Ext Antenna model STR-3 ??

[ArtT]

I have seen an antenna similar to the Gilsson for sale ...

Model STR - 3

Typical gain: 28db
Input voltage:2.5-12v
Current consumption: 5-11mA
Cable length: 3 metre
Connector: Angled MCX

Antenna is made in Taiwan

Anyone used one of these ? If so, how does it compare with the Gilsson.

Thanks

[jonasolof]

The STR3 is made by www.starsnav.com in Taiwan.

Specs are similar to the Gilsson, which , however, is claimed to have a three stage amplfier. It contains a voltage regulator that, like the Gilsson, reduces voltage between 2.5-12 V to the operational level. Now, the iQue or other low voltage gps receivers won't give you 2.5 volts under load, but that news hadn't reached Taiwanese antenna makers.

I'd buy the Gilsson because we know it works and the price is good.


The StR 3 is said to contain protective circuitry. But albeit the need for protection exoists in theory, there haven't been any problems reported.

The pdf with specifiactions show some interesting non encapsulated varietres. All contain an RF shield between the antenna element and the low noise amplifier. There is also a variety with a 3 inch flange/reflector, the first time I see a consumer grade antenna with a decent reflector in the design.

The pdf also contains a bklock diagram, first I have seen. But compare it to the diagram for the active 3 variety. That diagram shows two amplification stages of 16 dB each with a ceramic band pass filter in between whereas the first block diagram only shows one 16 dB amplification stage.

These diagram makes me think that maybe Gilsson counts the band pass filter as a stage since the amplfication is the same in all cases, nominal 30 dB minus some losses.

I bought a STR1 for 25 euros at the holux stand at Cebit. It is marked A-10302 Rikaline, another taiwanese GPS and antenna maker but has the STR1 capsule and power rating, so it is probably the same.

A quick test (awful weather) shows that amplification is about 3 dB higher with the Gilsson antenna. This might be due to a slightly lower power consumption due to less add on circuitry. Thus, the voltage wouldn't drop that much. I'll check better later.

[ArtT]

Thank you for the informative reply. I will source a Gilsson.

The reason why I was attracted to the STR was that it was available in Europe at a decent price. I have yet to find the Gilsson at a good price without ordering from USA.

Any sellers in UK or Germany/Holland/Belgium ?

[jonasolof]

ArtT,

Where do you find the STR3 in Europe? Would the same vendor have the other models?

Gilsson has worldwide delivery, which gives you best price. I choose US global priority mail, cost 9 USD, takes one week. I didn't have to pay any customs or VAT in Beglium. Later, i ordered 5 Gilsson antennas. That delivery cost 14 USD.

[Chrisq]

I got the Gilsson MCX via the web about a month ago.
It was delivered to the Netherlands within one week.
Shipment costs USD 7.95, which brought the total to 22 euros.

I can recommend it!

Chris

[ArtT]

That's sounds good. I am surprised they didn't charge more for shipping as I think the standard charge for US was USD 7.95

Pay Pal is not my favourite method mind you.

I assume you got it from Gilsson direct?

[jonasolof]

ArtT,

Who sells the STR-3 in Europe?

I see that gpsgeek now use Pay Pal. That is unfortunate, since it is a lot of hassle. Only time I used Pay Pal I didn't even receive the product I had paid for. Also, there are scams going on with Pay Pal. I try not to use it.

One can buy the Gilsson antenna over Yahoo also:

http://shop.store.yahoo.com/gilsson/products.html

I didn't go all the way through paying so I can't guarantee that PayPal isn't involved.

[GPSgeek]

Greetings,

We are familiar with STR-1 and STR-3 antennas from Stars Navigation in Taiwan. In fact, we used to sell their antennas until we received so many complaints hikers and drivers regarding its need for an external ground plane. For Garmin eMap and GPS 76 models, one would actually see a decrease in signal strength on the Satellite Info page, without attaching the antenna to an external ground plane. Further, we requested the makers of Stars Nav to optimize their STR-1 and STR-3 antennas for the Garmin GPS receivers, yet they still haven't done so for us. Hence, we stopped selling their STR-1 and STR-3 antennas. (Although we still have several hundred units in stock presently, we just could not recommend their antenna knowing our Gilsson antenna performs better.)

If one were to compare signal from both antennas (sitting side by side), one would clearly see the STR-1 and STR-3 antennas not performing as well as our Gilsson antennas. It is a guarantee.

Rikaline is not the manufacturer of the STR antennas. Rikaline simply reselling the Stars Nav antennas. I know both Mr. Andy Yang of Stars Navigation and Mr. Hsieh of Rikaline.

Kindest regards,
Ming Ho
Gilsson Technologies / GPSgeek
ming@gilsson.com
Toll Free: 877-GPS-GEEK

[GPSgeek]

Dear Jonas,

Good to "see" you again after my recent absence from the forum.

No, our Yahoo site does not use Paypal. One is free to charge the order onto Visa or MasterCard without opening an account with Yahoo nor Paypal.

Kindest regards,
Ming Ho
Gilsson Technologies

Originally posted by jonasolof
ArtT,


One can buy the Gilsson antenna over Yahoo also:

http://shop.store.yahoo.com/gilsson/products.html

I didn't go all the way through paying so I can't guarantee that PayPal isn't involved.

[bamb_i]

i been thinking about going in for the gilsson antenna.

one question pops to mind...does the lengthof the cord effects power drain or signal quality in any way?

[GPSgeek]

Originally posted by bamb_i
i been thinking about going in for the gilsson antenna.

one question pops to mind...does the lengthof the cord effects power drain or signal quality in any way?

There is a slight 2dB signal reduction between 9 and 16 feet cord lengths. There is no effect on the power draw between various lengths.

Kindest regards,
Ming Ho
Gilsson Technologies / GPSgeek
ming@gilsson.com
Toll Free: 877-GPS-GEEK

[kostik76]

Mounting external antenna... I am thinking of mounting points in my car and one that comes to my mind is under rear window. My concern is that it may pickup some interference from my speakers. What do you guys recommend? Also if I was to mount it outside, how well does it fair against elements?

[GPSgeek]

In responding to your inquiry, one of our staff members went outside and moved the antenna around the rear window and speakers, and saw no noticeable difference in performance. Generally, it is best to stay away from electrically noisy elements though. I would certainly place the antenna away from speakers, especially poorly shielded ones.

In terms of mounting the antenna outside the car, it would be fine. The linear amplifier circuit is located within a silicone-sealed chamber. We have seen our antennas mounted outside of cars and airplanes.

Kindest regards,
Ming Ho
Gilsson Technologies / GPSgeek
ming@gilsson.com
Toll Free: 877-GPS-GEEK

[PFflyer]

Running out of regulation, which will happen on most antennas with iQue, will result in some loss of gain, but that is not the major issue.

The important factor is how much tuning shift and noise figure degradation there is as the parametrics of the LNA device change due to the bias shift from the designed and tuned regulated voltage target.

Falling out of regulator input voltage limit also can cause shifts in performance over temperature because the LNA looses some temperature bias compensation control.

iQue need to have 6 or 7 mA pulled out of it to get the electronic switchover fully active to the external antenna port. Pull too little current and you get more loss in the iQue external port switch. If you pull more then 10 mA you drag the iQue output supply voltage below 2.0 vdc. So there is a narrow range of current and voltage for optimum iQue performance.

As to comments about ground plane, no magic here. The best performance needs at least a 4 inch ground plane. Since the patch tuning center is influenced by the amount of ground plane (or lack there of) you can tweek the best match (return loss) of the patch for insufficient ground plane which is better then having a mistuned patch and insufficient ground plane but there is no substitue for have sufficient ground plane. Poor return loss on the patch causes the noise figure of the amplifier to go up. Most the mag mount antennas assume you are putting on roof of car so they should be tuned with a good ground plane (car roof).

Hikers need to get a steel plate in their head.

[jonasolof]

Originally posted by PFflyer
.
------------------

iQue need to have 6 or 7 mA pulled out of it to get the electronic switchover fully active to the external antenna port. Pull too little current and you get more loss in the iQue external port switch. If you pull more then 10 mA you drag the iQue output supply voltage below 2.0 vdc. So there is a narrow range of current and voltage for optimum iQue performance.



My iQue is down to 1.99 volts already with the Gilsson drawing 5.55 mA.

How would you know that the activation of the external antenna port is gradual with increasing drain? Could you direct readers to some tech info?

When mentioning 3 dB down, this is actually not amplification but CN/o i.e. decodability as shown by Garmin's GPSinfo application which also gives sat signal "strength" in table form. So this figure would include all effects of signal degradation due to the voltage being too low.

I guess the best set up to try the effects of the various parameters is to use a bias-tee that lets you control the drain off teh iQue plus the feed voltage to the antenna. Wi-sys made me one likte that, but it is stuck with the tests in Amsterdam so I'll order another.

Re your comments on ground plane, I noted the remarks on the STR-3 antenna needing a ground plane in this thread earlier. I thought all patch antennas need a groundplane, because they would be designbed for it. Sam Storm van Leuven, who is doing the testing in Amsterdam already noted that some pre amplified antennas are detuned.

This is from a brief preliminary report from Sam:

"1. I measured the zenith gain as a function of power supply voltage: first
I looked at the gain and current at nominal voltage (mostly 5.0VDC). Next I
lowered voltage until the gain was 3dB down, and noted current again.
I did this for all your and mine antennas, 16 alltogether! For gain
measurement
I used my 'very poor mans spectrum analyzer', and noted that the
center frequency is often off by several MHz ! Also the -3dB frequency
bandwidth
shows some variation.
I attached a typical spectrum analyzer plot (one of my Big Brothers). The
analyzer output is in dBuV, I still need to convert to calibrated dBm.

2. I tried to measure absolute zenith gain for each antenna, but I am not
satisfied
yet, because the linear polarized signal to the antenna causes a large
rotation assymetry in most patches (the helix should be superior in this
respect). So I'll repeat the test using a circular ploarized signal to the
antennas.

3. I plan to compare the antennas tracking performance under foliage this
weekend. Hope the weather stays OK."


Attached should be a plot of a Big Brother antenna

[PFflyer]

I am pretty sure there is a PIN diode in series with external jack because of the obvious diode drop as you start to d.c. load the jack. Depending on type of PIN diode you need a certain about of forward bias current to get its series resistance down. Lower current turn on PIN's are available with cost and shunt capacitance trade off but if LNA current for GA27 is 5 to 7 mA you would expect the PIN diode selection would be optimized for that bias range.

The bars you see on iQue signal strength are not purely RF power indication but CDMA code correlation power/background noise power. Adding more gain to external antenna path will not necessarily make the bar go higher.

If you use a bias-T keep in mind that you need to pull some d.c. load on iQue jack. Would be interesting to find out how much current is needed to be pulled to get the input loss at acceptable level. You probably need to use a Patch without an active amp to represent the loss accurately.

[jonasolof]

Thanks, very good points.

It is easy to find a bare patch for testing, examples:

http://www.sanav.com/gps_antennas/pa-series.htm. I'll order one from Sanav and bring up the topic of drain influence on the pi diode circuit. .

Until now, I haven't been aware that too little drain could be a disadvantage. In fact, this makes optimizing the design of preamplified antennas even more important. It is also, as you mentioned, highly probable that Garmin designed the external antenna port for it's own 27C antenna. As Sam mentioned, it's amplification doesn't change even well below 2 volts (although in absolute terms, the Gilson works better).

I believe that there is a group of GPS receivers that function with pi-diodes: Garmin 76, Emap, possibly the 60 C, and the iQue. Probably also Magellan Meridian Colour. That should be enough for creating a market for optimized design (1.8-2.2volt working voltage, drain 5-7 mA.)

I have an impression that some other receivers don't work the same way. My emtac bluetooth GPS yields 3.6 volts at the external antenna connector unloaded. Thus it is possible that it is fed directly from the LiIon battery.

Finally, by adjusting the drain through a bias-tee to the same level as the load imposed directly by the preamplfied antenna, it should be possible to measure the insertion loss of the bias-tee.

It just goes to show how little data is public for consumer goods that users need to experiment to find out how the stuff they bought actually works. If the lack of manufacturer-provided tech data leads to suboptimized accessories it's even worse.

[PFflyer]

Think about it a little more there may be a way to get a feel for minimum draw current to get minimum PIN diode loss effect.

With bias-T's (you need to build special or use two in series to accomplish both directions bias feeds) and a step attenuator between the bias-T's, setup for good full reception. 5-8 mA draw from iQue, proper spec level voltage for external antenna.

Observe signal strength bars level on iQue, increase step attenuation in external antenna path until bars just start to drop. This is point when there is not sufficient gain to take over iQue noise figure.

You should now be sensitive to additional loss due to insufficient PIN diode bias. Start to back down the iQue d.c. load draw measuring current. Bars should stay constant until you reach the bias point where the PIN diode bias starts to be insufficient. This is the spec we are looking for.

[jonasolof]

Good setup. A step attenuator for 1.5 GHz isn't on my shopping list so I'll forward the suggestion to Sam Storm van Leuven. A second modified (dc-isolated variable drain, variable feed) bias-tee should be OK to get. It might also be possible to have Wi-sys interested enough in the problem that they could contemplate building one designed to fit with a specific attenuator.

For anyone inclined, here is an offer on e-bay (http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3805532253&category=36316)

More specs: http://www.alanindustries.com/catalog/P16.htm

Attenuation range? 10 DB seems little, 70 dB much.

[jonasolof]

The test scheme as laid out by PFflyer above only works if the implicit preconditions of the following statement exist:

"Observe signal strength bars level on iQue, increase step attenuation in external antenna path until bars just start to drop. This is point when there is not sufficient gain to take over iQue noise figure".

What if the noise level of the ique is influencing decoding all the way, so that no cut off level can be established? Or is the fact that one hardly ever sees a CN/o reading > 52 dB indicating that "perfect" decodability has been reached? It could just as well be that 52 dB is the cut off level for measuring, but that signal quality will still improve beyond that figure.

Since I lack the knowledge of how the CN/o readings are made in detail, my only analogy is with Orinoco client software that I have for Wifi cards. It will tell rate of packet loss, which easily goes down to zero under good conditions. But here a conventional S/N ratio is still given, which can improve even beyond the signal level where there is zero packet loss.

I understand, that with the CN/o readings in GPS, it is only the decodeability which is measured.

BTW, there are two ways to get dB readings from the iQue. One is GPSinfo, Garmin's test app. The other is with IQbooster and a serial cable to a PC running VisualGPS. It shows signal bars and dB readings at the same time, so I find it easier to use.

[PFflyer]

The bars are reading individual satellite code power just like a CDMA cellphone. There is an AGC system (automatic gain control) that, for GPS, primary kicks in when you have an amplified external antenna. The object of 'starving' the gain in the measurement is to get to point where analog control part of the AGC is out of the picture.

I am guessing on iQue hookup again, but the wrong way to do external antenna system with internal antenna is internal patch ant. > switch > LNA/receiver input. The right way is internal patch > internal LNA > switch-> receiver input.

You don't want the switch between internal antenna and LNA on a already challenged system which would add another 1.5 to 2 db loss between patch antenna and internal LNA. (challenged, meaning patch with insufficient ground plane and likely some noise pickup by the close by flip patch to all the digital noise generation from the iQue).

If it is hooked up the 'right way' then the noise figure of the external port is realitively poor requiring some gain between an external antenna and external input jack.

I am afraid if you put just an antenna patch without some amplification to the external port you might not pick up satellites very well.

[jonasolof]

PFflyer,

So what is the range of attenuation that you'd recommend for testing?

[PFflyer]

That will depend on how much gain your external antenna has. I would guess you need to have the ability to wipe out all the external gain down to about 5 or 6 db net gain. So if you got a 28 db gain ext antenna you would need about 10 to 20 db of attenuation assuming you are going to lose 5-6 db in the RG174 cable length.

You need to have some tweeking range. 0 to 10 db step attn with 10 or 20 db fixed pad should do.