Episode 1439 I got some PC boards off eBay. Let's see how they do Be a Patron: / imsaiguy
ความคิดเห็น • 63
ปีที่แล้ว +16
Thanks for the video. The coax you used is called "Conformable', not "semi-rigid". Semi-rigid is much stiffer, as it uses a solid copper tube as the shield. Also, be careful with metal objects on your table when testing!
Every single antenna I've bought from China on ebay has had some serious flaw. Either the marketeers are liars, or the chinese designers are clueless. Also, a note for some viewers: Although they look similar, a log-periodic is VERY different than a Yagi-Uda. Logs typically have only moderate forward gain, high front-to-back ratios, and wide bandwidth. Yagis generally have higher forward gain, moderate front-to-back ratio and narrow (about 2%) bandwidth.
Would that be possible to cut some lines to march the frequencies? I think if you shorten the line the frequencies will be shifted upward. I'm by no means an antennas guy.
@@IMSAIGuy It's possible to build log antennas that achieve a good match and have some gain over a wide frequency range. But your tests show that one to be somewhat unimpressive.
A cool experiment. It will be fun to see one antenna transmitting and the other antenna receiving. PS - When I went to college I had a teacher who said "A radio with no antenna is like a dog with no ears". Those antennas would look cool on a robot dog :-)
Don’t you need to get that in the clear by at least one wavelength all around so the near field isn’t affected by the workbench, tools on the bench, the VNA, etc?
I don't think the coax should be soldered to the board this way. Its shielding should be isolated from the PCB, the only place where it should be connected to the PCB is the tip of that antenna, because the shielding running parallel to the trace on that PCB forms a balun. The input of any LPA is symetric, but the feeding coaxial line not. In my opinion, that coax should be wrapped into some insulation like kapton tape, or should be fed through some silicon tube or things like that, which should be somehow glued to the PCB to keep things parallel to form the required balun. If you connect it like you've done, it doesn't form the balun, which is not a big issue on such a high frequencies, but it detunes the whole antenna by adding some material to it, which is exactly what you've been observing on the VNA.
Bought three of these from the original designer at Hamvention many years ago. They work from around 900 MHz to 2.5 GHz. Never had any issues with the authentic antennas. They worked great on the Ericsson G30 / 35 fixed wireless terminal for GSM networks. Went from marginal coverage to full indication of coverage when we aimed the antennas towards a cell site. I also use these antennas for directional base coverage of the Marconi Bowman PRR squad radios. Get twice the range between fixed points.
Great work. In addition to the return loss which you measure, it would be very interesting to measure the gain between two of the same antennas. Naturally, all the standard considerations and standards need to be taken, and the interpretation of "gain" follow the same engineering standards. I'm pretty certain you understand those standards. Thanks for your good posts.
The coax insulation is more than likely PTFE (i.e. Teflon), it's rather good over significant temperature ranges - both high and low. Low loss too, which is good for RF.
@@uni-byte You're welcome! I've used it extensively in the lab, really easy to shape, nice for short for cables, making RF transformers, etc. 100% shield too. Sort of my 'go to' coax for building prototypes. I should clarify one of my comments tho: While Teflon itself is low loss, the small diameter of the coax does hinder performance somewhat.
Um, Take some tin snips & cut down the outside edges of the diagonals? I would also reduce the angle of the slope to narrow the bandwidth to something more approaching WiFi band coverage. Alternatively one go the opposite direction & could solder some solid copper wires on to lower the frequency towards the lower frequency cellular data bands
You should not solder the coax shield along the line like that, it’s part of a transmission line and doing it like that will mess up the impedance significantly.
I’m gonna try to find the time to try what I’m about to ask. But, if you’ve already tried it , I’d love to know the results… Q: What would the response be and how would characteristics change if a similar Log Periodic was made on flat material, laid out as you did but the rolled into a cylindrical body. I’ve come across designs that are similar to what I described but has a huge dip at around 750mhz that hold a flat line parallel to the X axis then climbing back up at around 3.2ghz. That’s gotta be fake right? Such a broad band response that is so clean seems like someone is trying to run a con. If you’re able to make the one I described, I’d love to see/ hear the outcome. Thanks , -Jason Burchell
These are HIGHY directional... so in "rolling" them up would defeat the purpose of such a design...what you are looking for is a helical antenna(or not...just the closest thing to any practicality from such a design)- also, as you see, these are highly specific to frequency....the measurements are carefully adhered to as to not stray off course. Your antenna rollup would have to be adjusted since the frequency unrolled would differ from rolled....Helical antenas are in some cases pretty much a spring- many stock antenas for handhelds like mobile communication + CB radio, if you cut those plastic nubs of antenas in half, you'll find the cheapo coiled of metal. Other thana wire, these are some of the simplest antenna design, and cheapest- why you see em in so many cheapo gizmos and Baofangs....not the crem del la crem of radiating patterns...,but practical... pretty much a donut shape radiating out if you hold the antenna up. Certain designs can make for some phalic radiating patterns- placing 4 helical antenas in a 2x2 grid/ array will provide you with such a beauty of signal- out the front you'll find a girthy directional strong signal and just to the sides and back, 1 on each side, circular radiating signal. This arrangement is wicked for FPV where longer distances are sought- this and an antenna tracker and you are golden pony boy. If this isn't what you'd been thinking, then myself, and 100% i am sure the community at large would love for you to share your adventures with the fruit rollup yagi antenna. Please do.
Would there be some kind of defect / mis-design / detail that would cause it to be "under" by 100hz? So if you added 100 to 1100, you'd get 1200 and then the doubling would be 2.4ghz?
Nice LP Antenna NanoVNA measurements !!....Pretty good chance your Bonding the Coax Shield to the Boom pc trace is fouling up the Expected bandwidth impedance measurements......End Fed LP's with significant "Drooped" supported coax feed line seem to "Wide band" mot of the time....Measuring a Non-Bonded coax feed line would be a Very interesting test.......Thx for the video......
no, it was designed with the coax soldered down, the backside trace had to be thickened in a revision to offset things. the antenna was designed that way from the start. I agree it is quite odd, but you will find many copies of that design in many products now.
Still, ....Very Interesting antenna design Dimensions for something Labeled 2400 Mhz, which one would Assume is the Center design frequency.....The Dimensions would suggest something closer to 1100 Mhz to 2400 Mhz Bandwidth range.....That would put your Sellected 1725 Mhz freq just about mid band.......Other than wide band operation, why not use a higher gain Yagi design ??.....
Is that NanoVNA-F V2 more accurate than the earlier ones? I do agree with others on here though, that these Chinese antennas rarely are the frequency they advertise. Wondering if measuring Chinese antennas with Chinese VNA's compounds the inaccuracies? 🤔
Probably a stupid question, but are you sure the coax should be soldered along the length? It seems as if that makes the feedpoint be at one end for the top and the other end for the bottom (of the PCB).
@@IMSAIGuy I use a very old Kester for RF soldering but don’t get quite consistent response for my solder maybe it’s time to get a new one. One more thing while soldering the center pin of SMA I normally fill it with solder and then push in the center conductor. I noticed you do it differently I am sure there must be some advantage to your method can you please explain. Thank you
Thanks for the video. The coax you used is called "Conformable', not "semi-rigid". Semi-rigid is much stiffer, as it uses a solid copper tube as the shield. Also, be careful with metal objects on your table when testing!
Every single antenna I've bought from China on ebay has had some serious flaw. Either the marketeers are liars, or the chinese designers are clueless. Also, a note for some viewers: Although they look similar, a log-periodic is VERY different than a Yagi-Uda. Logs typically have only moderate forward gain, high front-to-back ratios, and wide bandwidth. Yagis generally have higher forward gain, moderate front-to-back ratio and narrow (about 2%) bandwidth.
vk4ghz.com/2-11-ghz-wa5vjb-log-periodic/
Would that be possible to cut some lines to march the frequencies? I think if you shorten the line the frequencies will be shifted upward. I'm by no means an antennas guy.
@@IMSAIGuy It's possible to build log antennas that achieve a good match and have some gain over a wide frequency range. But your tests show that one to be somewhat unimpressive.
@@chrisscott1547 not enough elements for that frequency range
A cool experiment. It will be fun to see one antenna transmitting and the other antenna receiving. PS - When I went to college I had a teacher who said "A radio with no antenna is like a dog with no ears". Those antennas would look cool on a robot dog :-)
tomorrow
@@IMSAIGuy Cool. Thanks :-)
Don’t you need to get that in the clear by at least one wavelength all around so the near field isn’t affected by the workbench, tools on the bench, the VNA, etc?
Life hack: it is convenient to solder the central pin of the coax connector conclusion using soldering paste.
I don't think the coax should be soldered to the board this way. Its shielding should be isolated from the PCB, the only place where it should be connected to the PCB is the tip of that antenna, because the shielding running parallel to the trace on that PCB forms a balun. The input of any LPA is symetric, but the feeding coaxial line not. In my opinion, that coax should be wrapped into some insulation like kapton tape, or should be fed through some silicon tube or things like that, which should be somehow glued to the PCB to keep things parallel to form the required balun. If you connect it like you've done, it doesn't form the balun, which is not a big issue on such a high frequencies, but it detunes the whole antenna by adding some material to it, which is exactly what you've been observing on the VNA.
vk4ghz.com/2-11-ghz-wa5vjb-log-periodic/
Bought three of these from the original designer at Hamvention many years ago. They work from around 900 MHz to 2.5 GHz. Never had any issues with the authentic antennas. They worked great on the Ericsson G30 / 35 fixed wireless terminal for GSM networks. Went from marginal coverage to full indication of coverage when we aimed the antennas towards a cell site. I also use these antennas for directional base coverage of the Marconi Bowman PRR squad radios. Get twice the range between fixed points.
It would be interesting connecting up both antennas to the VNA and pointing them at each other, Look at S21.
tomorrow
Great work. In addition to the return loss which you measure, it would be very interesting to measure the gain between two of the same antennas. Naturally, all the standard considerations and standards need to be taken, and the interpretation of "gain" follow the same engineering standards. I'm pretty certain you understand those standards. Thanks for your good posts.
I'm surprised all that soldering does not compromise the insulator.
The coax insulation is more than likely PTFE (i.e. Teflon), it's rather good over significant temperature ranges - both high and low. Low loss too, which is good for RF.
@@Mark-hb5zf Thanks Mark. I've had little to do with this kind of coax in my travels so it's good to be able to learn something from time to time
@@uni-byte You're welcome! I've used it extensively in the lab, really easy to shape, nice for short for cables, making RF transformers, etc. 100% shield too. Sort of my 'go to' coax for building prototypes. I should clarify one of my comments tho: While Teflon itself is low loss, the small diameter of the coax does hinder performance somewhat.
I would trim each trace down to raise the frequency.
Um, Take some tin snips & cut down the outside edges of the diagonals? I would also reduce the angle of the slope to narrow the bandwidth to something more approaching WiFi band coverage. Alternatively one go the opposite direction & could solder some solid copper wires on to lower the frequency towards the lower frequency cellular data bands
You should not solder the coax shield along the line like that, it’s part of a transmission line and doing it like that will mess up the impedance significantly.
vk4ghz.com/2-11-ghz-wa5vjb-log-periodic/
@@IMSAIGuy yeah, just because he does it like that doesn't make it right...
Looks like a Yagi Uda antenna on a PCB.
NICELY DONE WITH JUST ENOUGH OF COMMENTS TOO.
I’m gonna try to find the time to try what I’m about to ask. But, if you’ve already tried it , I’d love to know the results… Q: What would the response be and how would characteristics change if a similar Log Periodic was made on flat material, laid out as you did but the rolled into a cylindrical body. I’ve come across designs that are similar to what I described but has a huge dip at around 750mhz that hold a flat line parallel to the X axis then climbing back up at around 3.2ghz. That’s gotta be fake right? Such a broad band response that is so clean seems like someone is trying to run a con. If you’re able to make the one I described, I’d love to see/ hear the outcome. Thanks , -Jason Burchell
These are HIGHY directional... so in "rolling" them up would defeat the purpose of such a design...what you are looking for is a helical antenna(or not...just the closest thing to any practicality from such a design)- also, as you see, these are highly specific to frequency....the measurements are carefully adhered to as to not stray off course. Your antenna rollup would have to be adjusted since the frequency unrolled would differ from rolled....Helical antenas are in some cases pretty much a spring- many stock antenas for handhelds like mobile communication + CB radio, if you cut those plastic nubs of antenas in half, you'll find the cheapo coiled of metal. Other thana wire, these are some of the simplest antenna design, and cheapest- why you see em in so many cheapo gizmos and Baofangs....not the crem del la crem of radiating patterns...,but practical... pretty much a donut shape radiating out if you hold the antenna up. Certain designs can make for some phalic radiating patterns- placing 4 helical antenas in a 2x2 grid/ array will provide you with such a beauty of signal- out the front you'll find a girthy directional strong signal and just to the sides and back, 1 on each side, circular radiating signal. This arrangement is wicked for FPV where longer distances are sought- this and an antenna tracker and you are golden pony boy.
If this isn't what you'd been thinking, then myself, and 100% i am sure the community at large would love for you to share your adventures with the fruit rollup yagi antenna. Please do.
Would there be some kind of defect / mis-design / detail that would cause it to be "under" by 100hz? So if you added 100 to 1100, you'd get 1200 and then the doubling would be 2.4ghz?
If only life was so simple....
I have played with a 6 element Yagi on 2.4 Ghz .
Very simple & cheap with good gain .
Fussy about it's mounting & stability 🤣
Nice LP Antenna NanoVNA measurements !!....Pretty good chance your Bonding the Coax Shield to the Boom pc trace is fouling up the Expected bandwidth impedance measurements......End Fed LP's with significant "Drooped" supported coax feed line seem to "Wide band" mot of the time....Measuring a Non-Bonded coax feed line would be a Very interesting test.......Thx for the video......
no, it was designed with the coax soldered down, the backside trace had to be thickened in a revision to offset things. the antenna was designed that way from the start. I agree it is quite odd, but you will find many copies of that design in many products now.
vk4ghz.com/2-11-ghz-wa5vjb-log-periodic/
Can you tell us the coax that you are using? I would assume a teflon dielectric would be needed to withstand the braid soldering without deformation.
The coax he used is called "Conformable', not "semi-rigid".
I bet you a nickle if you trimmed off some of the traces *shortening them, you could move around the peak frequency response... I know its a gamble.
Still, ....Very Interesting antenna design Dimensions for something Labeled 2400 Mhz, which one would Assume is the Center design frequency.....The Dimensions would suggest something closer to 1100 Mhz to 2400 Mhz Bandwidth range.....That would put your Sellected 1725 Mhz freq just about mid band.......Other than wide band operation, why not use a higher gain Yagi design ??.....
Bit of a joke that one, even if it had a dip at 2450 it would suck in all the cell noise on 2100 etc
CABLE ??? Can anyone please point out to the place where I can get this coax cable ? And what's the name of it ? Thanks in advance!
I'm not sure as the cable I have came out of the some equipment. you might search for 'Semi-rigid Coax Cable with Tinned Braid Outer Conductor'
Would a gamma match help improve s11 @ 2.4GHz ?
Is that NanoVNA-F V2 more accurate than the earlier ones? I do agree with others on here though, that these Chinese antennas rarely are the frequency they advertise. Wondering if measuring Chinese antennas with Chinese VNA's compounds the inaccuracies? 🤔
the old vna was 900MHz, this one runs to 4.4GHz. the frequency is accurate on all.
0:30 you mean 2.4 GHz not MHz, right?
Could you make an appropriate pcb layout for 2400 and add to PCBwyy afterwards?
Many already made up for you on Aliexpress....
Probably a stupid question, but are you sure the coax should be soldered along the length? It seems as if that makes the feedpoint be at one end for the top and the other end for the bottom (of the PCB).
yes, I'm basing that on looking at other products.
@@IMSAIGuy Thanks!
vk4ghz.com/2-11-ghz-wa5vjb-log-periodic/
Looks like it's based off WA5VJB's original designs.
yes
I was wonder which soldering wire do you use and do you have a recommendation for a Lead Free Soldering wire for RF application :)
Metcal Mx-500
I hate all lead free, I use Kester
@@IMSAIGuy I use a very old Kester for RF soldering but don’t get quite consistent response for my solder maybe it’s time to get a new one. One more thing while soldering the center pin of SMA I normally fill it with solder and then push in the center conductor. I noticed you do it differently I am sure there must be some advantage to your method can you please explain. Thank you
@@ahmedshafqat1409 if you solder one side first, there will be no flux for the second item.
They look like cheap knockoffs of the WA5VJB antennas complete with random giant metal lettering to mess with the nice tidy log periodic design.
Hm, that doesn't look right. It measures more like a shorted coax than an actual antenna.
Log Periodic antennas operate poorly over a wide range of frequencies. Immutable fact.