All,

My ubitx has quit working in receive and SSB transmit. CW transmit
still works.

Attached are spectrum analyzer traces at C216 and C210, i.e. the input
and the output of the 45Mhz crystal filter.

Does anyone else have a spectrum analyzer they can use to verify that
this is what I should be seeing?

It's hard to believe that the 45Mhz crystal filter has a 30db insertion
loss.

I have removed nd6t's agc circuit and the thump filter circuit. The
ubitx is basically in standard configuration. The CW sidetone comes
through the audio section just fine, just no received audio.

My test signal is at 3900khz. The 48.915Mhz signal (marker 4) is about
where it should be to put the 3.9Mhz signal at 45Mhz. I am a little
surprised at the vfo signal being so strong, I would have thought the
first mixer would have suppressed the vfo signal more than this.

Any help or comments would be appreciated.

tim ab0wr

The 45mhz filter is rather narrow If memory serves its a 15khz (wide) unit
which is typical for common use (FM first IF work).   So for low attenuation
you need to be between 44.9925 to about 45.0075mhz.

The one marker is 44.9625mhz  which is 37k Khz off of center, I'd expect a
fair amount of attenuation there compared to inband.

Could your LO cal be off or the RIT/PBT be active?

Allison

Something is mighty fishy.

Or the input frequency for the VFO as set should 3.915mhz.

48.915mhz -45.000= 3.915mhz

Try raising the signal generator from 3.8mumble to 3.915 and see if things get better.

The LO for converting from 45 to ~12mhz if should be either 33. or 57. withoutly a
few khz from those.  IF the SA is right that seems off too.

Allison

Allison,

I just set the markers manually using eyesight. I don't know if they
are exactly on frequency or not. Tomorrow I'll let the SA do a peak
search and see what the actual frequencies are.

No RIT or PBT.

Since I'm looking at the output of the first mixer, the VFO is the only
frequency that applies.

I'm using a cheap MH5200 signal generator so I won't guarantee it's
accuracy.

I'll rerun the test tomorrow and see what comes up.

tim ab0wr

On Thu, 03 May 2018 18:13:01 -0700
"ajparent1/KB1GMX" <kb1gmx@arrl.net> wrote:

The 45mhz filter is rather narrow If memory serves its a 15khz (wide)
unit which is typical for common use (FM first IF work).   So for low
attenuation you need to be between 44.9925 to about 45.0075mhz.

The one marker is 44.9625mhz  which is 37k Khz off of center, I'd
expect a fair amount of attenuation there compared to inband.

Could your LO cal be off or the RIT/PBT be active?

Allison

I don't think you are seeing the LO frequency on the scan. You are
seeing a mixing product. The scan shows spurious responses at about
8Mhz, 18Mhz, and 31Mhz.

I'll check the other LO and the BFO to see what frequencies they show.
It might be that the 5351 is hosed.

I haven't tried the base software to see what happens. That is
something else I can do.

Till I figure out what is happening with the 45Mhz IF there isn't much
use in moving down the chain.

tim ab0wr

On Thu, 03 May 2018 18:26:11 -0700
"ajparent1/KB1GMX" <kb1gmx@arrl.net> wrote:

Something is mighty fishy.

Or the input frequency for the VFO as set should 3.915mhz.

48.915mhz -45.000= 3.915mhz

Try raising the signal generator from 3.8mumble to 3.915 and see if
things get better.

The LO for converting from 45 to ~12mhz if should be either 33. or
57. withoutly a few khz from those.  IF the SA is right that seems
off too.

Allison

Allison,

I redid the scans with a 1khz resolution in order to help with the
frequency resolution.

Doing this shows the 45mhz signal to be right on. The 33mhz signal is
right on also except in a couple of the scans I picked the wrong peak.

Frequencies at the ends of the scans are off for some reason. The sig
gen *is* at 3900khz and the vfo is also at 48900khz. My freq counter
and my scope (rigol ds1074) both confirm that.

Why the 45mhz filter drops the 45mhz signal by 50dbm and the 48.9mhz
signal by only 30db still concerns me.

I've also included scans at a couple of other points in the receiver.

I must confess I'm at a loss here. My next step will probably be to
turn the drive off to the finals, connect up my 2-tone generator and
try to trace the signal through the transmit side to see where it is
disappearing! If it disappears in the 45mhz filter then I'll need to
contact Ashar to see what filter he is using.

Thanks for your help.

tim ab0wr

On Thu, 03 May 2018 18:13:01 -0700
"ajparent1/KB1GMX" <kb1gmx@arrl.net> wrote:

The 45mhz filter is rather narrow If memory serves its a 15khz (wide)
unit which is typical for common use (FM first IF work).   So for low
attenuation you need to be between 44.9925 to about 45.0075mhz.

The one marker is 44.9625mhz  which is 37k Khz off of center, I'd
expect a fair amount of attenuation there compared to inband.

Could your LO cal be off or the RIT/PBT be active?

Allison

Arv,

It's the 45Mhz crystal filter on the ubitx. Not one that I designed. If
the 45mhz crystal filter had a major design flaw it would be affecting
the hundreds of other ubitx units out there.

I've taken the radio all apart. I'm going to build up a new w0eb
radi2cino daughterboard to use with the existing motherboard to see if
I have a bad connection on the radi2cino I previously soldered
together. I already found one and fixed it. And the 5351 seems to be
putting out the right frequencies. CW transmit works fine but not SSB
transmit or receive.

I might try jumpering around the 45mhz crystal filter and see if that
helps anything.

tim ab0wr

On Fri, 4 May 2018 12:28:46 -0600
"Arv Evans" <arvid.evans@gmail.com> wrote:

Tim AB0WR

If you were to load your filter components into LTSpice and do a
gain-frequency run
on the design it might highlight where the problem is. With LTSpice
it is possible to
do what-if changes to optimize your filter designs.

I usually include the stage before and the stage following an existing
filter because
that evaluates the filter using its actual in-circuit impedances and
drive levels.

Arv K7HKL
_._


On Fri, May 4, 2018 at 12:12 PM, Tim Gorman <tgorman2@cox.net> wrote:

Allison,

I redid the scans with a 1khz resolution in order to help with the
frequency resolution.

Doing this shows the 45mhz signal to be right on. The 33mhz signal
is right on also except in a couple of the scans I picked the wrong
peak.

Frequencies at the ends of the scans are off for some reason. The
sig gen *is* at 3900khz and the vfo is also at 48900khz. My freq
counter and my scope (rigol ds1074) both confirm that.

Why the 45mhz filter drops the 45mhz signal by 50dbm and the 48.9mhz
signal by only 30db still concerns me.

I've also included scans at a couple of other points in the
receiver.

I must confess I'm at a loss here. My next step will probably be to
turn the drive off to the finals, connect up my 2-tone generator and
try to trace the signal through the transmit side to see where it is
disappearing! If it disappears in the 45mhz filter then I'll need to
contact Ashar to see what filter he is using.

Thanks for your help.

tim ab0wr

On Thu, 03 May 2018 18:13:01 -0700
"ajparent1/KB1GMX" <kb1gmx@arrl.net> wrote:

The 45mhz filter is rather narrow If memory serves its a 15khz
(wide) unit which is typical for common use (FM first IF work).
So for low attenuation you need to be between 44.9925 to about
45.0075mhz.

The one marker is 44.9625mhz which is 37k Khz off of center, I'd
expect a fair amount of attenuation there compared to inband.

Could your LO cal be off or the RIT/PBT be active?

Allison

3.83333+45= 48.3333  Or if on the other side it would be 45-3.8333=41.1667
image C30 look at marker 1D as that likely the LO leakage so that about right.
Image C216 has a solid signal at 45MHZ but the LO indicates 48.91333

I see neither frequency in the scans that close enough to pass through the filter (its only 15khz wide!).

Have you simply dialed the sig gen or the VFO until the signal appears at 45.0000mhz (or the speaker)?
Based on what the SA is telling you your not putting a signal in the passband of the filter.  Since I have
a DSA815T (its a handy machine!) I'm inclined to believe its telling you that.

If you have the 815T use the tracking generator. Simply put TG ar 45mhz with say 50khz span and inject
that into the filter matching network (50 ohm point) and look at the output.  You should see the filter pass
band though the insertion loss with external circuits loading it will be higher it should be under 10db on
a guess.  Hint use a lower sweep rate or auto locked to RBW.

The only way that filter would die is if you applied force to it ripping the leads out or shattering the quartz.
Even then have pulls I've dropped on concrete that are fine.  Recently I helped someone with another kit 
and the problem was a ball of solder under the crystal causing a lead to ground short.  It Ohmed out.
If I had to guess the series inductor (ohm it) is open or maybe a short at the cap should be checked for.

If you truly believe its dead and there are no shorts bridge it with a wire.

Allison

The input frequency was 3900khz. verified by my freq counter and my
oscilloscope. The 815 just doesn't do a good job on frequencies at the
edges of a large range apparently. I didn't show it but when I had the
scan range from 3500khz to 4khz it the 815 showed the carrier to be at
3900khz.

The same thing probably applies for the 48.91333 signal. For a 3900khz
signal the vfo would be at 48900khz so the 815 is off by about 13khz at
the edge of the scan.

The 45Mhz signal should pass through the 45Mhz filter. Yet it is
knocked down further than the 48.9333mhz signal. That's why I wonder if
the 48.9Mhz signal isn't just being picked up out of the air by my
probe.

I will be working in the yard most of today. If I get a chance tonite
to put everything back together I'll try using the tracking generator
as you suggest. I don't know why I didn't think of that! I appreciate
the suggestion.

That should even work if I just set the TG to run from 3875khz to
3950khz and feed it in the antenna while leaving the vfo set to
3900khz.

I set the sig gen to 3900khz and the radio to 3900khz as well during my
tests.

Thanks for the help.

tim ab0wr

On Fri, 04 May 2018 15:52:33 -0700
"ajparent1/KB1GMX" <kb1gmx@arrl.net> wrote:

3.83333+45= 48.3333  Or if on the other side it would be
45-3.8333=41.1667 image C30 look at marker 1D as that likely the LO
leakage so that about right. Image C216 has a solid signal at 45MHZ
but the LO indicates 48.91333

I see neither frequency in the scans that close enough to pass
through the filter (its only 15khz wide!).

Have you simply dialed the sig gen or the VFO until the signal
appears at 45.0000mhz (or the speaker)? Based on what the SA is
telling you your not putting a signal in the passband of the filter.
Since I have a DSA815T (its a handy machine!) I'm inclined to believe
its telling you that.

If you have the 815T use the tracking generator. Simply put TG ar
45mhz with say 50khz span and inject that into the filter matching
network (50 ohm point) and look at the output.  You should see the
filter pass band though the insertion loss with external circuits
loading it will be higher it should be under 10db on a guess.  Hint
use a lower sweep rate or auto locked to RBW.

The only way that filter would die is if you applied force to it
ripping the leads out or shattering the quartz. Even then have pulls
I've dropped on concrete that are fine.  Recently I helped someone
with another kit and the problem was a ball of solder under the
crystal causing a lead to ground short.  It Ohmed out. If I had to
guess the series inductor (ohm it) is open or maybe a short at the
cap should be checked for.

If you truly believe its dead and there are no shorts bridge it with
a wire.

Allison

THe 815 is likely doing it right.   I should have warned you that with wide scans the number
of data points are spread out and interpolation is the result with soso accuracy.  For better results 
use a narrower scan (span).    All but the older full analog SAs have that problem.  Somewhere
in the manual it will give you the number of point it actually takes measurements at.  Typically
its some amount of memory limit or the width of the screen in Pixels.

The best approach is setup for 45mhz, maybe 50khz span and stuff that though if you have
the 815T.  

If you have the plain 815 (no tracking generator) set it for write hold and sweep it very slowly
with signal generator or a high output noise source and you will get the passband  outline.
Its slow doing it that was but it works.  I've used this on my 8568B as that does not have
a tracking generator so I use a noise source to fill it.

FYI you can dial the VFO (5351) down to 45mhz and enough of it will be
there to see in and out.

Allison

Hi Tim,

I'm interested to know how your connecting your SA to the circuit, from my recollection the majority of SA equipment has a 50 ohm input impedance, an input impedance that low is not conducive to hanging a probe off the circuit at any convient location. 

-Justin N2TOH  

I'm using a direct probe. The 815 has a 50ohm input which should be a
match to most points in the ubitx. At least a close enough match to not
directly affect the circuits operation. I suppose I could use a 10x
scope probe but I'm not sure that would make much difference.

tim ab0wr

On Sun, 06 May 2018 01:03:42 -0700
freefuel@gmail.com wrote:

Hi Tim,

I'm interested to know how your connecting your SA to the circuit,
from my recollection the majority of SA equipment has a 50 ohm input
impedance, an input impedance that low is not conducive to hanging a
probe off the circuit at any convient location. 

-Justin N2TOH

This why I picked r216 and r210 as measurement points. I assumed
those points should be close to being 50ohm points as they are outside
the L network matching to the filter.

tim ab0wr

On Sun, 6 May 2018 21:33:06 +0530
"Ashhar Farhan" <farhanbox@gmail.com> wrote:

You must use the L network of the ubitx or use 1:9 balun to get the
proper shape of the fiilter. The 45 MHz crystal filters usually have
a termintion specified for around 600 ohms.

Let me bullet this :
* All filters have the specified shape only when they are terminated
at the specified impedance on both sides.
* The passive filters, including the crystal fitlers have no
measurable termination to speak of at all. They have different
responses at different impedances.
* The capacitors of the ladder filter set the bandwidth.
* The termination impedance sets the ripple. Lower impedance brings
steeper skirt at the cost of ripples in the bandpass. To remove the
ripples you increase the impedance and lose the sharpness of the
skirt.


On Sun, May 6, 2018 at 7:45 PM, Tim Gorman <tgorman2@cox.net> wrote:

I'm using a direct probe. The 815 has a 50ohm input which should be
a match to most points in the ubitx. At least a close enough match
to not directly affect the circuits operation. I suppose I could
use a 10x scope probe but I'm not sure that would make much
difference.

tim ab0wr

On Sun, 06 May 2018 01:03:42 -0700
freefuel@gmail.com wrote:

Hi Tim,

I'm interested to know how your connecting your SA to the circuit,
from my recollection the majority of SA equipment has a 50 ohm
input impedance, an input impedance that low is not conducive to
hanging a probe off the circuit at any convient location.

-Justin N2TOH

IF the filters are not matched to 50 ohm the insertion loss due to mismatch is very high as well
never minding the bandpass distortion.  That means the network on the board has to be part
of the path on both sides.  Farhan is not pulling your leg on needing the matching networks.

Its very difficult to probe a board using a scope probe for a SA due to the 50 ohm input
and most scope probes are nominal 1MOhm using a high impedance cables.  When 
I do it insitu I make up a cable using RG316 (teflon rg174) and make very direct connections.
Failure to do that means the results at best will be an poor approximation.   When I have
to use a higher impedance probe to not load the circuit I use a active RF probe that
presents a 1Mohm load at maybe 1pF as well, the electronics are in the probe head.
The fly lead for ground is made as short as possible as well. 

THe two pole 45mhz filter are common. Standard stuff is 15khz bandwidth and
common in UHF or dual band HT and other VHF and up radios.  The 4pole 7.5khz
wide filters are less common but I found a few for a project, they weren't cheap.

Allison

Allison,

I am not an expert on digital anything, scopes or spectrum analyzers.
Thanks for the information. Someday I need to find time to study up on
all this but then I also need time to get the garden going, mow the
yard twice a week, etc! Never enough time.

I do have the 815tg. When I get everything put back together I'll give
the narrow span a try.

I assume you mean by taking the vfo to 45Mhz you mean setting the dial
to zero?

tim ab0wr



On Sat, 05 May 2018 15:03:46 -0700
"ajparent1/KB1GMX" <kb1gmx@arrl.net> wrote:

THe 815 is likely doing it right.   I should have warned you that
with wide scans the number of data points are spread out and
interpolation is the result with soso accuracy.  For better results
use a narrower scan (span).    All but the older full analog SAs have
that problem.  Somewhere in the manual it will give you the number of
point it actually takes measurements at.  Typically its some amount
of memory limit or the width of the screen in Pixels.

The best approach is setup for 45mhz, maybe 50khz span and stuff that
though if you have the 815T.  

If you have the plain 815 (no tracking generator) set it for write
hold and sweep it very slowly with signal generator or a high output
noise source and you will get the passband  outline. Its slow doing
it that was but it works.  I've used this on my 8568B as that does
not have a tracking generator so I use a noise source to fill it.

FYI you can dial the VFO (5351) down to 45mhz and enough of it will be
there to see in and out.

Allison