Date   
Re: Harmonic performance - SSB vs CW

ajparent1/KB1GMX <kb1gmx@...>
 

Nick,

Best layout is similar,  Bus down one site to the NO contacts
and the wiper is the end of the filter and when in the off poistion the
wiper sits on the NC contacts to ground shorting the filter.  So a filter pick 
enables the relay and filter and no pick the filter is shorted.  Reduces
the effects of stray capacitance.

That and copper pour under everything and opposing side. lots of vias to 
tie them together especially along signal tracks.

Allison

Re: Raduino Clone kit from W0EB-N5IB #ubitx

Jim Sheldon
 

Gentlemen would you mind taking this discussion to another thread?  You are cluttering up my offering of the Raduino clone to the group members.

Thanks,
Jim

Re: Harmonic performance - SSB vs CW

Nick VK4PP
 

Glen I hear you, I did look on eBay the omrons can be had for $1.36 each. i think 9 of them are still a reasonable cost given the LPF components can be salvaged from the original board... Also possible allows us to have 5 LPFs... This will requir moving to I2C lcd for extra IO ports...

Cheers.

Re: Harmonic performance - SSB vs CW

Glenn
 

Nick, I'm more inclined to stick with the current sized DIP relays for cost reasons if I did any more on this. Even though the Omrons are probably better relays, costs add up quickly.

glenn

Re: Harmonic performance - SSB vs CW

Glenn
 

Hi Nick,  it's just doodling right now.  I had also been thinking about doing a whole uBITX (with mods) in modular form also.  It could get expensive though re board making unless some smarts are applied to sizes and panelisation to hit the best cost/size ratio from current China suppliers.

It's also hard to form a 'final' schematic though. One also depends on the wisdom in this group for what is good and what is not so good, re mods. to take into a design.

glenn


On Tue, Aug 7, 2018 at 09:53 AM, Nick VK4PLN wrote:
Hi Glen.
How about using omron g6s-2, they are 5v and just need the reverse diode, I reckon then can be switched directly buy the NANO..
and about half the size...

You beat me to it, I was about to fire up EasyEDA and slap a board together..
I am still thinking of doing an entire uBitx in modular format... For this exact reason, easy to replace a section when a major fault is identified...

Great work!
73 Nick VK4PLN

Re: Raduino Clone kit from W0EB-N5IB #ubitx

Gary Anderson
 

Hi Glenn,
As far as I know, and have used, the physical layer (e.g. hardware) is defined as an open drain / open collector.  I haven't seen the Ap note.  If you have a link you could share, I would like to review.
For bit banging with a standard digital I/O port.  You set the port as an input, set the drive state to 0.  You bit bang between input and output states.

For the case we have here, between a 3.3V and 5V device, the pull-up to 3.3V Si5351 meets/ exceeds the VIH of the 5V Nano, so all is good.
However, if you need to interface between say a 1.2V or 1.8V to a  5V  I/O you would need to level shift......  I see how this Ap note could be going now.

I will admit, that I read you post under a fog of too many folks are destroying Nanos. I'd rather have a discussion then a fellow ham wondering why something went wrong when it worked fine yesterday.

Cheers,
Gary

Re: Harmonic performance - SSB vs CW

Nick VK4PP
 

Here is my take so far....

Re: Harmonic performance - SSB vs CW

Nick VK4PP
 

Hi Glen.
How about using omron g6s-2, they are 5v and just need the reverse diode, I reckon then can be switched directly buy the NANO..
and about half the size...

You beat me to it, I was about to fire up EasyEDA and slap a board together..
I am still thinking of doing an entire uBitx in modular format... For this exact reason, easy to replace a section when a major fault is identified...

Great work!
73 Nick VK4PLN

Re: Harmonic performance - SSB vs CW

Arv Evans
 

Kees

We did that back in the BITX20A and BITX17A days.  The cap in parallel with L2
can be used to make a steep notch at 2nd harmonic or at 3rd harmonic.  Testing
with LTSpice indicates that a cap across each inductor could provide the notch at
a pair of frequencies. 

Arv
_._


On Mon, Aug 6, 2018 at 4:03 PM Kees T <windy10605@...> wrote:
Since we're all using W3NQN's LPF design/layout from the 1999 QST (Hans, Farhan, me too). Typically 3 inductors, capacitive input/output, etc. The only difference I see is that my designs use a capacitor across L2 Unless someone has already looked into that, I'll  fire up "Elsie" and see what it shows.

73 Kees K5BCQ

Re: Harmonic performance - SSB vs CW

ajparent1/KB1GMX <kb1gmx@...>
 

Kees,

Mica is good but not at all cheap and somewhat scarce and big

For SMT MLC are good fairly cheap too (at the 5000 piece level).

For the rich ATC 100B or of the other series in that desired size
like 600C or F.

Also using T30 size is a bit small for 10W and up.

Allison

Re: Harmonic performance - SSB vs CW

Glenn
 

More to see if it's possible rather than anything else, I was working on a PCB that would fit in same space as the original LPF area, but switching on both ends of each filter. (ie 3 more relays)  I didn't alter the actual switching sequence just added the extra relays controlled in the same manner as the original ones. Picture is of the PCB. No switching transistors are on the PCB, yet.  You can see space either side though.

This is not all routed as you can see, just the RF paths in and out of the relays. (The inductors are shown with 4 holes, just because i  had that part in my LIB.) Orientations may change.

glenn vk3pe

Re: Humble suggestion future Ubitx versions #ubitx

RICHARD <k6kwq@...>
 

Who’s complaining?  Yes it is a kit, much like the 10/11 meter mods in the past “cut the green wire”.   It is a kit that anyone can buy,  and with a little help can operate 1-30 MHZ.   I’m an NRA member and see the same thing, “ buy this gun, move this screw” and it will do this.

But that’s just me, been a ham for 52 years seen a lot.

K6KWQ

 

Sent from Mail for Windows 10

 


From: BITX20@groups.io <BITX20@groups.io> on behalf of AA9GG <paul.aa9gg@...>
Sent: Monday, August 6, 2018 2:45:09 PM
To: BITX20@groups.io
Subject: Re: [BITX20] Humble suggestion future Ubitx versions #ubitx
 
We are hams...it's NOTHING to build a filter to put on it....It is sold as a "kit"...so FINISH building it and stop complaining....

On Mon, Aug 6, 2018 at 12:59 PM, RICHARD <k6kwq@...> wrote:

Last month QST did a write up on the BITX40, with all the published lab tests.  It would be interesting if they were do the same with the uBitx.    It might get FCC attention and stop all future imports to the USA,  is that what we want?  Maybe it is time to tone down all the attention to these problems.

Just a thought.

K6KWQ

 

Sent from Mail for Windows 10

 


From: BITX20@groups.io <BITX20@groups.io> on behalf of ajparent1/KB1GMX <kb1gmx@...>
Sent: Monday, August 6, 2018 9:30:20 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Humble suggestion future Ubitx versions #ubitx
 
Iz oos,

For some its was the FT817 for 109$.  There is a lot of distance and people complain about the FT817!

I also agree if you transmit, do so cleanly.  Its a kindness to friends around the world  The bitx40 is far cleaner.

As to many rigs... guilty.

For hf I have two old Tentec the 505 QRP 5 band and its bigger brother the 340 100W 5 band.
I have no worries depite the fact they were design in the 19070s that they meet current standards
or maybe even exceed them.

More modern a Ft817 and the Tentec Eagle.

That and KNQ7A, KD1JV Slopbucket20, WM20, and Diz's 1W, for the
kit radio world.  Then I have my first 20M SSB , 10M SSB, 15M SSB,
radios from the ground up.  All meet the numbers required and then some.
All have decent carrier suppression.  The only one with issues that needed
help was the KNQ7A that being a really loud pop on TX, easy fix too.

Others manage to do it well and inexpensively and with attention it works.
As is the ubitx has more of my time in it than a few scratch builds including
building their  filters and PLL systems(pre Si 5xx parts).

Mine however is in the "junkbox" as a result of slicing it up to get to root causes.
Its salvageable but a clean sheet would be less time consuming.

Allison




--
Paul Mateer, AA9GG
Elan Engineering Corp.
www.elanengr.com
NAQCC 3123, SKCC 4628

Re: Raduino Clone kit from W0EB-N5IB #ubitx

Glenn
 

Gary, my understanding is that the I2C bus is only open drain when reading back the "ACK"  9th bit on the data line. I may well be wrong here.  Bit banging implementations for I2C will not all be open drain either.
Philips publish an Ap note on using logic level translation on the I2C bus when different voltage levels are used. They use FET's and pullups as no doubt you have seen. Since they invented I2C I assume they know what they are talking about?   In any case, it costs pennies only and is in my view, good practice.
glenn



On Mon, Aug 6, 2018 at 11:13 PM, Gary Anderson wrote:
Glen, concerned about your logic level translator implementation w.r.t. I2C bus.  This is an open drain bus.  Nothing should be actively driven High.  The high state is handled by pull-up resistors.  Pull-up to 3.3V from Si5351 supply.   If one part is actively driving HIGH and another part on the bus drives LOW, there is bus contention with high currents on the I/Os which could lead to part destruction.  

Rgds,
Gary

Re: Harmonic performance - SSB vs CW

Kees T
 

Another aspect of filter design, and to reduce insertion loss, are the capacitors used and their characteristics. In my opinion (and some experience) Mica is the best but expensive, Polystyrene is second best but difficult to find, Ceramic NPO is third, and I'm sure there are exotic types out there too but they cost too much or are not readily available. You have to watch voltage ratings but not so much in the QRP range.

73 Kees K5BCQ 

Re: One question only...

Timothy Fidler
 

one 74LS42 with bcd firing from the three output ports assuming they are TTL compatible  and  say 5 BS250s P channel (or four given the number of  relays now available)  say 80 40 20 and 10 mtr  BPF drivers) small signal Mosfets for switch on zero, (or a ULN 2003 (30c US at Futurlec.com in one off qty)..

Say total cost $1.50 could have put a stake thorough this demon...KISS ?? of course it is board space and cost that drives these decisions.

Alternative for physical relays for switching RF up to say 6W level in 50 z impedance circuits...

 a possibly foolish thought for likes of Kees and Ashar F  for any weakware rework concept.. if you fire the gate of a 2N7000 N mosfet  and drive it hard on from an isolated 10V dc supply* with a limiting resistor and a decent say 1mH mini inductor in both the  supply and return legs of the 5V psu - will the thing pass RF as if it were a small signal relay contact ?  RDS fully on for above part approx 1.2 ohms.

* homemade with a Hartley osc design to reduce the snap and crackle.

Re: Harmonic performance - SSB vs CW

MadRadioModder
 

I’m using this one  https://www.ebay.com/itm/192131587815, but it’s only because I have a 150W amp after the uBITx inside the same box.  It’s smaller than the uBITx though and has a directional coupler on board.  I would think, as others have mentioned,  Hans’s QRP-Labs modules would work as a good starting place for those looking for a cheap alternative (https://www.qrp-labs.com/lpfkit.html).  They are quite good and $4.60 each band.  I’ve used them a lot..hard to beat.  There are some other QRP radios that have good outboard filter boards that are excellent too…Kees has some kits that are very well designed.  The one at HFProjects.com is pretty good too…

 

 

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of m5fra2 via Groups.Io
Sent: Monday, August 6, 2018 11:39 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonic performance - SSB vs CW

 

Which after-market board is that? I would like to keep any new boards in the same case.

 

Colin – M5FRA

 

 

 

From: BITX20@groups.io <BITX20@groups.io> On Behalf Of MadRadioModder
Sent: 06 August 2018 16:45
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonic performance - SSB vs CW

 

“rip out the filters and the relays and even the TR relay (KT1-3 and K3) and route everything to an external low pass board. “

 

Yup, done, works… gains almost 30bd in suppression with a good/cheap after-market board.

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of ajparent1/KB1GMX
Sent: Monday, August 6, 2018 9:52 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonic performance - SSB vs CW

 

Warren,

Not so much disagree as wow! that's terrible... can't be!  Yes, its horrific.

5 images.
NF1 is RX mode with no filters (default is 30mhz) selected and RF applied at RX amp output
and measured at antenna connector.  It is really bad!

F3.5 is TX mode, 3.5mhz filter selected.  The blow by limits us to barely 25DB at 30mhz.
F70 is tx mode, 7mhz filter selected.  Blow by limits us to maybe 25db.
F21 is tx with 21mhz filter selected. blow by limits us to about 30db at 30mhz, not impressive.
F300 is TX with last filter 30mhz, span extened to 50mhz so we can see it.  at 45mhz we
are barely 30db down.

Results, we can argue magnitude but over all the filters are unacceptable in situ.
The first one (NF1) was the one that caught my eye as its pure blow by as there is no
direct path its all random coupling.  That has to be much better before we even
consider filters.

At this point I'd rip out the filters and the relays and even the TR relay (KT1-3 and K3)
and route everything to an external low pass board.  At that point I think we have a chance
with the external board performance being unknown but for certain cannot be worse.

 

Image removed by sender.

Virus-free. www.avg.com


--

…_. _._


--

…_. _._

Re: Harmonic performance - SSB vs CW

ajparent1/KB1GMX <kb1gmx@...>
 

Warren,

Pin diodes have a lot going for them.  I like them.

What I have to try is two of the filters from the board and using the
poor mans pin the 1n4007 diode in pairs two on each side back to back.   

Worth a try and what is the cost if it fails.... not much. For two filters 
and 8 diodes its maybe 20cents in parts.

I've used that to 5W for TR switching with good result.

Allison

Re: Harmonic performance - SSB vs CW

Kees T
 

Since we're all using W3NQN's LPF design/layout from the 1999 QST (Hans, Farhan, me too). Typically 3 inductors, capacitive input/output, etc. The only difference I see is that my designs use a capacitor across L2 Unless someone has already looked into that, I'll  fire up "Elsie" and see what it shows.

73 Kees K5BCQ

Re: Harmonic performance - SSB vs CW

MadRadioModder
 

I’m considering some BAP64’s for switching tiny bandpass filters in and out in the uBITx right now.  Elecraft nailed this concept in their K3S design, using them everywhere from the front end almost up to the audio stage for switching things in and out…

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of Warren Allgyer
Sent: Monday, August 6, 2018 4:38 PM
To: BITX20@groups.io
Subject: Re: [BITX20] Harmonic performance - SSB vs CW

 

The reason I mention PIN diodes is these are used in the pristine RS-HFIQ SDR which is a superb all band performer. A significantly higher price point and it is not a standalone. But it is a superb radio and it uses PIN diodes for its filter switching. Yes there is significant current drain but there is also relay holding current in this design. Not advocating PIN, simply raising the possibility.

WA8TOD


Virus-free. www.avg.com

--

…_. _._

Re: Harmonic performance - SSB vs CW

Kees T
 

Warren, 

That is great that you tested the coils you removed from the uBITX on a perf board. Were the capacitors also the ones off the uBITX board ? If not, could you add those to your test (I know that asking is a lot easier than doing). You would like to completely duplicate what is on the uBITX board.

As far as it being a layout problem, contact resistance, added wiring capacitance and the actual parts used would be the determining factors. 

This is what I posted on the other thread:
I just looked at the uBITX schematic for 2017 (I believe it to be level V3) and looked how the band switching of 4 LPFs is done with 3 Relays KT1 - KT3. I'm told V4 is the same.

The 3 relay control signals are TXA (KT1), TXB (KT2), and TXC (KT3) which come from the Raduino board.  

The 4 stages of TXA/TXB/TXC are (0 indicates relay not "picked", 1 indicates relay "picked") : 
0,0,0 which selects the High Band LPF and uses 3 series relay contacts, 
1,0,0  which selects the Hi Midband LPF and uses 5 series relay contacts.
1,1,0  which selects the Lo Midband LPF and uses 7 series relay contacts
1,1,1 which selects the Low Band LPF and uses 7 series relay contacts.

That's a lot of series (resistive) relay contacts and associated (capacitive) wiring. Adding one more relay,  and associated driver, could cut that down to a max of 2 series relay contacts for each band (one on Input and 1 on Output) and "bus" type LPF wiring (results in less overall wiring). 

73 Kees K5BCQ