Title: E/S Umschaltzeit bei CW und digital_mod
Post by: S53DZ on 17. August 2017, 10:52:17
Hi there,
There is always a question: Should I change mcHF T/R switch pin diodes and use a relay instead or not? I did it, however .....
I have checked, the used PIN diodes MA4P7102 are designed for 200V rev. voltage.
1. T/R with PIN diodes as from M0NKA:
- the D3 diode starts to conduct with the positive peak of a TX signal at a power of 1/4 W
- the Rx/Tx delay is --> 0ms
- the complete Rx is well protected by conducting D4
2. T/R with a relay
- great isolation
- low I.L.
- but larger delay at switch over (not defined and possible bouncing)
As Andreas some time ago already said, to keep the D3 reverse polarized, we would need much higher reverse voltage than +5V.
U_rev P_threshold
(V) (W)
-------------------------
+5 0.25
+12 1.4
+30 9
+48 23
The table is valid for a sin wave signal.
So, for those of you who would want to have a full-break-in functionality or a zero time delay from receive to transmit, there is a challenge to try to add a DC/DC up-converter (well shielded, of course) and then measure the eventual TX signal degradation.
I have not tested that but is should work to improve the T/R switch behavior at TX. DC/DC in attachment is only as an example from the web.
73 Bojan
|
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: DB4PLE on 17. August 2017, 11:07:13
Hi,
just to give some additional input:
G6K-2F-Y data sheet: Release/Close time: 3 ms max.
Durability
Mechnical: 50,000,000 operations min. (at 36,000 operations/hour)
Electrical: 100,000 operations min. (with a rated load at 1,800 operations/hour) Rated current is 2A @ 50 Ohm -> 200VPP -> 100WPEP
So even in worst case we can expect quite a lot of life from a relais and 5ms (3ms + safety margin are not too bad). I think we have more latency in the software than that 5ms. This does not say, pin diodes are not interesting, it just says, we can do a lot in software probably to get closer to an almost full break-in even with a relay.
73
Danilo
|
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: DF8OE on 17. August 2017, 11:07:30
Hi Bojan,
you have 100% and well understandable described the problematic which causes signal distortions if you use PIN diodes and do apply only 4V (the 5V are not reached under working conditions). Because of an external voltage multiplier again produces RF and is much more difficult to implement as a simple relay I have proposed the "relay solution". |
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: S53DZ on 17. August 2017, 12:00:20
Hi,
Let me say I am not advertising any of the two solutions, as I said, I am currently using a relay and it works fine for me. But this is only because I left the full-bk mode some 40 years ago, when I constructed it myself, hi.
I am just trying to make some facts here and be open to differences between the two options.
With the G6K data it should be added also the temperature factor. And the fact maybe that also with flash RAM write cycles we all are a bit conservative at using them ...
Additional to the table: I on purpose did not count the diode forward voltage where it fully conducts. So, that balances then out.
And thanks to both of you for considering it.
73 Bojan |
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: DB4PLE on 18. August 2017, 05:52:35
Hi Bojan, Andreas,
I think this topic is a highly and regularily debated topic. Why not put the facts (and some discussion based on them) in the Github-Wiki?
This way we can conserve the discussion a little better and easier to find. Compared to the forum we can keep the page much more condensed compare to more or less lengthy debates in Yahoo, OV I40 Forum or elsewhere. I think we are not in any kind of disagreement here when it comes to the two options and their pros and cons.
73
Danilo
|
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: S53DZ on 18. August 2017, 10:31:44
Well, this is a great idea, but in this case one should add some measurements data to it. To my understanding it can be quite a difference across the bands since the change of impedance influences the behavior great deal. For example, take just this brief calculation:
U_rev P_threshold Ptreshold
(V) (W) at 50 ohm (W) at 20 ohm
----------------------------------------------
+5 0.25 0.6
+12 1.4 3.6
+30 9 22.5
+48 23 58
73 Bojan
|
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: DF8OE on 18. August 2017, 11:09:45
Hi Bojan,
a WIKI lives form the contents all of us will put in it. So there is no need to create a complete page already containing all possible and neccessary tables / diagrams etc. Start with an idea and put some contents to the page. Others can (and will!!) add their measurements, schematics etc. We can learn from this, can correct rhings which are wrong etc. And we all can better discuss if there is something we can discuss about and discussions about things which are not possible will get rare - because of the reasons are already on the WIKI page....
So, please be kind, start a WIKI page about relay and diode switching @ GitHub...
EDIT:
For creating "nice" pages there are two possibilities:
1) RTFM. The way for getting bold, small, large characters, tables, lists etc. is documented.
2) If you want to start and not read manuals you can take a look at an already existing WIKI page which contains what you do want to add to your page. Switch to "edit mode" of this page and then you can see the syntax which will lead to your wanted part... Leave edit mode without saving and so you can "steal from the best" (the BEST way to learn...)
vy 73
Andreas |
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: S53DZ on 23. August 2017, 07:10:12
Danilo, Andreas,
ich habe unter "Known Issues" ein paar Sätze eingefügt.
(Obwohl ich das nicht ausprobiert habe. :-\)
73 Bojan |
Title: Re:E/S Umschaltzeit bei CW und digital_mod
Post by: DF8OE on 23. August 2017, 07:50:10
Hallo Bojan,
sehr schön! Genau so haben wir uns das vorgestellt :D
vy 73
Andreas |
Diskussions- und Newsboard des DARC-Ortsverbandes I40 | Powered by YaBB SE
© 2001-2003, YaBB SE Dev Team. All Rights Reserved.
|