¿ªÔÆÌåÓý

Building my own ubitx - question about mixers and mixer transformers


 

Hello (and thanks to the moderator for approving my request to join)!

I am implementing my own ubitx and have a couple of questions about the mixers and mixer transformers (which don't seem to be specified on the project website):

1. I could not find specification for the transformers used in ubitx (alongside the BAT45S diodes), but I did eventually track down a mention from bitx40 on hfsigs.com as follows:

T1,T2,T4,T5,T6,T7 : 10 turns trifilar on FT 37-43 core

Is this appropriate for the ubitx frequency regime (12 and 45 MHz)?

2.

There were some mentions of "tv transformer cores" - binocular - but no specification or identifier.
Are these a suitable alternative?

3.

Could I use packaged minicircuits mixers (such as ADE-1 or similar Level-7 mixer) in lieu of the transformer+bat45s design?
If so, how do I "unbalance" the ADE-1 for CW transmit in the 45 MHz-to-RF mixer stage? Would injecting a positive bias voltage into the IF port work (if so, recommended voltage, current and means to isolate normal RF flow)?

Thanks very much, looking forward to this project!!

73 Dave WB0GAZ wb0gaz@...


 

Dave WB0GAZ

In the early days of BITX most of us built our own transceivers that loosely followed the
original design by Farhan.? Historical information on those builds cover most of your
questions and can be found by browsing back over the group discussions.?

  • You can build your own mixers.? These might be simple as-built units with
    1N4148 diodes and FT37-43 transformers.? !0 turns may be a bit much
    for operation at the upper extremes of HF range, but you can test this by
    measuring input versus output of the transformers with an appropriate load.

  • Many of the commercial mixers can be used (SBL, etc.).? Almost all should
    perform adequately.? There is much room for discussion (argument...?)
    about whether commercial mixers or carefully built homemade mixers are
    best.? Lower threshold diodes should require less LO drive voltage.? Carefully
    matched diodes and transformer windings should also improve performance.

  • Binocular transformers do find use in the RF PA of may QRP rigs.? You can
    use stacks of conventional toroidal cores as work-alike alternatives.? Same
    winding technique is used for both binocular cores and stacked toroid cores.
There are many myths, opinions, facts, and opinionated facts regarding which mixer
is best or worse.? I personally would build my own carefully designed mixers, but leave
room for substituting alternative mixers for testing and evaluation purposes.? This
way you can determine for yourself which is best for you.? It may be that the best?
HF mixer will be different from the best BFO mixer.? It is probably not a one-size-fits-all
situation.

Arv
_._


On Wed, May 31, 2017 at 9:55 AM, David Feldman via Groups.Io <wb0gaz@...> wrote:
Hello (and thanks to the moderator for approving my request to join)!

I am implementing my own ubitx and have a couple of questions about the mixers and mixer transformers (which don't seem to be specified on the project website):

1. I could not find specification for the transformers used in ubitx (alongside the BAT45S diodes), but I did eventually track down a mention from bitx40 on as follows:

T1,T2,T4,T5,T6,T7 : 10 turns trifilar on FT 37-43 core

Is this appropriate for the ubitx frequency regime (12 and 45 MHz)?

2.

There were some mentions of "tv transformer cores" - binocular - but no specification or identifier.
Are these a suitable alternative?

3.

Could I use packaged minicircuits mixers (such as ADE-1 or similar Level-7 mixer) in lieu of the transformer+bat45s design?
If so, how do I "unbalance" the ADE-1 for CW transmit in the 45 MHz-to-RF mixer stage? Would injecting a positive bias voltage into the IF port work (if so, recommended voltage, current and means to isolate normal RF flow)?

Thanks very much, looking forward to this project!!

73 Dave WB0GAZ wb0gaz@...





 

Most any level 7 mixer should work, such as the ADE-1 or SBL-1. ?The commercial ones are designed to give a flat response up to a ghz or two, so an appropriately hand built design appropriate for the operating frequency may outperform it, though choosing the correct core and number of windings is not something I'll be giving advice on. ?If you check the datasheets, you will see that the commercial mixer schematic is identical to what Farhan has on his diode ring mixers, so the same method of transmitting CW should work. ?The dual diode modulator on the uBitx is operating at 12mhz, and can be identical to what is on the Bitx40.?

There are a few issues with the uBitx as it now stands.?

With a 45mhz IF and a high side VFO, the receiver will be susceptible to images in the FM broadcast band. ?For example, when receiving at 25mhz, the vfo will be at 45+25 = 70mhz. ? An FM transmitter at 70+25 = 95mhz would be received as an image. ?So if you have strong FM broadcast transmitters in your area, you might consider upping the IF frequency some from 45 mhz. ?That would also make it easier to build a suitable front end lowpass filter that still lets through a 10m signal.

The second harmonic out of the push-pull IRF510's is not sufficiently attenuated unless you happen to have a spectrum analyzer handy to get the balance right. ?So, for example, you will likely need separate lowpass filters on 80 and 40 meters to be legal. ?Use C0G or NP0 caps in those filters and anywhere the actual capacitance value is critical, and the caps in in the lowpass filters need to be good for several hundred volts.

There might be some trouble with crosstalk when using all three outputs from the Si5351, especially if the SI5351 outputs are set to the higher drive levels. ?Will certainly work, though might give us extra birdies.

Farhan is still developing the uBitx, but has encouraged scratch builds as a way of getting more information on possible issues.
I think it will prove to be a very nice rig, better in many ways than the Bitx40 has been.

Jerry, KE7ER



On Wed, May 31, 2017 at 08:55 am, David Feldman wrote:
1. I could not find specification for the transformers used in ubitx (alongside the BAT45S diodes), but I did eventually track down a mention from bitx40 on hfsigs.com as follows: ? ?T1,T2,T4,T5,T6,T7 : 10 turns trifilar on FT 37-43 core
Is this appropriate for the ubitx frequency regime (12 and 45 MHz)?

?


 

Thanks, Arv and Jerry, both very helpful.

My application is to use uBitx as IF for a UHF transverter (and this project is basically a learning exercise for me), so the transmit chain isn't going to be included (in fact, the third mixer 45>HF won't likely be used.)

As for synthesizer output crosstalk, I suspect this is somewhat PC layout dependent (I am doing a PC board, focusing on the two bidirectional amplifiers, the two mixers, and the crystal/LC filters at 12 and 45 MHz.)

Dave


 

Hi!?

Many old Bitxr's might remember instructions , pictures and excellent videos by Leonardo on the kit.



He has devoted a complete page on ferrite transformers and its turns for Bitx....





73

Rahul VU3WJM




On Wednesday, 31 May 2017 3:23 PM, David Feldman via Groups.Io <wb0gaz@...> wrote:


Thanks, Arv and Jerry, both very helpful.

My application is to use uBitx as IF for a UHF transverter (and this project is basically a learning exercise for me), so the transmit chain isn't going to be included (in fact, the third mixer 45>HF won't likely be used.)

As for synthesizer output crosstalk, I suspect this is somewhat PC layout dependent (I am doing a PC board, focusing on the two bidirectional amplifiers, the two mixers, and the crystal/LC filters at 12 and 45 MHz.)

Dave