Dave,
If you want, I made a layout for the same component footprint as you have. I can upload Eagle "brd" file for it in the file section tomorrow. I think KiCAD can import Eagle. I can also generate gerbers for manufacturing if anyone is interested.
73,
--
Ion
VA3NOI
|
Dave, I eliminated the SMAs. ?Just put the boards next to each other, bridge with? small wires. 73, Gary WB6OGD?
toggle quoted message
Show quoted text
On Feb 11, 2020, at 6:17 PM, Dave VE3LHO <dave@...> wrote:
?It is actually pretty nice isn't it? Gives me inspiration to make something as nice :-) Ugly style is going to be my first attempt but I have 50 of the filters and a bunch of the inductors and caps (at pennies a piece I couldn't just order the 3~4 I really needed)? so I'm going to experiment. We'll see if the etched pcb works better (or worse) than my ugly version. Everything you say is true but none the less pcb are generally "better". I am a little concerned about the SMA connector center pin rotating. In my case it would be both a cap and an inductor that could be destroyed and I am considering some kind of small small land at each connector so I can solder them. Dave VE3LHO On Tue, Feb 11, 2020 at 08:10 AM, Jerry Gaffke wrote:
Erik,
Your ugly style SAW filter board is actually quite pretty!
And if you are getting 70dB of stopband suppression, plenty effective.
Could work better than many attempts at an etched board.
An etched board that only has a groundplane on the back
would need lots of feedthroughs to connect ground points.
Feedthroughs have inductance, and that groundplane is 1.5mm further away.
(In this case, an etched board could have an almost solid ground plane on the front.)
I have had that center SMA pin try to rotate when connecting a cable,
so I try to leave a land of copper at an SMA to solder the pin to.
Otherwise those caps can break.
An alternative might be some 1 meg resistors in a 1206 package
from center pin to the ground pins,?adding strength there.
Sufficient grounding as you have done is important.
Not obvious, but you have left copper under each SAW filter
connecting ?the two halves of the ground plane, that's important.
I'd probably leave the entire ground plane intact except for
a small island on each end for the SMA pin.
I would have thought it might need a shield through the middle?
to isolate one SAW filter from the other, but 70dB is good enough.
Jerry, KE7ER
On Tue, Feb 11, 2020 at 01:09 AM, <erik@...> wrote:
Maybe not as nice but building the filter dead bug style does deliver good performance.
The out of band rejection is at least 70dB, Combined with the SI4432 out of band rejection you get about 100dB which is sufficient.
|
Thanks. Looks good overall.
toggle quoted message
Show quoted text
On Tue, Feb 11, 2020 at 12:59 PM, Ion Petroianu, VA3NOI wrote:
Footprint QCC8C is the same as Murata footprint SM5050-8, for which I have found libraries (symbol and footprint) and uploaded them in the files section, in the folder VA3NOI.
Part number is a resonator (RO3101C)
73,?
Ion
VA3NOI
|
It is actually pretty nice isn't it? Gives me inspiration to make something as nice :-)
Ugly style is going to be my first attempt but I have 50 of the filters and a bunch of the inductors and caps (at pennies a piece I couldn't just order the 3~4 I really needed)? so I'm going to experiment. We'll see if the etched pcb works better (or worse) than my ugly version. Everything you say is true but none the less pcb are generally "better".
I am a little concerned about the SMA connector center pin rotating. In my case it would be both a cap and an inductor that could be destroyed and I am considering some kind of small small land at each connector so I can solder them.
Dave
VE3LHO
toggle quoted message
Show quoted text
On Tue, Feb 11, 2020 at 08:10 AM, Jerry Gaffke wrote:
Erik,
Your ugly style SAW filter board is actually quite pretty!
And if you are getting 70dB of stopband suppression, plenty effective.
Could work better than many attempts at an etched board.
An etched board that only has a groundplane on the back
would need lots of feedthroughs to connect ground points.
Feedthroughs have inductance, and that groundplane is 1.5mm further away.
(In this case, an etched board could have an almost solid ground plane on the front.)
I have had that center SMA pin try to rotate when connecting a cable,
so I try to leave a land of copper at an SMA to solder the pin to.
Otherwise those caps can break.
An alternative might be some 1 meg resistors in a 1206 package
from center pin to the ground pins,?adding strength there.
Sufficient grounding as you have done is important.
Not obvious, but you have left copper under each SAW filter
connecting ?the two halves of the ground plane, that's important.
I'd probably leave the entire ground plane intact except for
a small island on each end for the SMA pin.
I would have thought it might need a shield through the middle?
to isolate one SAW filter from the other, but 70dB is good enough.
Jerry, KE7ER
On Tue, Feb 11, 2020 at 01:09 AM, <erik@...> wrote:
Maybe not as nice but building the filter dead bug style does deliver good performance.
The out of band rejection is at least 70dB, Combined with the SI4432 out of band rejection you get about 100dB which is sufficient.
|
Hi, ? I hacked the standalone mode by adding another trace showing the combined gain of the LNA and PGA during a sweep.? If the AGC is enabled then the gain can be read in Register 0x69 bits 4-0 The vertical scale for the gain is 5dB/division.? T he yellow trace is the 0-255 RSSI value from the SI4432,? I was too lazy to change the scale at the side for the quick hack.This is how the gain varies with agc enabled (x69 = x60).? The dBm value at the peak includes the correction for the filter/mixer loss, and yes I know it should say dBm not dB - fixed after the photos were taken.  Interesting how the gain drops early, and the spike at each change of gain.? Also interesting how the gain starts increasing before we get to the main signal spike.? Presumably the gain algorithm is working on the signal strength before the FIR filter? Fixing the gain at the value the ALC chose, 37dB we get this:  go up to 43dB and we see some spurious signals and the peak RSSI drops - ADC overload?:  Drop down to 25dB and the peak RSSI stays at -13dBm but the minimum detectable signal level increases:  So it looks like the AGC does a great job of setting the best gain level, the RSSI value is compensated for the gain change, but there are unwelcome spikes at the changes in gain.? Perhaps some sort of on-demand "Optimise gain" button is needed where the gain at the peak (or perhaps the minimum through the sweep) is recorded and then used to keep a fixed gain throughout the sweep. I tried detecting a change in gain then delaying and retaking the reading, but that did not offer an improvement.? Perhaps there is some instability and the gain is swapping between different levels? 73 M0WID
|
Footprint QCC8C is the same as Murata footprint SM5050-8, for which I have found libraries (symbol and footprint) and uploaded them in the files section, in the folder VA3NOI.
Part number is a resonator (RO3101C)
73,?
Ion
VA3NOI
|
Hello,
This email message is a notification to let you know that the following files have been uploaded to the Files area of the [email protected]
group.
Uploaded By:
m0wid <david.wilde@...>
Description:
Kicad symbol and pattern files
Cheers,
The Groups.io Team
|
Si4432, 20dbm perhaps a bit much
This is a continuation of an old conversation that deserves its own thread. ? ?? /g/HBTE/message/942? ?? /g/HBTE/message/944Summary: Driving a level 7 mixer such as an ADE-1 or SBL-1 with 20dBm may damage it. It is definitely too much when driving the SAW filter in signal generator mode. The code used to set the transmit power level in the Jan 17 tinySA.ino firmware may not be correct for an Si4432, it appears to be written with some other part in mind. I think that all writes to the transmit power level register of the Si4432 should follow this pattern: ? //? drive:? 0x0=-1dBm? 0x1=2dBm 0x2=5dBm 0x3=8dBm 0x4=11dBm 0x5=14dBm 0x6=17dBm 0x7=20dBm ? SI4432_Write_Byte(0x6D,? 0x30 | (drive&0x7)); Erik's build instructions explicitly state that the mixer should be chosen to handle 20dBm: > The mixer is preferably a level 13 or better mixer. An example of a good mixer is the ADE-25MH > but any passive double balanced diode mixer (such as a ADE-1) that can be used till 500MHz > on its RF and LO port is usable. The mixer should be able to handle +20dBm input on the LO port. A level 7 mixer such as an ADE-1 could work fine if power into the LO port is reduced to 8dBm with: ? ? SI4432_Write_Byte(0x6D,0x33); as per post 944.? Too much power can damage such a low level mixer. Also, I suspect that asking for more power from the Si4432 than needed may create extra noise in the Si4432, and perhaps in the mixer. I doubt the Si4432 has any way of directly connecting the TX port to the RX port using the lna_sw bit as the Si4431 does.? ?So that bit may not be an issue. The B3730 SAW filter that Erik uses can deal with an absolute max of 10dBm. So if going into signal generator mode with the "P6 3" command, be sure to reduce transmit power for the RX Si4432 before enabling the transmitter. No reports yet of this damaging a B3730 SAW filter. The B3790 SAW filter I plan to use can deal with an absolute max of 5dBm. So quite likely that 20dBm would damage it. There are Si4431 modules available, capable of a maximum of 13dBm instead of 20dBm. That may be a safer choice for most of us, but has not been tried. The power levels available are down 7dB from what the Si4432 can provide: ? //? drive:? 0x0=-8dBm? 0x1=-5dBm 0x2=-2dBm 0x3=1dBm 0x4=4dBm 0x5=7dBm 0x6=10dBm 0x7=13dBm ? SI4432_Write_Byte(0x6D,? 0x30 | (drive&0x7)); We could choose to have the variable "drive" be calibrated in dBm rather than arbitrary numbers. This might be more convenient, but is definitely more confusing: #if USESI4432 #define? MIN_DBM? ?-1 #else #define? MIN_DBM? ?-8 #endif SI4432_Write_Byte(0x6D,? ?0x30 + (((drive - MIN_DBM)/3) & 0x7)); Disclaimer: I have not verified any of this by trying it out on hardware, just reading datasheets for now. Jerry, KE7ER
|
Ah, that makes sense.
And I vaguely recall now that you may have stated?
you were using tape in a previous post.
I thought I was seeing the fiber in the exposed fiberglass.
Jerry
On Tue, Feb 11, 2020 at 08:15 AM, <erik@...> wrote:
I use Teflon tape to isolate parts of the PCB such as under the sma center line. The PCB is solid copper
--
HBTE Files section:?/g/HBTE/files
Erik, PD0EK
|
Correction:? Those are inductors, not caps.
toggle quoted message
Show quoted text
On Tue, Feb 11, 2020 at 08:10 AM, Jerry Gaffke wrote:
I have had that center SMA pin try to rotate when connecting a cable,
so I try to leave a land of copper at an SMA to solder the pin to.
Otherwise those caps can break.
|
I use Teflon tape to isolate parts of the PCB such as under the sma center line. The PCB is solid copper -- HBTE Files section:? /g/HBTE/filesErik, PD0EK
|
Erik,
Your ugly style SAW filter board is actually quite pretty!
And if you are getting 70dB of stopband suppression, plenty effective.
Could work better than many attempts at an etched board.
An etched board that only has a groundplane on the back
would need lots of feedthroughs to connect ground points.
Feedthroughs have inductance, and that groundplane is 1.5mm further away.
(In this case, an etched board could have an almost solid ground plane on the front.)
I have had that center SMA pin try to rotate when connecting a cable,
so I try to leave a land of copper at an SMA to solder the pin to.
Otherwise those caps can break.
An alternative might be some 1 meg resistors in a 1206 package
from center pin to the ground pins,?adding strength there.
Sufficient grounding as you have done is important.
Not obvious, but you have left copper under each SAW filter
connecting ?the two halves of the ground plane, that's important.
I'd probably leave the entire ground plane intact except for
a small island on each end for the SMA pin.
I would have thought it might need a shield through the middle?
to isolate one SAW filter from the other, but 70dB is good enough.
Jerry, KE7ER
toggle quoted message
Show quoted text
On Tue, Feb 11, 2020 at 01:09 AM, <erik@...> wrote:
Maybe not as nice but building the filter dead bug style does deliver good performance.
The out of band rejection is at least 70dB, Combined with the SI4432 out of band rejection you get about 100dB which is sufficient.
|
I have made many a custom footprint in ExpressPCB, how hard could it
be?
You just make a pad, copy it a few times and then group it as a
component.
I do have Kicad, I do know it is almost impossible to get started
on...
But I see it has a component wizard.
73,
Gary
WB6OGD
On 2/11/2020 4:36 AM, Ion Petroianu,
VA3NOI wrote:
toggle quoted message
Show quoted text
I have received a message from Mouser telling me
that my request for QCC8C footprint has been discarded due to
insufficient information in the data sheet.
We are left with the only option of creating it
ourselves.
I will try to do it in Eagle and see what comes
out.
Ion
VA3NOI
On Mon, Feb 10, 2020, 12:21
Dave VE3LHO < dave@...> wrote:
The last
bits arrived last week and I'm working on my initial
implementation of the TinySA, building the filters "ugly"
style. I'm also considering making small PCBs for the filters,
both low pass and the IF bandpass filter.
I'm using the EPCOS B3555 434MHz SAW filter from the eBay
seller in Poland. It is in a QCC8C package and I've hit a
snag. I can't find an existing footprint for that package.
Does anyone know of/have a footprint for it for any hobbyist
available CAD package? My preference is kicad but others (e.g.
Eagle) would be fine.?
If no one has this I guess I'll need to consider learning how
create kicad footprints ... or maybe just live with ugly
construction :-)
Dave L.
--
Ion
VA3NOI
|
I have a footprint I made myself (KiCAD), but at work at the moment so no access to the PC.? I will put on the files section this evening if I remember.
73
Dave M0WID
|
I have received a message from Mouser telling me that my request for QCC8C footprint has been discarded due to insufficient information in the data sheet. We are left with the only option of creating it ourselves. I will try to do it in Eagle and see what comes out. Ion VA3NOI
On Mon, Feb 10, 2020, 12:21 Dave VE3LHO < dave@...> wrote: The last bits arrived last week and I'm working on my initial implementation of the TinySA, building the filters "ugly" style. I'm also considering making small PCBs for the filters, both low pass and the IF bandpass filter.
I'm using the EPCOS B3555 434MHz SAW filter from the eBay seller in Poland. It is in a QCC8C package and I've hit a snag. I can't find an existing footprint for that package. Does anyone know of/have a footprint for it for any hobbyist available CAD package? My preference is kicad but others (e.g. Eagle) would be fine.?
If no one has this I guess I'll need to consider learning how create kicad footprints ... or maybe just live with ugly construction :-)
Dave L.
--
Ion
VA3NOI
|
Maybe not as nice but building the filter dead bug style does deliver good performance. The out of band rejection is at least 70dB, Combined with the SI4432 out of band rejection you get about 100dB which is sufficient. -- HBTE Files section:? /g/HBTE/filesErik, PD0EK
|
On Mon, Feb 10, 2020 at 01:11 PM, Dave VE3LHO wrote:
A pretty good suggestion but unfortunately that SAW filter is in a different package (DCC6 versus QCC8).
Yeah, I didn't search through all his PCB's for the different components he offers... though know he has quite a few.? Not sure if any are? QCC8 package... though figured might be worth asking him.?
|
On Mon, Feb 10, 2020 at 12:04 PM, jafinch78 . wrote:
On Mon, Feb 10, 2020 at 09:21 AM, Dave VE3LHO wrote:
I'm using the EPCOS B3555 434MHz SAW filter from the eBay seller in Poland. It is in a QCC8C package and I've hit a snag. I can't find an existing footprint for that package. Does anyone know of/have a footprint for it for any hobbyist available CAD package?
.. stuff deleted ...
I'm thinking if you ask, he might share, since he may have the layout readily available assuming the package footprint is the same.?
A pretty good suggestion but unfortunately that SAW filter is in a different package (DCC6 versus QCC8).
|
On Mon, Feb 10, 2020 at 09:21 AM, Dave VE3LHO wrote:
I'm using the EPCOS B3555 434MHz SAW filter from the eBay seller in Poland. It is in a QCC8C package and I've hit a snag. I can't find an existing footprint for that package. Does anyone know of/have a footprint for it for any hobbyist available CAD package?
I made an order yesterday for:?
Since they already had pre-made PCB's and the size is so small.... just in case I ordered one:?
I asked about the performance of the Johanson Tech. 39nH inductors (?with this package per the B39431-B3740 datasheet and David (from rfextra) replied that he will be performing some testing with the SAW filter, PCB and inductor per the matching network spec and will post in the listing and reply.? I'm thinking if you ask, he might share, since he may have the layout readily available assuming the package footprint is the same.?
I think this is the last piece of hardware to make a like-for-like type design for comparison of performance of my work.? Being that this SAW listing has ten per order... I have a few to practice making my own PCB's with.? I still haven't practices soldering the 0603 size yet and was thinking about buying an inductor kit also to have a range of inductors.? Thinking will order the kit from mdflyelectronics to have with the resistor and capacitor 0603 kit for future experimenting with.?
I do have a roll of 0ohm resistors to practice with too before hand and see there are board practice kits for about a dollar also thinking won't hurt the budget... then again probably just as easy to desolder and resolder some salvage boards to practice on:?
?
?
|
Jerry,
No, I DO have it working fine, with 5V Nano clone (with level
shifters).
I do have a custom pin map (to make construction easier and not
conflict rx/tx).
Erik's code is fine.. except for a couple serial reads in the
command read section.
For instance the 'P' command has an extra " if
(Serial.available..."? that caused a problem,
I just commented it out.? It seems to me it should have been ok that
way but it wasn't.
? if(inData == 'P' || inData == 'p')
? {?
??? //if(Serial.available()>0) {
???? parameter = Serial.parseInt();
????
It does somehow have to do with running on 3V.? I did not change the
16MHz clock, that may
be the problem.? It reads and writes registers just fine in Serial
Monitor, it DOES run, so I conclude its some
kind of delay problem but???? Strange..
When I build my final version I will just use a 5V Nano with voltage
dividers...? I think the level shifter board
is overkill.
73,
Gary
WB6OGD
On 2/10/2020 11:01 AM, Jerry Gaffke via
Groups.Io wrote:
toggle quoted message
Show quoted text
Gary,
I assume that you do not yet have a hardware configuration for
which Erik's code works.
My code is still bit-banging the SPI port just like Erik's, but it
is using pins that
could use the Nano's SPI hardware interface if we ever decide to
turn it on.
Yes, Erik's pin choice conflicts with the serial port out of the
Nano,?could be your problem.?
He is not using a Nano.
> I did convert a Nano to 3V, mostly worked but I think it
needs a delay changed somewhere or something,
>? the noise level steps up and down by 10s of db erratically.
I assume that is with Erik's tinySA.ino and tinySA.exe.Is it still
using Erik's pin choices, but on the Nano?
Does it otherwise seem to work?
Have you ever had hardware up that was working better?
Remember, D0 and D1 are already used for serial communications.
Not sure how, but I suppose this could explain the erratic jumps
in noise level
if you are still using the pins Erik selected.
I doubt the symptom you report has much to do with running the
Nano at 3v.
I assume your Nano's processor clock is still at 16mhz, a bit out
of spec but should work fine.
If you have divided that Nano clock down to 8mhz, do try it at
16mhz to see if that cures your issue.
Jerry, KE7ER
On Mon, Feb 10, 2020 at 10:39 AM, wb6ogd wrote:
Jerry,
No, if it doesn't communicate in Serial Monitor, it won't work
with tinySA.exe, unless
I am mistaken.
Yes, I believe it should be "Init"? coming out.? Locking up
after 2 chars is the problem.? Getting
2 says its sort of working.
Oh, I didn't change to my pins..?? that could be it..? I think
you were going to use the SPI hardware
port?? I am not, Erik's code is bit bang on any pin and I relied
on that.
I had to make some small changes to Erik's serial read? but my
proto analyzer is working fine with that.
I did convert a Nano to 3V, mostly worked but I think it needs a
delay changed somewhere or something,
the noise level steps up and down by 10s of db erratically.
73,
Gary
WB6OGD
|
Ion Petroianu
jafinch78 .