开云体育

with the ubitx, i have extensively used the EMRFD tools : the ladbuild, the gpla, the fba and the cascade. maybe i should do a video that explains how to use it.

- f

On Thu, Jun 22, 2017 at 1:33 PM, Trent Trent <vk7hrs@...> wrote:

Is that how you designed the bitx

On 22 Jun. 2017 3:13 pm, "Ashhar Farhan" <farhanbox@...> wrote:

Here is a screenshot of my simulation. i run it under ubuntu, hence the weird window style.

- f

On Thu, Jun 22, 2017 at 10:06 AM, Ashhar Farhan <farhanbox@...> wrote:

ben,

i simulated the amps using fba.exe that accompanies the EMRFD cd. you can download this and other goodies from . it is somewhere under the EMRFD link from the home page.?

i do remember that the match was very close to 50 input. but you could be right. the more complicating problem is that the operating frequency for this amp is vhf. the simple relationship of the resistor ratios doesnt hold up there.

i will build and measure the return loss at 50 ohms soon and let you know. thanks for pointing this out.

- f

On 22 Jun 2017 7:40 a.m., "Ben Bangerter, K0IKR via Groups.Io" <bwbangerter=yahoo.com@groups.io> wrote:

The design that Farhan has put forth for the ubitx transceiver project utilizes bi-directional termination-insensitive amplifiers (TIAs) based on work done by Wes Hayward, w7zoi, and Bob Kopski, k7nhi, and described in the white paper “A Termination Insensitive Amplifier for Bidirectional Transceivers,” which can be found on Wes’ web site: .? Feedback amplifiers of this type have stable and predictable gains, and most importantly, well-defined input and output impedances, which are most important for terminating filters and mixers.

I have been studying the ubitx schematic in some detail, and I conclude that the TIA shown has an input impedance of ~100 ohms.? Lets look at RXamp1, which employs Q10, Q11 and Q12, and is paired with what we may call TXamp2, with Q20, Q21, and Q22.? The input impedance as calculated by Farhan’s approximate relation is:

?Zin = ?R11xR13/R14 = 100 ohms,

And the impedance as calculated by Hayward’s more accurate relation (described in his white paper “The Feedback Amplifier with a Simple Model,” available on his web site) is:

?Zin = R13x(R14+R11)/(R14+R13) = 105 ohms.

When RXamp1 is paired with TXamp2 to create the bi-directional construct, in receive mode (when RXamp1 is powered and TXamp2 is not) the input to RXamp1 is in parallel with a load (for RF) of the series combination of R26+R27 = 517 ohms.? As a consequence, the output of mixer 1 is terminated in an impedance of 84 ohms (100||517) rather than the desired 50 ohms.? Actually, there is additional loading by R105 in the keying circuit, which lowers the impedance seen by the mixer to 77 ohms.? The output impedance of the bi-directional TIA in receive mode is 50 ohms in parallel with 708 ohms from the combination of R20, R21, and R24, or 47 ohms, which is the impedance presented to the input of the 45 MHz bandpass filter.

In transmit mode, my analysis shows that the 45 MHz bandpass filter is terminated in 84 ohms, and the impedance presented to the first mixer is 45 ohms.

The bi-directional TIA formed by RXamp2 and TXamp1 has the same departure from ideal behavior.? In receive mode, the second mixer is terminated in 84 ohms, and in transmit mode the 12 MHz crystal filter is terminated in 84 ohms.

Does all this have any significant effect on how well the ubitx performs?? Perhaps not.? I think the effect of non-ideal termination of an LC or crystal filter affects mainly passband ripple, and non-ideal termination of a mixer affects the amplitudes of mixing products and may increase the amplitudes of “birdies” in the receiver.

If I got “lost in the weeds” with my analysis, please set me straight!