Hi all What is the "right" total amplification factor of an IF amplifiers chain in a radio receiver ?
For example a receiver employs 4 IF amps cascade , what is the proper total voltage amplification of that IF cascade ?
Thanks in advance
EC
On a sunny day (Fri, 1 Feb 2013 03:05:43 -0800 (PST)) it happened RealInfo wrote in :
radio receiver ?
Just multiply the voltage gain of each stage, or add the db gain of each stage.
I reality IF gain will likely be under influence of some AGC[1], and vary between 'next to nothing[2]' and 'a lot[3]'.
[1] Automatic Gain Control. [2] Can be less than 1. [3] Depends..There are also limiting IF amplifiers, for example for FM, so then output no longer increaes when the input increases, you can than speak of a minimum signal level required for limiting.
duh
radio receiver ?
I liked putting the total gain of the RF, mixers, and IFs so that the system would start AGC on noise. Then I could use the RF gain or If gain control to back off on gain for specific applications. I also liked to put a primary bandwidth filter as close to the antenna as possible, usually at the IF input, in order to eliminate adjacent channel signals, AND, if economics allowed, putting another primary bandwidth filter at the IF output, to set the system noise bandwidth.... If good Xtal filters are used, this is nearly always too expensive for commercial production, so there is usual a compromise position determined, depending on the system requirememts....
So my answer to your question is that the IF gain should be combined with the other gains/losses to develop an overall response.
in a radio receiver ?
al
As Jan said,
It can be next to nothing or a lot. It all depends on architecture. If you are digitizing the signal you likely do not need a whole lot.
I was taught to build a table which tracked noise power through the channel on one end and how close to compression the highest expected signals got on the other end. Build a table which shows the increases /decreases in noise power and also for the strongest expected signal coming in. This table should get you started
radio receiver ?
Yup. Just enough RF gain to override the noise of a nice strong mixer, followed by a band setting filter.
AGC generally hurts linearity pretty badly, so for instrument use it's usually better to cascade fixed-gain amps and variable attenuators.
Radio is fun.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 USA +1 845 480 2058 hobbs at electrooptical dot net http://electrooptical.net
n in a radio receiver ?
otal
I forgot to add, you need to also track the weakest expected signal and keep track of S/N ratio degradation throughout.
a radio receiver ?
Microdyne built analog Telemetry recivers with linear AGC. It was one of their selling points. The rest of the industry used standard AGC, except for Scientific Atlanta who did an almost exact copy of one of our products. That ended up in court so Microdyne ended up building those, too.
Depends on the required level at the demodulator, the gain (if any) at the mixer, any pre-mixer (RF amplifier) gain, the available signal voltage. If it's AM, AGC is usually used. FM receivers are usually designed to hard limit (clip).
--
"For a successful technology, reality must take precedence
over public relations, for nature cannot be fooled."
(Richard Feynman)
in a radio receiver ?
al
What these guys are talking about is something called ?gain line-up?. It?s the system architecture design that balances the ?large signal? distortion effects vs small signal s/n requirements.
As they've said, final ?baseband? requirements will dictate overall spe cs.
I'd tell you to do a search, but you most likely already have; it's not a t rivial exercise if you don't have any rf experience.
good luck
a radio receiver ?
One fairly cute method at low frequency is to use an SA614 FM IF strip, and use its RSSI output to drive the gate of a 2N7002 used as a variable attenuator on the input. You need to AC-couple 100% of the input signal onto the FET gate to get decent linearity though.
JFETs are said to need 50%, but 2N7002s need 100%, and 150% is even a bit better, e.g. two in series with the gates AC-coupled to the drain of the upper FET--that1s 200% for the lower FET and 100% for the upper one.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
As you've just found out -- it's complicated, and it depends not only on what the radio needs to do, but on choices you make for mixer and RF gain stages. It gets even more complicated if you're trying to design for cost (or simplicity, if it's for home-built).
You have considerable latitude in trading off gain in the IF stage for gain in the baseband, but at some point it gets very difficult to isolate the input of any amplifier from the output. The extreme example of this is probably a direct-conversion receiver using a diode-ring mixer with no RF preamp. In that case, there is no IF, no RF gain, and mixer loss. When you look at radio designs you'll see that those radios are designed to use headphones -- in my personal experience, I haven't been able to get one to work with speakers for the oscillation (although I haven't tried hard).
Breaking up the gain, either by adding RF gain, using a mixer stage with gain (the NE602/612 is popular with radio amateurs), or adding an IF, is a help. Clearly, part of the attraction of an IF is a place to add gain, but there have been a number of successful radios used for amateur work that had only a small fixed gain at IF, with all of the "real" gain (and AGC, if any) done at audio.
I have Wes Hayward's "Radio Frequency Design" -- it is good if you have an EE background. I don't have his "Experimental Methods in RF Design", but I know that it comes highly recommended if you are more of a technician than an engineer. I suggest you get one or both and go through them. I think the ARRL Handbook touches on the subject as well, but it won't go into any great depth.
-- My liberal friends think I'm a conservative kook. My conservative friends think I'm a liberal kook. Why am I not happy that they have found common ground? Tim Wescott, Communications, Control, Circuits & Software http://www.wescottdesign.com
radio receiver ?
Here's a link to hp's old appCad.
in a radio receiver ?
total
This was a dual conversion receiver, intended for diversity reception and used 70 MHz for the low IF frequency. The AGC system in each receiver was 0-5 volts, and was available as a buffered output on each receiver to feed the diversity combiner. The AGC control system had to be linear to no more than 1.5 mV at both extremes, and contained up to
14 opamps, depending on the selected function. You could select the time constant, as well.o
in a radio receiver ?
tal
stage.
So, if the smallest signal input is 1 microvolt, and you want 1 volt out, the gain is 120db? And that includes audio gain of maybe 20db?
-Bill
a radio receiver ?
stage.
But, but, but dB only refers to power, right? ;-)
Info
ain in a radio receiver ?
total
ch stage.
ary
it's just a conversion, note the reference in the following;
dBm,dB(W),dB(uV),dB(V),etc.
also note rf systems generally are 50 ohms, cable 75 ohm.
take that into account when doing the conversion.
a radio receiver ?
stage.
Decibels are fertile ground for confusion, much like percentages are for bankers. The basic definition is 10*log10(P1/P2). _If_ both are electrical signals working into the same impedance, this is the same as 20*log10(V1/V2). The trouble is that not everyone is equally careful with this.
For example, control system engineers employ 20*log10(V1/V2) and totally ignore impedance. They even use dB to express ratios between different units, e.g., as the ratio between say, voltage and pressure of a transducer. (I once got into an argument over that with Dr. Middlebrook.) RF engineers will usually, but not systematically, take the impedance into account. For example again, my Rohde & Schwartz 5075 matching pads are marked as having the same attenuation in both directions, while my Radiall matching pads have two different values, despite doing exactly the same thing in the same way.
Then there is the jungle of reference levels, with an abundant number of definitions, sometimes conflicting, often illogical. For example, dBm is dB referred to 1 mW, but dBu is referred to 1uV. I wouldn't be at all surprised if with some digging, we could come up with over a hundred different definitions of reference levels that are in common use in one corner or another.
Jeroen Belleman
in a radio receiver ?
total
stage.
it's pretty straight forward if people are clear when they use the terms.
gain = dB.
input and output ports should always have an absolute value referenced to them.
keep the impedance straight and it becomes simple arithmetic.
btw, it's virtually impossible to do a gain line-up correctly if you don't keep this stuff straight when using mixed impedances and equiv nf and IM specs.
Info
ain in a radio receiver ?
total
ch stage.
ary
The original question was about voltage gain. 20dB is a factor of 10 times the voltage. 20dB=10 , 120dB= 1 million.
-Bill
radio receiver ?
IF? Whazzat? Everything I see and do is either direct conversion (zero Hz IF), software defined radio (digital everything), or TRF (tuned RF for instruments). Even the common AM/FM broadcast band radios are now all digital.
Well, I guess I can make a rough estimate.
You probably want to hear signals down to about -110dBm. Your earphone probably wants about 100mw into 32 ohms or about +20dBm into
50 ohms. Power_gain = +20dBm - -110dBm = 130db Of course, that only works if both the input and output impedances are 50 ohms. Converted to voltages +20dBm @ 50 ohms = 2.25Vrms -110dBm @ 50 ohms = 710uVrms Voltage_gain = 2.25 / 710*10^-6 = 3170 Well, that's not exactly the IF voltage gain, but more like the system gain. Each stage along the RF->IF->Audio path has either gain (amplifiers, active mixers) or loss (passive mixers, filters, demodulators). Each stage also has a finite limit as to how much signal it can handle without clipping, distorting, or catching fire. If your boss or instructor has an interest in dynamic range, then your task is to distribute this 130dB of gain in such as way that you have sufficient sensitivity (mostly controlled by the front end RF or IF stage NF and gain), without overload. In this case, the ideal maximum dynamic range design is when all the stages overload or distort at the same RF input level.Of course, this is a gross simplification. For example, if you're demodulating FM, you will want additional IF gain to take advantage of IF impulse noise clipping and AM rejection, and provide for noise squelching. The design really depends on what you're trying to accomplish, which you haven't bothered to specify or explain.
JDC recommended HP AppCAD for doing the calculations. I agree. See the Signals-Systems -> NoiseCalc page. This is what it looks like:
Good luck.
-- Jeff Liebermann jeffl@cruzio.com 150 Felker St #D http://www.LearnByDestroying.com Santa Cruz CA 95060 http://802.11junk.com Skype: JeffLiebermann AE6KS 831-336-2558
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.