mcHF V06 TX-RX antenna switching: Pin diodes vs Relay

See also related post.




This is the modification that I just made in my mcHF V06, inspired in df8oe RF-04/05/06-H-029  upgrade. In my case, I had to make the modification due to failure of pin diode D3.


The variant has been to connect a G6S-2 signal relay as a switch of the antenna signal, but taking care that capacitor C81 is earthed in TX mode. Else, in RX mode, the antenna is isolated from PA out. In this way, the shunt function of D4 pin diode is emulated. Protection by diodes 1N4148 -or similar- has always been discussed because the intermodulation effect that would cause an incoming strong signal exceeding 500 mV.Unlike df8oe, I used an NPN transistor to pilot the relay. Now switching transition between RX and TX is done the way that the RX band-pass is completely disconnected and grounded before any RF power from the PA can go into the LPF. This way no blowing up RX Bandpass circuit can ever happen independent of the switching speed of the relay or a complete failing Relay. 


I have not been able to check if the sensitivity of the receiver increase, as indicates Andreas, df8oe, since I do not have another pin diode MA4P7102F in good condition to do the check. In any way, in the original versión, when RX mode, the incoming signal has to cross diodes D3. The voltage at D4 is less than 0.3 volts, given by the voltage divider R53 - R54, else D4 is not polarized directly, although it is a strange situation. A part of the RF signal arriving from TX_PA_OUT fron antenna would not reach RX_ANT and would end up derived to ground through internal capacity of diode D4. At lower bands, RF2 and RF3 presents lower impedance, but there is an additional problem with D3, as we will see later.

MA4P7102F pin diodes have a reverse voltage of 200 V (absolute maximum value) according to their data sheet. In lower bands (40 ... 160 meters) at full power and SSB mode, maximum peak-to-peak voltage at the low-pass input should reach 70 V. That's why Cris, m0nka, suggests using 200-volt capacitors in the low-pass filters -and other parts- if full power is to be used.


Original TX-RX antenna switching wiht PIN diodes.

When  ANT_TX_ON is set to "1"(+5V) and mandatory ANT_RX_ON is set to "0" -both are CMOS logic compatible output signals-  D4 is direct biased and derives the possible TX strong signal through the small residual capacity of D3 to ground and protects BFR93A. But in receive mode is a little more complex.
Original TX-RX antenna switching wiht PIN diodes.
DC in red, RF in blue. Diode PIN D3 is reverse polarized RF, D4 shunt to ground.

DC in red, RF in blue. D3 is -poorly- biased, about 1 mA, but  RF from tx to antenna passes through.





First let's see how signals ANT_TX_ON and ANT_RX_ON are generated from signal PTT_3V.  Remenber: both are digital  CMOS compatible signals, generated in a quad NOR gate 74HCT02 (U7). In fact, 74HCT02 inputs are TTL compatible.


Characteristic curves of MA4P7102F PIN diodes.

The direct bias current of D3 around 1 mA causes an appreciable series resistance close to 10 Ω (left graphic curve), that would decrease the sensitivity of the receiver.  D4 receives a direct bias voltage of 0.26 volts. It is not enough to start direct bias driving, but it is not reverse bias either. D4 is in a strange region where everything can happen, including the possibility that it may be a source of noise, precisely at the beginning of the RX amplification chain.
The relationship between R53 and R54 is critical. R54 should be of lower value to better biasing D3 -near 20 mA serial resistance will be only 1 - but then would require a lower value for R53 too, but each NOR output of 74HCT02 can not supply more than 20 mA to the load (same as 10 TTL inputs), which limits the minimum value of R53. The problem is served. A signal relay is better optión.





df8oe option. 

C81 is not contected to ground when not in RX mode.



















ea8arx option.

C81 grounded on TX mode and RX isolation from T3.








Comments

  1. I enjoyed your analysis of the switching trouble, I have it also on 40m band above 2 watts. I am preparing to perform the relay modification. I'm a bit puzzled by your schematic diagram immediately above, I see relay pin 9 going to BPFs via C81, but what is happening from relay pin 4? It goes to LPFs, but originally PA and LPF is joined by R26 0 ohm resistor, I guess you removed this and connected either pad to relay pins 4 and 9?

    ReplyDelete
  2. Hi, Ian
    See this related post.

    https://ea8arx.blogspot.com/2020/04/mchf-v06-tx-rx-relay-antenna-switching.html

    ReplyDelete
  3. As I explained in the post, initially I had to do the modification due to the D3 PIN diode breaking. And I based it on the df8oe proposal, although I observed his proposal was incorrect and I improved it by isolating the secondary of the T3 transformer during reception (TX_PA_OUT from LPF) and grounding the RX_ANT signal during TX, all done with the same relay. Later I was able to verify a better linearity in the emitted SSB signal and an improvement in receiver sensitivity, when i did a second mcHF kit. My mistake was considering that the theoretical explanations were sufficient for everyone and not documenting everything with photographs. Fortunately Wolfram Perrey, PY2BND, implemented my proposal and took photos.

    ReplyDelete
  4. Hello, I do not understand, why someone could make modify PIN diode T/R switching to relay...?
    It is ruin the QSK ability as relay will never be enough fast and even worse it will start to clicking like old Singer sawing machine... :)

    ReplyDelete
    Replies
    1. Hello, Petr.
      The modification is not mandatory, in fact I had to do it because a PIN diode was damaged, everyone is free to do it. In my case and in that of other radio amateurs who did it as I implemented it, a better linearity of the SSB signal is obtained in transmission and an improvement in sensitivity in reception. The problem is caused by the PIN diodes in the mcHF being polarized from a TTL gate, with a voltage of 5 V you can barely get a current of 1 mA to directly polarize the PIN diodes. That is insufficient, the manufacturer recommends at least 20 mA, with which an internal equivalent resistance of 1 Ohm is obtained. Look at the characteristic curves of the diodes used. MY implementation differs from that of DF8OE in that I take advantage of the two relay circuits to separate the TX part from the RX part in antenna switching. And the relay noise is not heard even on CW. Look at the related post in the article.

      Delete
    2. Please see also:

      https://ea8arx.blogspot.com/2020/04/mchf-v06-tx-rx-relay-antenna-switching.html

      Delete
  5. This comment has been removed by a blog administrator.

    ReplyDelete

Post a Comment

Popular posts from this blog

MiniPA-70 and MINIPA-100 step by step

Modification in mcHF V06 to make accessible the jumper P6 on UI-board