how-to block ads
|
|
Share Topic
 |
 |
|
|
|
| reply to grohgreg
Re: Minimum cable length said by grohgreg :
Since a transmitter with 62dB gain would be about the size of Delaware
In the Ku-band world, BUCs (Block UpConverters, the equivalent of the transmit portion of the TRIA) run about 50 to 55 dB of Gain. I figure that WB has a slightly higher gain, since the modem output levels are lower than the Ku-band version.
The transmitter in the TRIA is a fixed gain device, with a maximum output power of (let's assume) 4 watts. Converting to dB, 4 W = 6 dBW = 36 dBm. If we assume 62 dB gain, this means that to drive the transmitter to saturation we need (36-62=) -26 dBm input to the transmitter. If we need only 1 watt (0 dBW or 30 dBm) out of the transmitter, the input level would be -32 dBm. So the transmit power is controlled directly by the output level from the modem.
Note that there is an unknown amount of cable loss between the modem and the transmitter, and manufacturing variations in the trasmitter gain, not to mention beam coverage and dish pointing variations. This is why there's a feedback system from the SMTS to continually adjust the power from the modem. The trick is getting everything engineered so that you are never at the end of the adjustment range.
said by grohgreg :
And if DC voltage is "not a control mechanism", why would transmit power be "adversely affected"?
The Transmiiter needs to be able to draw enough power to produce it's 4 watts. If the voltage is too low (or the cable is too long) it won't be able to hit its max output. | |
| said by joebob42 :
The Transmiiter needs to be able to draw enough power to produce it's 4 watts. If the voltage is too low (or the cable is too long) it won't be able to hit its max output. So do I get this, WB didn't figure on this problem in the design of the modems and Gateway equipment?...and, just going with a total of 75'-150' of cable won't solve everyone's issues with variable transmits...in some locatons, regardless of cable length, an attenuator will be required? | |
| said by randyvsatus :
and, just going with a total of 75'-150' of cable won't solve everyone's issues with variable transmits...
WB didn't initially specify a minimum cable length. I think they were caught by the system performing better than expected.
said by randyvsatus :
...in some locatons, regardless of cable length, an attenuator will be required?
I think that by specifing a minimum, they won't ever need an attenuator. | |
| said by joebob42 : said by randyvsatus :
...in some locatons, regardless of cable length, an attenuator will be required?
I think that by specifing a minimum, they won't ever need an attenuator. There actually have been some posts here and on the other forum that attenuators have already been shipped to some customers and WB has given them instructions to connect it between modem and transmit cable....might this be in lieu of a service call to add cable length and the costs associated therewith? Don't know the cost of an attenuator, but a truck roll is a good $125 raw cost for one hour...maybe more. | |
| said by randyvsatus :
Don't know the cost of an attenuator...
It looks like the attenuators are about $10... certanly cheaper than a truck roll. WB is probably using these to fix any existing problem users (which they can easily identify from stats at the SMTS.) Future problems are eliminated by the minimum cable length.
And, I did learn from this that there are inexpensive L-band attenuators that pass DC... I didn't think that they existed. (What I've seen previously cost over $100...) | |
 grohgregDunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
4 edits | reply to joebob42
said by joebob42 :
The transmitter in the TRIA is a fixed gain device, with a maximum output power of (let's assume) 4 watts. Converting to dB, 4 W = 6 dBW = 36 dBm. If we assume 62 dB gain, this means that to drive the transmitter to saturation we need (36-62=) -26 dBm input to the transmitter. If we need only 1 watt (0 dBW or 30 dBm) out of the transmitter, the input level would be -32 dBm. So the transmit power is controlled directly by the output level from the modem.
OK. It's becoming clear that somebody's feeding at least part of this stuff to you, since till now you didn't seem to realize that there is a thousandfold difference between a milliwatt (dBm) and a watt (dBW). In the professional world, dBm is usually specified. When an unqualified "dB" is used, it's usually assumed dBW. Hence, a 62dB transmitter the size of Delaware. The SurfBeam specs list the modem input/output in dBm. To avoid confusion you must carry the same unit of measurement forward in the discussion - and in the calculations. Forget about the 4 watts, and start with the modem; known output -35 to -5 dBm. Then figure 6.3dB loss over 75' of CS5729. Then add your 62dB "transmitter gain", then add the known antenna G/T of 15.4
-35 - 6.3 + 62 + 15.4 = 36.1 dBm or ~4 watts (EIRP). Discounting the antenna gain, that's about 0.117 watt. So far so good.
-5 - 6.3 + 62 + 15.4 = 66.1 dBm or 4074 watts (EIRP). Discounting antenna gain, that's still ~117 watts out of the transmitter.
See the problem I have with your 62 db gain figure?
said by joebob42 :
The Transmiiter needs to be able to draw enough power to produce it's 4 watts. If the voltage is too low (or the cable is too long) it won't be able to hit its max output.
Nor do you seem to have a real good handle on the difference between voltage and current. When you say "draw enough power", that means the transmitter drawing current from the modem. Consider voltage as little more than the delivery vehicle for current. But - voltage changes CAN be used as the trigger to change polarity, change frequency, change output power. Hence the confusion over fixed versus ramped.
//greg//
--
DW4000CE/1.2 Ghz Tualatin/512 PC133 - W2K Pro/SP4 - SRS G11/1370H - RSL83/ACP81 - v4.2.1.10C - RWIN 513920/MTU 1500 - proxy switch - Gateway/DNS 205.177.62.91/66.82.4.8 plus rollovers - Firefox 1.02 | | |
|
| said by grohgreg :
When an unqualified "dB" is used, it's usually assumed dBW.
I'm sorry, I'll have to disagree with you here. In 25 years of satellite engineering, I've never seen this assumption. If it's "dB", then it's either gain/loss, or a ratio (like C/N). Otherwise, a 10 dB attenuator would reduce the signal by 10 Watts, which doesn't make sense... I would be interested if you could point me to an example of this usage. And I certanly know the difference between dBm and dBW (it's 30 dB  )
said by grohgreg :
Nor do you seem to have a real good handle on the difference between voltage and current.
I've got a very clear handle on this; though I did blur the destinction in my response. Sorry for that.
The key is that the voltage is not used to control the TRIA in any way. | |
grohgregDunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY | said by joebob42 :
I would be interested if you could point me to an example of this usage. No problem: see »www.atis.org/tg2k/_dbw.html. Note the definition; "dB referenced to one watt". So when you say "dB", I see an all-encompassing term. Generic, if you will. True, "dB" expresses a "ratio", but relative to what? It could be dBA, dBC, dBV, dBm, dBi, dB0, dBW et cetera.
To see what I'm talking about - the difference between a simple "dB", and the specific "dBm" - see »www.isa.org/Content/ContentGroup···_dBm.htm. dBm is specific to miliwatts. When you're talking negative power levels (-5 to -35 for example), it's absolutely imperative to keep the discussion in dBm.
//greg//
--
DW4000CE/1.2 Ghz Tualatin/512 PC133 - W2K Pro/SP4 - SRS G11/1370H - RSL83/ACP81 - v4.2.1.10C - RWIN 513920/MTU 1500 - proxy switch - Gateway/DNS 205.177.62.91/66.82.4.8 plus rollovers - Firefox 1.02 | |
| reply to grohgreg
said by grohgreg :
Then add your 62dB "transmitter gain", then add the known antenna G/T of 15.4
I'm sorry, it looks like it's my turn to call this gibberish.
First off, G/T is a measure of receive performance, you can't add it to the transmit side like you're trying to do.
Second, the 3 watt spec (sorry, I had the 4 watts wrong) is for the output of the transmitter, and has nothing to do with the antenna it's connected to. The Transciever spec (referenced earlier) says that the typical maximum output power (P1dB) is 34.8 dBm, or 4.8 dBW, or 3 watts. So your examples become:
-35 dBm - 6.3 + 62 = 20.7 dBm, or about a 10th of a watt.
-5 dBm - 6.3 + 62 = 50.7 dBm, which is limited to 34.8 dBm by the transmitter capabilities.
Now, if you want to talk about EIRP, figure the transmit gain is about 45 dBi. (Again, I haven't seen the specs for WB's particualr dish, 45 is a spec for a Patriot Ka band dish of similar size.) This is added to the above results, and do give what seem rediculously large numbers. But, EIRP is how much power you would need if you had an isotropic antenna (non-directional), not a dish.
Does the 62 dB figure make more sense now? (Actually 48 to 63, as it turns out I was picing the high end.) | |
| reply to grohgreg
said by grohgreg :
Note the definition; "dB referenced to one watt".
This is the definition for dBW. dB is always a ratio. Nothing here says that "dB" can be assumed to be dBW. If you're talking about the gain of a device, it's gain in dB is derived from the ratio of its output power to its input power, at its normal operating point.
Your second reference states quite clearly: Use dB when expressing the ratio between two power values. Use dBm when expressing an absolute value of power. My usage is consistant with that. So, I still don't see why you assume when I write dB, I really mean dBW. | |
| reply to joebob42
said by joebob42 :
figure the transmit gain is about 45 dBi.
To correct myself, this should be 39.5 dBi for the WB dish. | |
Mr_DNS
join:2005-09-07
Laredo, TX | reply to randyvsatus
said by randyvsatus:So do I get this, WB didn't figure on this problem in the design of the modems and Gateway equipment?...and, just going with a total of 75'-150' of cable won't solve everyone's issues with variable transmits...in some locatons, regardless of cable length, an attenuator will be required? .
.
Let's backup for a minute. First, if you look at both of the beam coverage maps that exist out there, you will see the first one shows circles, whereas the second one is more defined for beam coverage. My conclusion is these maps were drawn by artists for conceptual purposes...hence, they are only ballparks...not the reality.
.
Beam EIRP Contours - Keep in mind in certain geographical areas, they have more cloud cover [rain fade]; consequently, the engineers would have to account for expected user uplink/downlink power requirements, as a design baseline....this is a no brainer...the weather.
.
Now, back to cable issue, where somewhere around 150' was the max for RG6. Someone stated, in the beginning, there was only a max length restriction...so, let's do a cable calc quickly.
Let's pretend the TX freq on the cable is 1500 MHz, with 75' cable you have about 5.4 dB loss and at 150' you have about 10.8 dB loss, with a difference of 5.4 dB....
Now, I spoke with a Sat Guru recently about those beam circles, and he estimated in ballparks around 4dB loss from beam center to perimeter.
Hence, if you look at
1. what is "required" now [75'-150'],
2. User's short cable has a hot signal
3. 4 dB loss when user on beam perimeter
4. 75' cable has about 5.4 dB loss
Then, you can conclude for those in the beam's center must have 75' min cable length, and those on the perimeter can have short cable runs since about 4 dB signal loss is similiar to about 55' cable....so even 10' of cable run would be OK, if you lived on the perimeter area...
| |
| said by Mr_DNS:...Then, you can conclude for those in the beam's center must have 75' min cable length, and those on the perimeter can have short cable runs since about 4 dB signal loss is similiar to about 55' cable....so even 10' of cable run would be OK, if you lived on the perimeter area... Makes sense to me....and sending $10 attenuators (if that is the WB cost) to those running "too hot" (short runs in the center of beam) also makes sense - lot's cheaper than a truck roll. My question then becomes why running at say 3 watts transmit (assuming that is the max) all of the time cause a tria to crap out...or is that really why some are going south on people - it should have no problem running at max power for 100,000 hrs like the specs say...right? | |
grohgregDunno. Ask The Chief
join:2001-07-05
Dawson Springs, KY
2 edits | reply to joebob42
said by joebob42 :
First off, G/T is a measure of receive performance, you can't add it to the transmit side like you're trying to do. Granted, earth terminal G/T is the figure of merit at the receiver antenna. The standard formula is "G/T = Receive Antenna gain – 10 log(system noise temperature). Bear with me here, cuz I can't edit subscript into the next few sentences. But consider that since this same antenna transmits as well, the Gt (gain-transmit) is also related to G/T. And since gain is a function of frequency, Gt is always better than Gr (on the same antenna). I will concede though, that using 15.4 to represent antenna gain in the previous example was redundant. Rather than grab the convenient G/T value, I should have crunched the numbers.
But no, a transmitter gain of 62dB didn't compute initially, until you qualified it as dBm. 62 dBm equates to 32 dBW. I originally interpreted your post as meaning 62 dBW. Big difference. Anyway, we agree that
-35 dBm - 6.3 + 62 = 20.7 dBm, and
-5 dBm - 6.3 + 62 = 50.7 dBm.
But to arrive at EIRP, you must add Gt. Using your numbers:
-35 dBm - 6.3 + 62 = 20.7 + 39.5 = EIRP 60.2 dBm (30.2 dBW)
-5 dBm - 6.3 + 62 = 50.7 + 39.5 = EIRP 90.2 dBm (60.2 dBwW)
ViaSat data rates the 62cm configuration at EIRP 48 to 58 dBW. So now we're pretty close at the top end. I'm sure the tenths will be variable relative to cable loss. Not sure what's going on at the bottom though (30.2).
Backing up a bit, we tentatively agree that -5 dBm - 6.3 + 62 = 50.7. And I'm guessing you just picked 34.8 dBm because it conveniently equates to 3 watts. What I'm still missing is how the 50.7 becomes "limited to 34.8 dBm by the transmitter capabilities".
//greg//
--
DW4000CE/1.2 Ghz Tualatin/512 PC133 - W2K Pro/SP4 - SRS G11/1370H - RSL83/ACP81 - v4.2.1.10C - RWIN 513920/MTU 1500 - proxy switch - Gateway/DNS 205.177.62.91/66.82.4.8 plus rollovers - Firefox 1.02 | |
Mr_DNS
join:2005-09-07
Laredo, TX | reply to randyvsatus
said by randyvsatus:My question then becomes why running at say 3 watts transmit (assuming that is the max) all of the time cause a tria to crap out... .
.
I think Holler posted on this topic here, and at Wildblue.CC.
The large silver screw, if loose or gone, caused moisture to seep in unit, and moisture took it out.
The screw on mine was not snugg...as was others.
. | |
| reply to grohgreg
said by grohgreg :
But no, a transmitter gain of 62dB didn't compute initially, until you qualified it as dBm.
I still don't think we're quite on the same page, but the end result may still be the same.
I never said the transmitter gain was 62 dBm. dBm only makes sense in terms of transmit power.
said by ghrogreg :
And I'm guessing you just picked 34.8 dBm because it conveniently equates to 3 watts.
No, I picked 34.8 dBm direct from the U.S.Monolithics datasheet (»www.usmonolithics.com/products/c···1024.pdf), as the Output Power at the 1dB compression point. From the same datasheet:
Small Signal Transmit Gain (over freq. and temp.):
Min: 48 dB
Typical: 56 dB
Max: 63 dB
Note the units: power as dBm, gain as dB. As it turns out, when I guessed 62 dB for the gain, I was high, but still within the spec.
For more on how RF amplifiers are rated, and the meaning of P1dB, take a look at »www.evaluationengineering.com/ar···iers.htm .
Disclaimer... I hadn't seen the spec sheets when we started this discussion... but now we have more information... | |
| reply to grohgreg
said by ghrogreg :
Rather than grab the convenient G/T value, I should have crunched the numbers.
Thanks for clarifying that.
It looks like the antenna datasheet is at »raven.co.uk/satellite-products/d···heet.pdf
Though they show Gt at 14.25 GHz, rather than at 30 GHz, so I'm not sure... | |
|
