AVS Forum banner

1 - 20 of 162 Posts

·
Registered
Joined
·
6,318 Posts
Discussion Starter #1
The RCA TVPRAMP1R preamp is one of the few preamps available with separate or combined VHF/UHF inputs. Except for gain, there are no specifications available for this preamp. This post will correct that.

I purchased one of the preamps and performed various measurements with my Rigol spectrum analyzer, an Ailtech noise source and an HP 75 ohm directional coupler. Here are the measurements with analyzer images attached.

The power supply is 9V and the preamp draws 68ma.

1) Full band gain measurement. The UHF gain is 2-3 dB higher than claimed.
2) UHF gain to 1000 MHz. I wanted to see how badly a nearby cell phone tower might affect the preamp. The gain is => 20 dB up to 1000 MHz.
3) An expanded look at VHF and the FM trap. With the FM trap switched out the gain is flat from low to high VHF.
4) Isolation. This is referenced to 0 dB and shows how much signal leaks through the preamp with the power off. Strong signals, especially on VHF will still be received even with no power applied.
5) Noise Figure UHF. This is better than most preamps.
6) Noise Figure high VHF. Good enough for most situations.
7) Noise Figure low VHF. Better than required since the environmental noise will be higher than the noise figure.
8) Return Loss for low, high and UHF combination input. The markers show the lowest return loss (highest SWR) for each band. I converted those to SWR.
9) Return Loss for the separate VHF input. The markers show the lowest return loss (highest SWR) for the low and high VHF bands. I converted those to SWR.
10) Return Loss for the output port. The markers show the lowest return loss (highest SWR) for each band. I converted those to SWR. This is much better than I've seen on other preamps.

I also preformed a 1 dB gain compression test for VHF and UHF. There are no graphs for this test. P1dB on VHF is +15 dBm and on UHF is +10 dBm.

For such an inexpensive preamp I think the measurements are all good except for P1dB. I'd really like to see +20 dBm. My guess is that this preamp probably should not be used on a high gain UHF antenna with TV Fool Noise Margin predictions of much over +40 dB. This should be an excellent preamp where one is required.
 

Attachments

·
Registered
Joined
·
6,963 Posts
The RCA preamp is supposedly similar to the original Channel Master 7778/7777 preamps in terms of design and offering both separate UHF and VHF inputs. The noise figure on the CM was claimed about 2.0, whereas the RCA is reportedly about 2.8 unless you state otherwise. But apparently a good inexpensive preamp that is supposedly resistant to overload, which should help if you have both strong and weak signals.
 

·
Registered
Joined
·
693 Posts
I've had this preamp for 4 years now, & it has helped at getting 4 LPTV UHF stations, & the only Class A VHF-LO station. The full power stations were never the problem, as I'm about 30 miles SE of Chicago. It has always been the LPTV stations that are weak in my direction that have been the problem, & the pre-amp has helped it. It especially helps with LPTV VHF stations that broadcast less than 1kw (maybe 2kw as well).
 

·
Registered
Joined
·
6,318 Posts
Discussion Starter #4
The RCA preamp is supposedly similar to the original Channel Master 7778/7777 preamps in terms of design and offering both separate UHF and VHF inputs. The noise figure on the CM was claimed about 2.0, whereas the RCA is reportedly about 2.8 unless you state otherwise. But apparently a good inexpensive preamp that is supposedly resistant to overload, which should help if you have both strong and weak signals.
I measured an old CM7777. I thought I had posted the results but I can't find it. If that post exists it will have the P1dB results. Most of the preamps I measured have P1dB in the mid teens. The Tin Lee I measured was +20 dBm, the best UHF preamp I found. A few old CM7777 measurements are attached. The RCA has a slightly better noise figure and slightly higher gain.
 

Attachments

·
Registered
Joined
·
797 Posts
Chuck,

The DS815-TG can do a very good facsimile of a P1dB measurement and you can capture its screen. Use the power sweep function of the generator over a very tight frequency span and you'll get something that looks like the attachment. From the step display, it's fairly straight forward to determine the P1dB level.

Several caveats:
1. Generator non-linearity. The TG on our Rigol is +/- about 1 dB. Pick a frequency where it's fairly accurate.
2. Depending on the amp to be tested, you may need either additional attenuation or amplification of the TG's -20 to 0 dBm range.

I tested the RCA about two years ago but haven't revisited it recently. Your results sound fairly consistent with what I recall.
 

Attachments

·
Registered
Joined
·
6,318 Posts
Discussion Starter #6
Chuck,

The DS815-TG can do a very good facsimile of a P1dB measurement and you can capture its screen. Use the power sweep function of the generator over a very tight frequency span and you'll get something that looks like the attachment. From the step display, it's fairly straight forward to determine the P1dB level.

Several caveats:
1. Generator non-linearity. The TG on our Rigol is +/- about 1 dB. Pick a frequency where it's fairly accurate.
2. Depending on the amp to be tested, you may need either additional attenuation or amplification of the TG's -20 to 0 dBm range.

I tested the RCA about two years ago but haven't revisited it recently. Your results sound fairly consistent with what I recall.

Thanks for the tip. I've been using an HP-8640B as the signal source, a JFW 75 ohm 1 dB step attenuator and the Rigol as the power meter.
 

·
Registered
Joined
·
147 Posts
Calaveras, thank you for your detailed review of the RCA preamp and the follow-on measurements of the CM7777.

New data often raises questions::

1, 2, & 3 - Did you measure the separate UHF gain (down thru lower VHF)?

Curious how much FM gets through the independent UHF section.

4 - Wonder what the power-off isolation is thru the UHF section only (separate input)?

(Assuming your spectrum 4 was with combined mode).

6 & 7 - Did you see any difference in VHF noise figure when FM Trap was in?

9 - Did FM Trap affect VHF input Return Loss?

Seems you have some nice test equipment.

--------------------------------

ADtech,

Thanks for the measurement data you have posted over several months.
 

·
Registered
Joined
·
6,318 Posts
Discussion Starter #8
Calaveras, thank you for your detailed review of the RCA preamp and the follow-on measurements of the CM7777.

New data often raises questions::

1, 2, & 3 - Did you measure the separate UHF gain (down thru lower VHF)?

Curious how much FM gets through the independent UHF section.

4 - Wonder what the power-off isolation is thru the UHF section only (separate input)?

(Assuming your spectrum 4 was with combined mode).

6 & 7 - Did you see any difference in VHF noise figure when FM Trap was in?

9 - Did FM Trap affect VHF input Return Loss?

Seems you have some nice test equipment.

Your questions point out that with a preamp like this there are many possible combinations to measure. I did not do them all but I have a couple more which are now attached below.

1) VHF Gain with FM trap in Combined (yellow)/Separate (purple) inputs. They are virtually identical.
2) VHF return loss on VHF separate input.

When I get a little more time I'll review you questions that the additional measurements don't address.

The Rigol analyzer is the centerpiece of the test equipment. To convert the 50 ohm Rigol to 75 ohms I use a couple of Anzac TPX-75-4 transformers. After a long search I found an Ailtech 7618 noise source. As far as I can tell there's no such thing as a 75 ohm noise source so to use it in a 75 ohm system I added a JFW 57ZT-1G 75:50 ohm transformer with about 1 dB loss on UHF and 0.7 dB loss on VHF. For return loss I'm using an HP-86207A 75 ohm directional coupler.
 

Attachments

·
Registered
Joined
·
147 Posts
... with a preamp like this there are many possible combinations to measure.
Yes quite a few combinations, and such measurements can consume a good bit of time..

Thanks for the additional scans.

...The Rigol analyzer is the centerpiece of the test equipment. .... For return loss I'm using an HP-86207A 75 ohm directional coupler.
Yes seems most precision lab test equipment is based on 50 Ohms. Often, hobbyists have to come up with innovative combinations to address 75 Ohm measurements. Looks like you have found your test combo to do the job.
 

·
Registered
Joined
·
797 Posts
The zip file attached below is the data I pulled from my RCA sample about a year and a half ago using the HP8970A Noise Figure Meter with an HP346A noise source.

The individual .txt file names are self-explanatory.

A note about the data from the 8970A: If the measured noise figure value exceeds the calibrated limit of the meter, usually between 30 and 35 dB, the meter inserts wildly invalid values for both gain and noise figure. When I pull the NF meter data into an Excel spreadsheet, I usually do a search for these values and replace them with a standardized value, for example 35 dB for NF and 0 for the gain if I want to keep the graph from having extraordinary spikes.

The Rigol has a built in conversion function for 75 ohm measurements when using a standard 5.7 dB insertion loss minimum loss pad.

This app note may be of interest: http://www.maximintegrated.com/en/app-notes/index.mvp/id/3250
 

Attachments

·
Registered
Joined
·
6,963 Posts
UHF Noise Figure?

I am not an engineer and these graphs and charts confuse me. So, long story short, what was the UHF noise figure on the RCA preamp determined to be? I had read where it reportedly was around 2.8, whereas the old Channel Master was claimed to be about 2.0. All I know is the lower the better.
 

·
Registered
Joined
·
797 Posts
Read the text files I posted for my readings. They're all in there.
 

·
Registered
Joined
·
6,318 Posts
Discussion Starter #13
I am not an engineer and these graphs and charts confuse me. So, long story short, what was the UHF noise figure on the RCA preamp determined to be? I had read where it reportedly was around 2.8, whereas the old Channel Master was claimed to be about 2.0. All I know is the lower the better.
I measured 1.9 to 2.2 dB which are the values I calculated from the display and I added to the image.

I use the noise source On/Off measured on a spectrum analyzer method to calculate noise figure. This method requires the Excess Noise Ratio (ENR) of the noise source to be known. Most noise sources have a table of ENR vs frequency on the device. The HP-8970A that ADTech uses has this table entered into the unit. ENR is interpolated between frequency points.

In this post I briefly explain the method and include the formula to calculate noise figure using the noise On/Off difference and ENR.

http://www.avsforum.com/forum/25-hdtv-technical/381623-official-avs-antenna-related-hardware-topic-524.html

Basically the higher the noise figure of the device being measured, the smaller the noise On/Off difference.
 

·
Registered
Joined
·
6,318 Posts
Discussion Starter #14
New data often raises questions::

1, 2, & 3 - Did you measure the separate UHF gain (down thru lower VHF)?

Curious how much FM gets through the independent UHF section.

4 - Wonder what the power-off isolation is thru the UHF section only (separate input)?

(Assuming your spectrum 4 was with combined mode).

6 & 7 - Did you see any difference in VHF noise figure when FM Trap was in?

9 - Did FM Trap affect VHF input Return Loss?

I think I've addressed your questions. New images attached.

1) Gain - VHF through UHF using the VHF separate input
2) Gain - VHF through UHF using the UHF input with input switch set to separate
3) Isolation - VHF Separate Input
4) VHF Noise Figure Trap In/Out (Purple - Noise off, Yellow Noise on with trap in, Cyan - Noise on with trap out) No difference in low/high VHF noise figure. Marker 1 is at 88 MHz and Marker 2 is at 174 MHz.
5) VHF Return Loss with trap In/Out (It looks much different but the SWR is good either way
 

Attachments

·
Registered
Joined
·
147 Posts
The zip file attached below is the data I pulled from my RCA sample about a year and a half ago using the HP8970A Noise Figure Meter with an HP346A noise source.
Thanks for the data files.

... Excel spreadsheet, I usually do a search for these values...
If no big mistakes, think the attached images reflect your data over the VHF and UHF bands.





One can see that the FM trap does not cover the lower end of the FM band (the response is above the 90 MHz marker).

Yes, good info. One can set up a spreadsheet utilizing those equations, or similar, and convert 50 Ohm derived S-parameters to 75 Ohm S-parameters and actual impedance values.

...what was the UHF noise figure on the RCA preamp determined to be? ...
tylerSC,

Calaveras' measurement:

...I measured 1.9 to 2.2 dB...
Those numbers are a bit lower than ADtech's (close to 2.5 dB), that is pretty good agreement considering measurement variables, equipment variables, and unit-to-unit variables.
 

Attachments

·
Registered
Joined
·
797 Posts
Here are a few other images from that period of testing that were in the same folder. All were done with our older HP8569B analyzer. I haven't revisited them for accuracy or expanded explanation.
 

Attachments

·
Registered
Joined
·
6,318 Posts
Discussion Starter #17
Thanks for the data files.

Those numbers are a bit lower than ADtech's (close to 2.5 dB), that is pretty good agreement considering measurement variables, equipment variables, and unit-to-unit variables.

The exact numbers depend on the ENR I used with the noise source. The problem is determining the loss of the transformer plus the adapters. I was using the numbers from the transformer spec but closer examination shows that might be as much as 0.5 dB too high. That means the NF numbers could be up to 0.5 dB too low. Low noise figures are hard to measure to within 0.1 dB because it's hard to know the ENR that accurately. Fortunately 0.5 db isn't going to be noticed by anyone watching TV. :)
 

·
Registered
Joined
·
147 Posts
I think I've addressed your questions. New images attached.
Yes, thanks for the additional measurements.

Quite interesting to me. May help with a puzzling question I had in mind. Need to think a bit more about the results shown in the graphs and how they relate to the actual preamp circuitry.

... a few other images from that period of testing that were in the same folder...
Good.

... Low noise figures are hard to measure to within 0.1 dB...
Yes, understood.
 

·
Registered
Joined
·
797 Posts
Chuck,

The problem is determining the loss of the transformer plus the adapters.
Additionally, the response of the transformers isn't going to be linear across the test band, in most situations.

When using the NF meter with min loss pads, all those extraneous insertion losses are calibrated out or accounted for during meter setup. For the min loss pads, since they have a very flat insertion loss (resistive elements), their insertion loss is known and it doesn't vary across the test frequency band. Additionally, the DUT's temperature must be accounted for and there's an entry that has to be made to account for that since the device's physical temperature will also affect the meter readings.

Fortunately 0.5 db isn't going to be noticed by anyone watching TV.
This is so true but is overlooked for so many who are looking for that last little bit of margin. Kind of like trying to squeeze an addition couple of horsepower out of an engine that's already near its limits.

You're certainly to be commended for putting together your numbers with your setup. You are definitely in the ball park!
 

·
Registered
Joined
·
6,318 Posts
Discussion Starter #20
Chuck,

Additionally, the response of the transformers isn't going to be linear across the test band, in most situations.

When using the NF meter with min loss pads, all those extraneous insertion losses are calibrated out or accounted for during meter setup. For the min loss pads, since they have a very flat insertion loss (resistive elements), their insertion loss is known and it doesn't vary across the test frequency band. Additionally, the DUT's temperature must be accounted for and there's an entry that has to be made to account for that since the device's physical temperature will also affect the meter readings.
I agree with you but there are tradeoffs to the methods. Although not applicable to what we're doing, a min loss pad lowers the noise source ENR and therefore lowers how high of a noise figure that can be measured. High noise figures require high ENR sources. It's been a couple of decades since I used an HP-8970A but I seem to remember that it ran a 28V square wave into the noise source so the meter measured noise on/off just like the spectrum analyzer method. High noise figures with a low ENR source produce a very small difference between noise on/off in turn producing large changes in the noise figure for tiny changes in the noise. We had a unit at work with such a high noise noise figure we had to use a noise source with an ENR of 25 dB.

Again it's been a long time, but I seem to remember that we went with the spectrum analyzer method because we had units that downconverted the input, say 16-18 GHz in to 2-4 GHz out. I don't think the HP-8970A could handle that situation, noise source at one frequency range and the measurement at a different frequency. The spectrum analyzer didn't care.

I looked at the issue of noise varying with temperature. Although it does vary, it takes a pretty big temperature change to see much difference. It's not like you're going to receive stations when it's 0°C out and not when it's 30°c.;)


This is so true but is overlooked for so many who are looking for that last little bit of margin. Kind of like trying to squeeze an addition couple of horsepower out of an engine that's already near its limits.
I'm guilty of this but I've found that squeezing out the last dB only buys you a few more reception minutes when the station is fading or when it's about to come back in. Signals just don't seem to sit nicely at 1 dB above the cliff for long.

There's kind of an equivalent in DTV to analog TV. An increase of 3 dB was about the minimum change you could barely detect in the analog picture. 6 dB was definitely noticeable and 10 dB was a real difference. In DTV 3 dB will improve the time a weak station can be decoded but nowhere near enough to make it reliable. 6 dB is a big improvement but the station will likely still have bad periods. A 10 dB increase can make a weak station pretty close to 100%.
 
1 - 20 of 162 Posts
Top