Date   

Re: Compliance Summary - other radios

Skip Davis
 

Bill I also have one of the Dentron radios for 20 meters, plus I have a spare board with the filter and I think also the RX frontend board in a box. One day I’ll get it out of the box and back on the air.

Skip Davis, NC9O

On Aug 11, 2018, at 12:02, Bill Cromwell <wrcromwell@gmail.com> wrote:

Hi,

Apparently Dentron never got that radio into full production. I have only ever seen mine for 80 meters and another I exchanged info with another ham. His was for 20 meters. We had the same 5 MHz VFO and 9 MHz xtal filter. I will finish building a radio on it.

73,

Bill KU8H

On 08/11/2018 06:41 AM, Timothy Fidler wrote:
arghgh rew WBand SSB that might be one of those MPa'70 home built amps
as in 20 dollars on ebay and no filter set or perhaps an RM Italia amp
with no filter (some yes, some no plenty of hoods off photos on the
internet to show the awful truth) and run right at the stops in terms
of drive. .. and the FCC does nix even though they are factory built
amps and should have compliance tests before entry.


Timothy E. Fidler : Engineer BE Mech(1) Auckland , NDT specialist AINDT
UT /RT3 , MT2
Telephone Whangarei 022 691 8405
e: Engstr@netspacenet.au



----- Original Message -----
From:
BITX20@groups.io

To:
<BITX20@groups.io>
Cc:

Sent:
Sat, 11 Aug 2018 06:03:33 -0400
Subject:
Re: [BITX20] Compliance Summary - other radios


Hi,

I have some other solid state gear and the filtering is more
serious. My
latest purchase (used) is a Hendricks PFR-3 and with only three
bands it
contains as many of those funny little donuts than the uBitx. They are
also well placed. My old Atlas has a complete set of low pass
filters on
the output AND the interstage filters. It resembles a box of donuts
from
a real donut shop. I have a Ten Tec receiver that covers from 300
kHz to
30 MHz and has seven switched filters in just the front end and some
more in the synthesizer section.

I also have a Dentron 80 meter rig rated for 12 watts with
absolutely no
filter after the final!! I have it sitting on the shelf just because I
cared to remove the cover and look. No filter shows up on the
schematic.
Can't be. But it is. Now that uBitx has my attention I will probably
roll a set of filters for that Dentron too.

UBitx is fixable.

73,

Bill KU8H

On 08/11/2018 12:32 AM, Dennis Yancey wrote:
> I wonder what other kits and factory radios actually look like in
> comparison on exactly the same teapots. For instance, I was
listening to
> a gentleman who is an extra class talking on 40 meters last night. He
> USA using a high dollar radio, so he said, a high dollar amplifier and
> he was 3 states away from me. His emissions were covering 8 kHz on
> either side of the frequency he was using. That is just one of many. .
> --
> 72 and God bless
> KD4EPG
>
--
bark less - wag more



Re: UBitx Crystal filter capacitor Q

Ashhar Farhan
 

We use NP0 anyway. I would look at what your termination impedance is and how strong it is. ideally, you must use series resistors to get a good, strong impedance at both ends.

- f

On Sun, Aug 12, 2018 at 9:04 AM, George via Groups.Io <rx3arg@...> wrote:
Caps should be NP0 to have stable result. Filter itself very simple with flat responce but but still the skirt or shape is too great - it is better to do real 8-crystal filter as it is not so difficult. I did so and got 2.5 filter/6dB near 2 shape/60 db. Caps were 68 to 82 on small experimental board. Best regards George RX3ARG/UR4CRG



Re: UBitx Crystal filter capacitor Q

George
 

Caps should be NP0 to have stable result. Filter itself very simple with flat responce but but still the skirt or shape is too great - it is better to do real 8-crystal filter as it is not so difficult. I did so and got 2.5 filter/6dB near 2 shape/60 db. Caps were 68 to 82 on small experimental board. Best regards George RX3ARG/UR4CRG


Re: Proposed fix for Harmonics Issue

Howard Fidel
 

Allison:
Thanks for your comments. You always are spot on. I forgot about  receive when I considered putting the transformer externally. Still, it is much simpler then tearing everything up. I planned on doing exactly what you said and try one to test it. As you may recall, I am waiting for replacement Arduino nano, so I can't do the work until I get them,  and I have some other commitments that   probably will prevent me from testing it until the end of the month. I don't think the increase in current is a serious issue, although you never know. The current will increase by 1.4 X.
It seems to me to be the simplest fix. I did the design for 80 meters, and was able to keep the same inductor values. Hopefully that scales to the other filters, but I didn't do the design for them yet. Actually  the new filter  is flatter in the passband then the original filter.

Howard


On 8/11/2018 9:42 PM, ajparent1/KB1GMX wrote:
Gordon, Howard,

With all the parallel thread it was somehow missed.

FYI memory says both capacitance and inductance change.

Looking at the idea.  I think the basis is sound.  It also may improve the trace
to trace coupling (less coupling).  I worry that the increased currents may
highlight other board level RF current loops.  The output transformer has
to be rewound and that in itself could be an improvement. The ouch is that
converting back to 50 ohms and in a spot before K3 (TR) to not impact RX.
Also transformers have loss and bulk.

The current are likely still low enough and the traces may be light but I think 
the worst that can come of it is some unwanted likely small loss.  They will
not burn up but it is always a risk running lower impedance as the I^2R
losses go up for conductors.  I can't easily evaluate as its RF and skin
effects and wire diameter are are higher and messy to calculate plus it
would have a lot of not validated assumptions.  Its faster to test than guess.

If someone can try it,  they only need to do the 80m filter, wind two transformers
to see and test...

FYI ALL:
I added a image NF1.jpg its a capture from the Rigol on the 6th showing the
blow by of the relays from where the power amp input to the filter selection
relays to the antenna terminal in receive mode. no relays active.  So what
your seeing going past the 10m filter to the antenna jack via layout and
open TR relay (K3)  contacts.   The markers in the table are set to the
3db cutoff or close for a reference point.  It translates to if the filters
are perfect the is as good as it gets due to wires and relays.  What
would I like to see for that case? better than -60db.  The only place
it gets there is 1mhz (extreme left edge).  The source signal is 0dbm
or 1mW.



Re: UBitx Crystal filter capacitor Q

Rahul Srivastava <vu3wjm@...>
 

so
>>do we want hi or low Q?

It is  advisable to use High Q capacitors in filter circuits as well as use of high Q crystals.

>>how is the “response” better?

Response is better in terms of filter skirt resulting in a better shape factor, lower losses and over all a better performing filter.. In my experiments I have noticed a major impact on the rounding off at the filter bandpass edges. They are much better when Q of capacitor is increased.

>>does paralleling capacitors increase or lower the Q?

Paralleling the capacitors will result in increase of Q.

>>what does “touchy”mean when talking about values between 82 and 100pF?

If you will notice as mentioned by others for a change of just about 18pF in capacitor value bandwidth changes almost by a factor of 2. So touchy here refers to capacitor values being critical, a couple of pF change has a significant impact and if you had been using a trimmer to tune this filter then the adjustment would be very sensitive and critical or Touchy...

If making filters I would highly recommend reading and archiving this paper by Wes Hayward:



Rahul VU3WJM


Re: Finding the right version of schematic, voltage and signal at test points #ubitx-help

ajparent1/KB1GMX <kb1gmx@...>
 

the voltages are not magic.

If a relay is picked the collector for that relay is around .8V or less.
and it base will be about .65-.7V.  The output from the micro is about 5.

For the not picked relay condition the collector will be close to power supply voltage.
For that I assume 12V or more. and the base will be 0V or well under .1V
The micro will be about .1V or less (essentially ground).

There is a caveat... all voltages are 0 (or close to zero) around the relays
and the transistors that drive them for RX as K1 supplies TX+ (transmit
DC whatever the power supply is set at) to the filter relays and they
actually activate (picked).  The outputs from the micro are based on
what dial frequency is set so you see essentially 0V or near 5V on
TX_A, TX-B and TX-C. also TX+  operates K3 (TR).

The info I gave is for what frequency the specific relays operate or not.
The relays are wired in a cascade tree so for the three signals they
are a partial decode. Clever but causes other issues.

Allison


Re: Spurs - BPF fix? #ubitx

ajparent1/KB1GMX <kb1gmx@...>
 

On Sat, Aug 11, 2018 at 05:53 PM, Glenn wrote:
Allison,
 By my reckoning, only three filters are needed. 
By that you mean 12/10m, 15m and added 20/17.

LPF for 17M band to remove 27MHz spurs
The 17M band does have a spur there but its caught by the known
anemic output filters as they are.  No need to filter but not a bad thing.
it just adds work.   The filter for that band is a LPF identical to the
existing 20/17m filter at the output( or move the corner frequency
up a tiny bit for prefernces).

However my comment on 4 filters was to allow low pass filters to the
input to the amp for the lower bands plus one for 10M as well.
If we do 15M then its 5 filters.  Those being 80/60, 40/30, 20/17,
using compact copies of the output fitlers plus new 15m BPF
and for a wide BPF 12 and 10.  That's a total of five done that
way and it can be in the RX path as it cleans it up and does
not limit tuning. That simplifies switching too. That assumes
we toss the whole 33mhz 9element deal as redundant.

BPF for 15M to remove 20MHz spurs
The 15m filter has to block the spurs at 23.55 to 23.8 not 20M.

LPF for 12-10M to remove 20MHz spurs.
Actually these are at 17mhz (10M) and 20mhz(12M) 

That's OK although the BPF for 15M is do-able I think but very tricky to get the peak correct and also with low losses.

The other issue is switching though.
      The current switching of the output LPF's is 000 which covers 15 to 10M
      We need to isolate though the 15M band BPF from the output LPF's. 
Depending on the switching needs, anything is possible.  First it need to be tested
if only with manual inputs.  Bottom line is we can select 4 filters with two lines
or create more as needed.

Depending on how the output filters are handled that may change.
I also posted a few weeks back I posted a mod to make the 4 filter
selection using two diodes and TX-B and TX-C, then TX-a was
free for use. The mod itself (adding two diodes and removing Q17)
is a trivial one.

Also I've written that if 4 more lines are needed the LCD lines can be multiplexed
as they tend to have long periods (in computer time) of no activity.  So creating
a whole new set of relays selectors is not a large or difficult task.  It is tested!
I used TX-A and a common cheap not likely to disappear chip for that as I
needed more filter selector for those filters at the input to the amp should
they not coincide.

The loading new code is a pain for some but its not coding as a new set
of files can be made available to just install. (or compile&install)
are we on the same page?


Re: #ubitx SSM2167 mic compressor speaker feedback issue - resolved #ubitx

John (vk2eta)
 

Hi Kevin, 

Glad you sorted it out. Enjoy.

73, John


Re: Spurs - BPF fix? #ubitx

Glenn
 

CORRECTIONs:

 By my reckoning, only three filters are needed. 

LPF for 17M band to remove 27MHz spurs

BPF for 15M to remove 23.55MHz spurs

LPF for 12-10M to remove 20MHz spurs.


On Sun, Aug 12, 2018 at 10:53 AM, Glenn wrote:
Allison,
 By my reckoning, only three filters are needed. 

LPF for 17M band to remove 27MHz spurs

BPF for 15M to remove 20MHz spurs

LPF for 12-10M to remove 20MHz spurs.

Thats OK although the BPF for 15M is do-able I think but very tricky to get the peak correct and also with low losses.

The other issue is switching though.
      The current switching of the output LPF's is 000 which covers 15 to 10M
      We need to isolate though the 15M band BPF from the output LPF's. 

With current firmware and logic for the output filters, we cannot isolate 15M pre PA BPF.

glenn

 On Sun, Aug 12, 2018 at 10:53 AM, Glenn wrote:
 By my reckoning, only three filters are needed. 

LPF for 17M band to remove 27MHz spurs

BPF for 15M to remove 20MHz spurs

LPF for 12-10M to remove 20MHz spurs.


Re: Proposed fix for Harmonics Issue

ajparent1/KB1GMX <kb1gmx@...>
 

Gordon, Howard,

With all the parallel thread it was somehow missed.

FYI memory says both capacitance and inductance change.

Looking at the idea.  I think the basis is sound.  It also may improve the trace
to trace coupling (less coupling).  I worry that the increased currents may
highlight other board level RF current loops.  The output transformer has
to be rewound and that in itself could be an improvement. The ouch is that
converting back to 50 ohms and in a spot before K3 (TR) to not impact RX.
Also transformers have loss and bulk.

The current are likely still low enough and the traces may be light but I think 
the worst that can come of it is some unwanted likely small loss.  They will
not burn up but it is always a risk running lower impedance as the I^2R
losses go up for conductors.  I can't easily evaluate as its RF and skin
effects and wire diameter are are higher and messy to calculate plus it
would have a lot of not validated assumptions.  Its faster to test than guess.

If someone can try it,  they only need to do the 80m filter, wind two transformers
to see and test...

FYI ALL:
I added a image NF1.jpg its a capture from the Rigol on the 6th showing the
blow by of the relays from where the power amp input to the filter selection
relays to the antenna terminal in receive mode. no relays active.  So what
your seeing going past the 10m filter to the antenna jack via layout and
open TR relay (K3)  contacts.   The markers in the table are set to the
3db cutoff or close for a reference point.  It translates to if the filters
are perfect the is as good as it gets due to wires and relays.  What
would I like to see for that case? better than -60db.  The only place
it gets there is 1mhz (extreme left edge).  The source signal is 0dbm
or 1mW.


Re: Harmonics measured by Warren. How bad?

Jerry Gaffke
 

>  FYI, Allison mentioned replacing the 4th LPF with a 10M BPF ...

https://groups.io/g/BITX20/message/55789


On Sat, Aug 11, 2018 at 05:26 PM, RCBoatGuy wrote:
FYI, Allison mentioned replacing the 4th LPF with a 10M BPF and gave a general description of the results back in a post on Aug 4th.  Sadly, I accidentally closed the window before I copied the link to the post, or I'd post it here.  :(


Re: Compliance Summary

Gordon Gibby <ggibby@...>
 

Alyson that is perfect !!   Thanks!!!


On Aug 11, 2018, at 20:56, ajparent1/KB1GMX <kb1gmx@...> wrote:

Gordon,

Better picture least it looks good on the big screen.
<no_relay.JPG>


Re: Compliance Summary

ajparent1/KB1GMX <kb1gmx@...>
 

I just realized how rough from all the mods tried and all the board was getting.

Allison


Re: Compliance Summary

ajparent1/KB1GMX <kb1gmx@...>
 

Gordon,

Better picture least it looks good on the big screen.


Re: Spurs - BPF fix? #ubitx

Glenn
 

Allison,
 By my reckoning, only three filters are needed. 

LPF for 17M band to remove 27MHz spurs

BPF for 15M to remove 20MHz spurs

LPF for 12-10M to remove 20MHz spurs.

Thats OK although the BPF for 15M is do-able I think but very tricky to get the peak correct and also with low losses.

The other issue is switching though.
      The current switching of the output LPF's is 000 which covers 15 to 10M
      We need to isolate though the 15M band BPF from the output LPF's. 

With current firmware and logic for the output filters, we cannot isolate 15M pre PA BPF.

glenn


Re: Harmonics measured by Warren. How bad?

ajparent1/KB1GMX <kb1gmx@...>
 

Carl,

That discussion is going on maybe three different threads. So ts easy to 
get it all scrambled.

I do get amazed that someone will create a whole new topic for something
already under discussion or to answer someone in that topic.  I can only
figure the email side of this really messes with the flow.

Allison


Re: Harmonics measured by Warren. How bad?

RCBoatGuy
 

Warren,

FYI, Allison mentioned replacing the 4th LPF with a 10M BPF and gave a general description of the results back in a post on Aug 4th.  Sadly, I accidentally closed the window before I copied the link to the post, or I'd post it here.  :(

73,

Carl, K0MWC


Re: Spurs - BPF fix? #ubitx

ajparent1/KB1GMX <kb1gmx@...>
 

Its not possible to remove the 23.55 spur as that generated when on 15M
band without a distinct band pass fitler for it   It varies from 23.55 to 23.8.
It is better to put 15M out of reach of the 12/10M filter with a slightly higher
cutoff.  The notch shown in the sweep would prevent 15M use anyway.

So for 15 you need a narrow band pass and your wide high pas with
the existing low pass for 12 and 10 and where legal 11M.

I did build a 15M Band pass as a tapped parallel,series top coupled,
tapped parallel. You can't model that in the free Elsie as too many
components pops up. so I split the filter and modeled as two sections
and built it to test.   I put it aside as the best I could get was about
25db attenuation in the 23.5 to 23.8 window with reasonable
insertion loss.  Looking back that would be enough.

 For 17M down to 80M:
You could take those parts off the board and replace them with jumpers.
Then make up a board or separate filters with connectors using the same parts.
I dead bugged mine using 1.00"x.75" pieced of copper clad into those tiny
pomona boxes (1.5x1.25 inch) with (I use SMA) connectors on the ends. 
IF you do that use the larger 2.25inch, less cramped.  The filters themselves
and the parts to build them are not at issue. In a separate box they perform
well and solve the problem as originally intended. no design work needed
other than to layout some form of switched board.

Allison


Re: Compliance Summary

jim
 



On Saturday, August 11, 2018, 3:27:08 PM PDT, ajparent1/KB1GMX <kb1gmx@...> wrote:


Jerry,

That ended up in a different topic.  Compliance summary....

 RBW Receiver Bandwidth Window


Allison

Back in the old days, RBW was called "Resolution Bandwidth" ..How close in frequency can 2 (or more) signals be to be seen as seperate  signals.  ...Still is on my hp-8569b SA

Jim


Re: Compliance Summary

Gordon Gibby <ggibby@...>
 

Hi Alyson, no big deal, the cell phone is fine, but if you’re able to get one with a little more even lighting that covers all the wiring underneath where the relays were,  that would be great.

I’ll capture it, and then mark  which traces need to be cut so I can create a little manual for this. 

Thanks!!!


On Aug 11, 2018, at 18:51, ajparent1/KB1GMX <kb1gmx@...> wrote:

Gordon,

If you need a better picture I can grab the camera instead of the cell.
I doubt there many have remove more than one relay and only
because it was bad.  Even with the right gear its still a pain.

What would helps is a board layout.  

The S meter approach will get you there. Just less convenient.

Allison