Topics

Harmonics and Relay Replacement

ajparent1/kb1gmx
 

What I'm seeing yet another wish for a single pill fix...

Using better relays even omron G6 series helps... to a point.  That point occurs when the board
layout under the relays (the real cause of issues) takes over.  For all that haven't looked there is
a wiki with pictures.

I've made those filters work with the existing relays and a simple re-wire.

Board layout is a extremely significant contributor to many of the issues.

Fire up your web browse and look here:
https://groups.io/g/BITX20/wiki/uBITX-spur-and-harmonics

Allison


iz oos
 

Welcome back Allison, I have been convinced too by these phony people to simply replace the relays. I suspect some of them work for either Axicom or the Farnell :)


Il 18/ott/2018 17:11, "ajparent1/KB1GMX" <kb1gmx@...> ha scritto:
What I'm seeing yet another wish for a single pill fix...

Using better relays even omron G6 series helps... to a point.  That point occurs when the board
layout under the relays (the real cause of issues) takes over.  For all that haven't looked there is
a wiki with pictures.

I've made those filters work with the existing relays and a simple re-wire.

Board layout is a extremely significant contributor to many of the issues.

Fire up your web browse and look here:
https://groups.io/g/BITX20/wiki/uBITX-spur-and-harmonics

Allison


Jerry Gaffke
 

These things are complicated, I certainly wouldn't say anybody's being phony.

There are multiple reports of Axicom relays mostly solving the issue of harmonics (except 40m?)
as measured by a spectrum analyzer.

The board layout looks problematic, it may prove necessary to at least replace
a few traces with coax while the relays are off the board.

There are a lot of things about the $120 uBitx that do not conform to standard practice
for an all band HF transceiver.  But we're very close to simple solutions that make the uBitx
meet regs.  No reason to disparage those contributing weeks of their available time in the attempt

The easily discouraged should hang back for another week or two till this all sorts out.
You didn't finish reading War and Peace yet, did you?

Jerry, KE7ER


On Thu, Oct 18, 2018 at 08:27 AM, iz oos wrote:
Welcome back Allison, I have been convinced too by these phony people to simply replace the relays.

Joe Puma
 

You nailed it with the rework. I probably wouldn’t go through that effort although I don’t mind, but different relays that can keep the issue at bay seems like a no brainer if you don’t want to roll the sleeves up or maybe don’t have the skill to do so. 

Joe



On Oct 18, 2018, at 11:11 AM, ajparent1/KB1GMX <kb1gmx@...> wrote:

What I'm seeing yet another wish for a single pill fix...

Using better relays even omron G6 series helps... to a point.  That point occurs when the board
layout under the relays (the real cause of issues) takes over.  For all that haven't looked there is
a wiki with pictures.

I've made those filters work with the existing relays and a simple re-wire.

Board layout is a extremely significant contributor to many of the issues.

Fire up your web browse and look here:
https://groups.io/g/BITX20/wiki/uBITX-spur-and-harmonics

Allison


RCBoatGuy
 

Given that the main problem is the layout, although better relays seem to help, has anyone tried shielding the relays from the PCB using copper tape between the board and relays, soldered to the relays ground pin?  I don't currently have a uBitX, or I'd give it a try. :(

You'd have to be careful to notch the copper so it doesn't short to any other relay pins, but it might help shield the relays...

73,

Carl,.K0MWC

ajparent1/kb1gmx
 

On Thu, Oct 18, 2018 at 08:27 AM, iz oos wrote:

I have been convinced too by these phony people to simply replace the relays. I suspect some of them work for either Axicom or the Farnell :)


No its legit.  By lowering the leakage past the relay itself the over all attenuation is better.
When I say better its in the form of just enough to help the issue and not a full solution.

The word often used in industry is a a bandaid (brand of small self stick bandage).

My experience and evaluation says if the layout were functional the old relays would have
done the job and this discussion would be moot.

Allison

ajparent1/kb1gmx
 

On Thu, Oct 18, 2018 at 09:07 AM, Jerry Gaffke wrote:
The board layout looks problematic, it may prove necessary to at least replace
a few traces with coax while the relays are off the board.

Jerry, 

Its more than looks, it is that way.  Whats in the wiki is verified work that does work.
Fix the layout and it works.  

More isolation in the relays helps but doesn't fix (harmonic on 40 for example) that
is because the blow by of the relays is not exceeded by the layout.

Hence my comment on the G6 relays... obviously the high cost  part should be
even better.  Except that layout cripples you too, its about as good as it gets..

As far as W&P, that took me a week of evenings, the Dune trilogy is a weekend read.

Allison


ajparent1/kb1gmx
 

On Thu, Oct 18, 2018 at 02:08 PM, RCBoatGuy wrote:
has anyone tried shielding the relays from the PCB using copper tape between the board and relays

Carl,

Yes back in early july.  Didn't help much at all maybe 2db at the high end of the band.

Look at the wiki pictures two of the longer lines have RF power on them and they all loop
back to KT1 and past all the others.  Not good.

Allison

Ashhar Farhan
 

I still havent finished Brothers Karamazov. Yes, I have read it cover to cover several times. Not yet finished.


On Fri, Oct 19, 2018 at 10:33 PM ajparent1/KB1GMX <kb1gmx@...> wrote:
On Thu, Oct 18, 2018 at 02:08 PM, RCBoatGuy wrote:
has anyone tried shielding the relays from the PCB using copper tape between the board and relays

Carl,

Yes back in early july.  Didn't help much at all maybe 2db at the high end of the band.

Look at the wiki pictures two of the longer lines have RF power on them and they all loop
back to KT1 and past all the others.  Not good.

Allison

Sascha Bohnet | DL5SMB
 

I would really like to try out Allisons solution with the complete rearrangement of the filters. However I am a big fan of "clean" solutions,
and all this wiring, cutting etc. looks really sloppy to me.

And I would really hate it, if it did not work out as it should and then i would have completly ruined my board.

Is maybe anyone good at pcb designing?

Then we could make a daughterboard, desolder the whole filter part on the main board and put everything on the daughterboard on top.
Maybe the placing of all the parts could remain the same on the additional board and only the traces would need some rearranging.

Gordon Gibby <ggibby@...>
 

​Well.......you can do all that, and I'm sure that Allison's solutions would work quite well!    


however, that is an AWFUL LOT of work, which is why I came up with a pcb daughterboard to just add right hand side relays and move just a FEW traces to wires heading outbound.   This had a very salutory effect, in almost every case dramatically reducing the harmonics.   


It is possible that there are other intermediate steps that might also solve the problem.   But Allison has already put a HUGE amount of effort into this and I doubt she is going to do any more ---- her suggestions for the current board would solve many of its probems;, and her contirubtions to a different company's power amplifier have provided a road map to a better power amplifier solution.    


So there are likely now several solutions that work and you can just take your pick of how much work you wish to do in order to improve things.   


Cheers,


gordon




From: BITX20@groups.io <BITX20@groups.io> on behalf of Sascha Bohnet | DL5SMB via Groups.Io <saschabohnet@...>
Sent: Friday, October 19, 2018 5:12 PM
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonics and Relay Replacement
 
I would really like to try out Allisons solution with the complete rearrangement of the filters. However I am a big fan of "clean" solutions,
and all this wiring, cutting etc. looks really sloppy to me.

And I would really hate it, if it did not work out as it should and then i would have completly ruined my board.

Is maybe anyone good at pcb designing?

Then we could make a daughterboard, desolder the whole filter part on the main board and put everything on the daughterboard on top.
Maybe the placing of all the parts could remain the same on the additional board and only the traces would need some rearranging.

Arv Evans
 

Problem, and solutions are not that simple.  Allison has added comments to the Wiki that
describe both problems and solutions.  Kees has made a daughter board available for the
relays and LPF components. 

Overdriving the audio apparently is a big contributer to some problems.  Stage gain
arrangement is another potential issue.  Board layout and component placement has
been identified as problematic.  There are ways to operate the uBITX that mostly keep
it within reasonable limits, but some apparently want to exceed those parameters.

Unfortunately those who have recently joined the group may not have followed the process
that has led up to present state of things. 

Arv
_._


On Fri, Oct 19, 2018 at 3:12 PM Sascha Bohnet | DL5SMB via Groups.Io <saschabohnet=googlemail.com@groups.io> wrote:
I would really like to try out Allisons solution with the complete rearrangement of the filters. However I am a big fan of "clean" solutions,
and all this wiring, cutting etc. looks really sloppy to me.

And I would really hate it, if it did not work out as it should and then i would have completly ruined my board.

Is maybe anyone good at pcb designing?

Then we could make a daughterboard, desolder the whole filter part on the main board and put everything on the daughterboard on top.
Maybe the placing of all the parts could remain the same on the additional board and only the traces would need some rearranging.

Dexter N Muir
 

...and having read it, have you watched it (Frank Herbert's)? and follow-on Children Of (also a trilogy)? Well worthwhile!  Lord Of The Rings did that to me, too: sit down, read it, get up. 40 hours or more straight: Total immersion!

Joe Puma
 

I’m surprised someone hasn’t fabricated a board that the relays and coils can be moved to and the board can be mounted on top of the ubitx with standoff’s and then wired up to Ubitx. 

Joe



On Oct 19, 2018, at 5:17 PM, Gordon Gibby <ggibby@...> wrote:

​Well.......you can do all that, and I'm sure that Allison's solutions would work quite well!    


however, that is an AWFUL LOT of work, which is why I came up with a pcb daughterboard to just add right hand side relays and move just a FEW traces to wires heading outbound.   This had a very salutory effect, in almost every case dramatically reducing the harmonics.   


It is possible that there are other intermediate steps that might also solve the problem.   But Allison has already put a HUGE amount of effort into this and I doubt she is going to do any more ---- her suggestions for the current board would solve many of its probems;, and her contirubtions to a different company's power amplifier have provided a road map to a better power amplifier solution.    


So there are likely now several solutions that work and you can just take your pick of how much work you wish to do in order to improve things.   


Cheers,


gordon




From: BITX20@groups.io <BITX20@groups.io> on behalf of Sascha Bohnet | DL5SMB via Groups.Io <saschabohnet@...>
Sent: Friday, October 19, 2018 5:12 PM
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonics and Relay Replacement
 
I would really like to try out Allisons solution with the complete rearrangement of the filters. However I am a big fan of "clean" solutions,
and all this wiring, cutting etc. looks really sloppy to me.

And I would really hate it, if it did not work out as it should and then i would have completly ruined my board.

Is maybe anyone good at pcb designing?

Then we could make a daughterboard, desolder the whole filter part on the main board and put everything on the daughterboard on top.
Maybe the placing of all the parts could remain the same on the additional board and only the traces would need some rearranging.

Joe Puma
 

Ahh so Kees has made a board. Is there a link to it? Are they for sale?

Joe



On Oct 19, 2018, at 5:45 PM, Arv Evans <arvid.evans@...> wrote:

Problem, and solutions are not that simple.  Allison has added comments to the Wiki that
describe both problems and solutions.  Kees has made a daughter board available for the
relays and LPF components. 

Overdriving the audio apparently is a big contributer to some problems.  Stage gain
arrangement is another potential issue.  Board layout and component placement has
been identified as problematic.  There are ways to operate the uBITX that mostly keep
it within reasonable limits, but some apparently want to exceed those parameters.

Unfortunately those who have recently joined the group may not have followed the process
that has led up to present state of things. 

Arv
_._


On Fri, Oct 19, 2018 at 3:12 PM Sascha Bohnet | DL5SMB via Groups.Io <saschabohnet=googlemail.com@groups.io> wrote:
I would really like to try out Allisons solution with the complete rearrangement of the filters. However I am a big fan of "clean" solutions,
and all this wiring, cutting etc. looks really sloppy to me.

And I would really hate it, if it did not work out as it should and then i would have completly ruined my board.

Is maybe anyone good at pcb designing?

Then we could make a daughterboard, desolder the whole filter part on the main board and put everything on the daughterboard on top.
Maybe the placing of all the parts could remain the same on the additional board and only the traces would need some rearranging.

 

I would really like to try out Allisons solution with the complete rearrangement of the filters. However I am a big fan of "clean" solutions,
and all this wiring, cutting etc. looks really sloppy to me.

And I would really hate it, if it did not work out as it should and then i would have completly ruined my board.
Ah Sascha,

Looks aren't everything eh?  You would be happier with the supermodel that has an awful personality?

The work need not look bad at all -- it's all up to how you do it and it gives one the chance to show their construction skills beyond the paint-by-numbers stage.  In the end it can be an accomplishment to be proud of!

And I'm not sure if the off-board solution works the same as Allison's plan?  Hers also reduces the number of relays engaged and that need switching so CW and PTT operation would be smoother.

73,


Mark.

PS:  Only read Crime and Punishment from Dostoevsky -- really dramatic and emotional.  Always thought the TV detective Colombo was modeled after the police detective in that book.  And Quiet Flows the Don was a very good read also. 

ajparent1/kb1gmx
 

Sascha,

The problem with yet another board is that there is no good way to couple the grounds between them.

As to clean solutions, just what is that?  You have a dirty board and the only clean solution is fix the board
as everything else is simply trying to make it work...maybe.

Allison

Donald <donwestpwl@...>
 

As a new member to this group I am offended. I have been reading posts, and reading posts, and reading posts. One problem is, to find the "good stuff", I have to read through posts like Arv's.


If the files are open to the public, I will quit this group.


Donald West, KB5PWL


On 10/19/2018 4:45 PM, Arv Evans wrote:
Problem, and solutions are not that simple.  Allison has added comments to the Wiki that
describe both problems and solutions.  Kees has made a daughter board available for the
relays and LPF components. 

Overdriving the audio apparently is a big contributer to some problems.  Stage gain
arrangement is another potential issue.  Board layout and component placement has
been identified as problematic.  There are ways to operate the uBITX that mostly keep
it within reasonable limits, but some apparently want to exceed those parameters.

Unfortunately those who have recently joined the group may not have followed the process
that has led up to present state of things. 

Arv
_._


On Fri, Oct 19, 2018 at 3:12 PM Sascha Bohnet | DL5SMB via Groups.Io <saschabohnet=googlemail.com@groups.io> wrote:
I would really like to try out Allisons solution with the complete rearrangement of the filters. However I am a big fan of "clean" solutions,
and all this wiring, cutting etc. looks really sloppy to me.

And I would really hate it, if it did not work out as it should and then i would have completly ruined my board.

Is maybe anyone good at pcb designing?

Then we could make a daughterboard, desolder the whole filter part on the main board and put everything on the daughterboard on top.
Maybe the placing of all the parts could remain the same on the additional board and only the traces would need some rearranging.

Joe Puma
 

Allison,

I haven’t looked at the wiki in a week or two. Do you have pics of the rig with the mods you made, removing traces, rewiring, etc. I just saw the board with the parts removed. 

Joe




On Oct 19, 2018, at 1:03 PM, ajparent1/KB1GMX <kb1gmx@...> wrote:

On Thu, Oct 18, 2018 at 02:08 PM, RCBoatGuy wrote:
has anyone tried shielding the relays from the PCB using copper tape between the board and relays

Carl,

Yes back in early july.  Didn't help much at all maybe 2db at the high end of the band.

Look at the wiki pictures two of the longer lines have RF power on them and they all loop
back to KT1 and past all the others.  Not good.

Allison

Bill Cromwell
 

Hi,

Nobody is going to see a "messy" looking solution if we keep the cover on :)

73,

Bill KU8H

On 10/19/2018 10:43 PM, ajparent1/KB1GMX wrote:
Sascha,

The problem with yet another board is that there is no good way to
couple the grounds between them.

As to clean solutions, just what is that? You have a dirty board and
the only clean solution is fix the board
as everything else is simply trying to make it work...maybe.

Allison
--
bark less - wag more