80m T1. It requires more care but it isn't a problem. There's also this suggested modified way of winding it, which makes winding easier: http://www.qrp-labs.com/qcx/qcxmods.html#80m

If you wish to use two toroids and adjust the number of turns then that should work as far as I know, though I have not tried it. I am not convinced that the number of turns would be exactly half. 

73 Hans G0UPL
http://qrp-labs.com

On Sun, Feb 2, 2020 at 10:03 AM on4aa <serge@...> wrote:

Hi, I intend on buying and building an 80m version QCX before, in 11 months time, import taxes are possibly raised on importing goods from the UK into the EU. Anyhow, lurking around, I noticed the issue with winding T1 for 80 and 60m.

This made me wonder if winding T1 could be simplified by stacking (and glueing) two T50-2 cores?

Doing so would double the cross sectional area of the core. T1 would therefore only require half the number of turns for each winding.

Has this ever been considered, Hans? Would this work?

Serge, ON4AA
 

Dear Hans,

In retrospect, I agree, the word "issue" may have sounded a bit too harsh. However, I did refer to exactly that same modification with my link. Then, there is also OE6FEG's mod of using thinner 0.15mm wire for winding the 80m T1. Finally, in all fairness, to quote the manual:

Note for 80m and 60m versions: the toroid ring is not large enough to neatly hold all of those turns in a single flat tidy winding. The large secondary winding WILL end up looking messy, with overlaps in some places. You should try to ensure any overlaps turns are evenly spaced throughout the winding! Do not try to wind one neat layer, then wind the remaining turns as an additional neat layer on top of that. Just go with the messy overlapping turns and don’t worry about it: everything will work fine, regardless.

There is a suggested modified way of winding the turns for 80 m version, which may make it easier; see http://www.qrp-labs.com/qcx/qcxmods.html#80m

So, perhaps it is safe to say that quite a few people are daunted by the prospect of winding T1 for either 80 or 60 m. This may or may not be because about 25% of the world population is affected by presbyopia.

Anyhow, all I wanted to do, is to suggest this modification of stacking two F50-2 cores in the hope that this may render this kit even easier to construct on 80 and 60 m. Incidentally, it may also improve the Q factor of T1. This is certainly something I would like to try out. From http://toroids.info/T50-2.php, I gather that one should be aiming at 22.66 µH at 3.550 MHz. I agree, the required turn count might not exactly be half (34), but I could initially add a few turns more and then test it with a calibrated VNA before mounting. The shorter windings appear to be less critical, but perhaps you can provide a design specification for those in this context? When I come around it, I will report back here about how it went.


As for the location and the absence of import taxes, that information deserves prominent mentioning on your website and webshop! Mind you, that I discovered your website about half a year ago. Since all prices on the site (not the shop) are mentioned in USD, I was living with the false impression that QRPLabs were a US enterprise. In fear of excessive EU import taxes and customs handling costs, I did not consider placing any order at that time. It is only very recently that it dawned on me that you are actually living in the UK, hence my earlier comment. I am very happy to learn that we will remain spared from import taxes in the future.

Hans Summers
 

Hi Serge 

In retrospect, I agree, the word "issue" may have sounded a bit too harsh. However, I did refer to exactly that same modification with my link.

Ah Ok yes, I missed the link... 

Then, there is also OE6FEG's mod of using thinner 0.15mm wire for winding the 80m T1. Finally, in all fairness, to quote the manual:

Note for 80m and 60m versions: the toroid ring is not large enough to neatly hold all of those turns in a single flat tidy winding. The large secondary winding WILL end up looking messy, with overlaps in some places. You should try to ensure any overlaps turns are evenly spaced throughout the winding! Do not try to wind one neat layer, then wind the remaining turns as an additional neat layer on top of that. Just go with the messy overlapping turns and don’t worry about it: everything will work fine, regardless.

There is a suggested modified way of winding the turns for 80 m version, which may make it easier; see http://www.qrp-labs.com/qcx/qcxmods.html#80m

So, perhaps it is safe to say that quite a few people are daunted by the prospect of winding T1 for either 80 or 60 m. This may or may not be because about 25% of the world population is affected by presbyopia.

People with Presbyopia (including me) have some optical aids to help. They will need that anyway generally for soldering the rest of it probably. Generally yes, T1 is the hardest part of the QCX assembly, but it should not really be considered daunting. As long as the steps in the manual are followed carefully and methodically, the result is success.  

Anyhow, all I wanted to do, is to suggest this modification of stacking two F50-2 cores in the hope that this may render this kit even easier to construct on 80 and 60 m. Incidentally, it may also improve the Q factor of T1. This is certainly something I would like to try out. From http://toroids.info/T50-2.php, I gather that one should be aiming at 22.66 µH at 3.550 MHz. I agree, the required turn count might not exactly be half (34), but I could initially add a few turns more and then test it with a calibrated VNA before mounting. The shorter windings appear to be less critical, but perhaps you can provide a design specification for those in this context? When I come around it, I will report back here about how it went.

I'd half the number of turns for the shorter windings, it is not critical and half should be a good enough approximation. I don't see any harm in stacking T50-2 cores but I would not consider it necessary, or bother with it, myself.  

As for the location and the absence of import taxes, that information deserves prominent mentioning on your website and webshop!

This information is in the FAQ http://qrp-labs.com/faq#where 

Mind you, that I discovered your website about half a year ago. Since all prices on the site (not the shop) are mentioned in USD, I was living with the false impression that QRPLabs were a US enterprise. In fear of excessive EU import taxes and customs handling costs, I did not consider placing any order at that time.

You can also price in euros or pounds, there is an icon at the top of the shop webpage in the middle of the screen, which will allow you to show all prices in euros or pounds if you wish, and also complete your payment in euros or pounds. 

It is only very recently that it dawned on me that you are actually living in the UK, hence my earlier comment. I am very happy to learn that we will remain spared from import taxes in the future.

I don't live in UK since 2011. Anyway don't worry about it :-D 

73 Hans G0UPL
http://qrp-labs.com

Serge, ON4AA
 

One final question, Hans:

f_res = (2 * pi * sqrt (L * C) )^1 = (2 *pi * sqrt(22.66E-6 * 91E-12))^-1 = 3.505 MHz

Is that the correct resonant frequency I should aim for with T1 on 80 m?

Or should it be a bit lower because of the winding capacitance?


I don't live in UK since 2011. Anyway don't worry about it :-D

You turned this thread in a (Dover) cliff hanger, Hans ;-)

Arv Evans
 

Serge

Hans and family live in Turkey.  QRP-Labs kits are designed, packaged, and sent from
that country.  Due to the relatively low cost most country customs will not impose a tax
on his kits (there are a few notable exceptions). 

Arv
_._


On Sun, Feb 2, 2020 at 5:02 AM Serge, ON4AA <serge@...> wrote:

Dear Hans,

In retrospect, I agree, the word "issue" may have sounded a bit too harsh. However, I did refer to exactly that same modification with my link. Then, there is also OE6FEG's mod of using thinner 0.15mm wire for winding the 80m T1. Finally, in all fairness, to quote the manual:

Note for 80m and 60m versions: the toroid ring is not large enough to neatly hold all of those turns in a single flat tidy winding. The large secondary winding WILL end up looking messy, with overlaps in some places. You should try to ensure any overlaps turns are evenly spaced throughout the winding! Do not try to wind one neat layer, then wind the remaining turns as an additional neat layer on top of that. Just go with the messy overlapping turns and don’t worry about it: everything will work fine, regardless.

There is a suggested modified way of winding the turns for 80 m version, which may make it easier; see http://www.qrp-labs.com/qcx/qcxmods.html#80m

So, perhaps it is safe to say that quite a few people are daunted by the prospect of winding T1 for either 80 or 60 m. This may or may not be because about 25% of the world population is affected by presbyopia.

Anyhow, all I wanted to do, is to suggest this modification of stacking two F50-2 cores in the hope that this may render this kit even easier to construct on 80 and 60 m. Incidentally, it may also improve the Q factor of T1. This is certainly something I would like to try out. From http://toroids.info/T50-2.php, I gather that one should be aiming at 22.66 µH at 3.550 MHz. I agree, the required turn count might not exactly be half (34), but I could initially add a few turns more and then test it with a calibrated VNA before mounting. The shorter windings appear to be less critical, but perhaps you can provide a design specification for those in this context? When I come around it, I will report back here about how it went.


As for the location and the absence of import taxes, that information deserves prominent mentioning on your website and webshop! Mind you, that I discovered your website about half a year ago. Since all prices on the site (not the shop) are mentioned in USD, I was living with the false impression that QRPLabs were a US enterprise. In fear of excessive EU import taxes and customs handling costs, I did not consider placing any order at that time. It is only very recently that it dawned on me that you are actually living in the UK, hence my earlier comment. I am very happy to learn that we will remain spared from import taxes in the future.

ajparent1/KB1GMX
 

Serge,

Stacking two cores increase the effective area but will not halve the turns due to 
lower effective mu of the total.  your would have to wind and measure with the
goal of about 23uh.

An alternate is to try a ft50-61 core with 18-19 turns as its easier to wind
(fewer turns) as it has a higher mu.

Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO

Serge, ON4AA
 

Dear Allison,

OK for the lower effective µ.

However, I fear I have to differ about the use of ferrite material instead of iron powder for this particular application.

An alternate is to try a ft50-61 core with 18-19 turns as its easier to wind (fewer turns) as it has a higher mu.

Narrow band tuned circuits achieve higher Q's with iron powder, as explained here.

Serge, ON4AA
 

The reason I am asking about the resonant frequency is because T1 wound as specified in the manual will have a substantial self-capacitance.

There is less ambiguity when specifying a tuned circuit by its resonant frequency instead of its inductance (23 µH). That is why I am asking Hans.

ajparent1/KB1GMX
 

Serge,

At 3.5mhz the loss to large amounts of wire approaches the loss using 61 ferrite
and in addition the network in this case is not narrow band. 

If Q were the critical function at 3.5mhz (or lower) you would best use spider
wound air cored inductor wound with Litz wire.  Bulky, but the Q is far higher
than ferrite or powered iron.

Trying to get higher Q from a single tuned circuit is not beneficial in this case.
Its one of those a little one way or the other and not at all that clear.  

Making it easier to wind is a consideration and reasonable trade.
I tried it and you would need test gear to discern the difference.

FYI: mine uses FT50-43 core with no large tuning coil (just the three
5 turn windings) as I tend to want to wander around multiple bands.
 I have the LPF from protection again higher band interference and
added a simple high pass with cutoff of 1.8mhz.  Performance is hard
to tell from stock without test gear.

Allison
-------------------------------
Please reply on list so we can share.
No direct email, it goes to bit bucket due address harvesting in groups.IO

Serge, ON4AA
 

To answer my own last question:

7.7 Band Pass receiver input filter characteristics

The following charts show measurements of the simple band pass receiver input filter implemented using transformer T1. For each band, one image shows the close-in response, the other shows the response across 0 to 30MHz. Note that in the circuit, the Band Pass filter sits behind the Low Pass filter. Therefore, the low pass filter response is ADDED to the band pass filter response. These charts show a relatively poor attenuation above the passband – but this is not a problem because in this region the Low Pass Filter provides high attenuation. For each band, the 3dB bandwidth was measured. The tuning range available was also measured (from having the C1 trimmer capacitor plates fully meshed, to having them fully un-meshed). Also shown is the attenuation at the centre frequency.

80m band

3dB bandwidth: 343kHz

Insertion loss: 1.10dB at 3.520MHz centre

Tuning range: 3.320MHz to 4.03MHz

T1 80 m frequency response

Now, I know what I need to do.