HD HomeRun HDHR5-2US vs Flex 4K Sensitivity

[WHJ]

Since my tests indicated almost identical sensitivity of the HDHR5-2US and the Flex 4K HD HomeRun tuners while others reported better sensitivity from the HDHR5, I thought it best to improve my overload and IMD tests of the tuners.

I knew that my overload tests were limited by the isolation of the combiners that I used. I added band pass filters to prevent the signal from my R&S SFQ ATSC 1.0 test transmitter getting back into my R&S SFE100 test transmitter and creating IMD (as well as vice-versa). That solved my problem. That made my test equipment path be as shown below.

My goal was to combine a weak DTV signal with a much stronger one and measure the level of the stronger signal which would hurt reception of the weak signal. I set the weak signal 5 dB above the minimum level for proper reception by the tuners. The amplitude of the other signal on a different channel was increased to the point where the weak signal could no longer be received. I started with the weak signal on channel 6 while the stronger signal was on channel 9. The Flex 4K had perfect reception with a 1 dB greater dynamic range than the HDHR5. However, the HDHR5 maintained intermittent reception with distorted video and noisy audio for significantly longer than the Flex 4K. The tuner would not need to recapture the signal again. Accepting that type of reception, the HDHR5 had 9 dB greater dynamic range than the Flex 4K.

Doing the same sort of test with the stronger signal on channel 21 instead of 9 produced almost identical results.

I also tried the almost adjacent channel for the stronger signal. Precisely adjacent channel would require crystal filters to use the same technique. I tried about as close as I could get—2 MHz above channel 6. I called it channel 6.3. That is still close enough that IMD from that channel 6.3 transmitter would hurt reception of the weaker channel 6 signal. The Flex 4K was marginally better—maintaining solid reception during the 2 dB range where the HDHR5 just had intermittent reception. The results are summarized below.

I have already replaced the GALI-52 input amplifier in one of my Flex 4K’s with a SBB2089Z. It just improved things by one dB. Unless I can increase that, I will probably not do that on my other Flex 4K’s.

So, a summary of the two tuners is that their sensitivities will be identical unless you have some RF between 50 – 650 MHz with a level more than your tuner’s limitations. When that first occurs the HDHR5 will maintain reception with a higher overload signal than the Flex 4K will. I do not see any filters in the tuner which will limit the overloading signal to just DTV signals. The strongest ones in that 600 MHz range will determine your overload potential.

[WHJ]

After my above tests I tried two additional RF amplifiers.

I first replaced the GALI-52 amplifier in my Flex 4K with a GALI-74. Like with the SBB2089Z, the improvement was not significant.

I then tried a TQP3M9009. It increased the dynamic range of my Flex 4K by 6 dB! That will be a worthwhile change. Due to the replacement IC's increased supply current and heat I will start with the replacement to two of my four Flex 4K's and see if they survive long-term tests.

[tripelo]

After my above tests I tried two additional RF amplifiers.

I first replaced the GALI-52 amplifier in my Flex 4K with a GALI-74. Like with the SBB2089Z, the improvement was not significant.

I then tried a TQP3M9009. It increased the dynamic range of my Flex 4K by 6 dB! That will be a worthwhile change. Due to the replacement IC's increased supply current and heat I will start with the replacement to two of my four Flex 4K's and see if they survive long-term tests.

WHJ, that is interesting.

From datasheets, looks like the change gained dynamic range at both top and bottom end of the signal range.

Numbers below at 600 MHz are approximate due to circuit implementation and variations of current, temperature, and Voltage.

Approx OIP3:

GALI-52+: 30 dBm
GALI-74+: 35 dB
TQP3M9009: 41 dBm

Approx Noise Figures:

GALI-52+: 3 dB
GALI-74+: 2.8 dB
TQP3M9009: 0.9 dB

Approx Gain:

GALI-52+: 22 dB
GALI-74+: 23.5 dB
TQP3M9009: ~26 dB



When a preamplifier is used, often the mixer is the limiting factor for instantaneous dynamic range.

That the TQP3M9009's higher gain and higher OIP3 allowed dynamic range improvement indicates significant loss must be present between preamplifier and the mixer.

High losses (if they exist) between preamplifier and mixer(s), may be due to the splits required for driving multiple tuners.

Of course AGC control may add to the loss between preamplifier and mixers. But AGC is not instantaneous, and probably unable to suppress RF channels near in frequency a desired channel, without also attenuating the desired channel.

You have some interesting experiments. Thanks.

[WHJ]

Numbers below at 600 MHz are approximate due to circuit implementation and variations of current, temperature, and Voltage.

My comments on the effects of the RF amplifier changes in my Flex 4K's were based on making the actual IC changes, with before and after sensitivity and overload figures measured using a R&S SFQ ATSC 1.0 DTV test transmitter feeding a Flex 4K. The measured dynamic range was the difference between the minimum level signal which could be received by the Flex 4K and the signal level which would overload the Flex 4K to the point of preventing reception of a weak signal from my R&S SFE100 ATSC 1.0 DTV test transmitter on another channel. Signals stronger than that loud signal from the SFQ could still be decoded and viewed, but weak signals could not be viewed then.

The test circuit shown on the GALI-52 specification sheets does not have any negative feedback. However, the GALI-52 circuit in the Flex 4K has 1,020 ohms of negative feedback. That reduces its gain some while increasing the amplifier's dynamic range. Likewise, the TQP3M9009 specification sheets do not show any feedback resistance. However, I left some in my modified Flex 4K. To show the effects on changing the value of the feedback resistance I made a TQP3M9009 test amplifier circuit out of the Flex 4K and measured the gain and dynamic range changes after changing the feedback resistance. That is shown below. That was measuring a channel 35 ATSC 1.0 DTV signal on a R&S EFA DTV test receiver.

[tripelo]

... based on making the actual IC changes, with before and after sensitivity and overload figures measured using a R&S SFQ ATSC 1.0 DTV test transmitter feeding a Flex 4K....

Thanks for the test info.

...The test circuit shown on the GALI-52 specification sheets does...That is shown below. That was measuring a channel 35 ATSC 1.0 DTV signal on a R&S EFA DTV test receiver.

Good to know about the resistive feedback. As you know, that feedback further degrades the not so great noise figure of the GALI-54.

Could not view your images. The 'Edge' browser didn't show a clue that there were attached images.Tried older version of Firefox that seems to work on most websites, it showed image symbols, but clicking displayed a black image with message indicating a '.png' image.

Also, same for your images posted earlier in this thread.

Don't know why a .png would not be decoded by two different computers on this website with either Firefox or Edge, but you might try posting a .jpg or .gif to see if that makes a difference.

[WHJ]

...Don't know why a .png would not be decoded by two different computers on this website with either Firefox or Edge, but you might try posting a .jpg or .gif to see if that makes a difference.

I tried again with .png images (see above). Particularly for graphs, .jpg images look faded and not as precise as .png images. More so if any additions are made to the graphs. So, that's what I generally use for things other than true pictures. My prior inclusions were probably not done the best way on that. I will double check on what else was left off above after I complete some new tests on the GALI-52 and a PGA-106-75 amplifier.

...feedback further degrades the not so great noise figure of the GALI-54.

Notice how the Minimum dBuV data (which you can hopefully now see on the table above) increases only slightly with increased feedback and decreased gain. That results in increased dynamic range when the gain is reduced. I think those tests are more relevant to the actual performance of the Flex 4k than a spec sheet noise figure since my tests use an actual 6 MHz ATSC 1.0 signal and performance of the Flex 4K being tested.

[tripelo]

I tried again with .png images (see above).

Yes, now the schematic and table are viewable with both browsers (Edge & Firefox).

...Particularly for graphs, .jpg images look faded and not as precise as .png images. More so if any additions are made to the graphs. So, that's what I generally use for things other than true pictures.

Yes, jpg with lossy compression is mainly good for real life images (photos,etc). PNG and GIF being lossless work better for schematics, graphs, etc. PNG being much newer relative to GIF has advantages.

...Notice how the Minimum dBuV data ....increases only slightly with increased feedback and decreased gain. That results in increased dynamic range when the gain is reduced. I think those tests are more relevant to the actual performance of the Flex 4k than a spec sheet noise figure since my tests use an actual 6 MHz ATSC 1.0 signal and performance of the Flex 4K being tested.

Understandably, seems in your case the instantaneous high dynamic range is very important.

The change of min dBuV from 17.7 to 20.7 represents an increase in noise figure of about 1.3 dB.

For distant reception, if one may already be using a large antenna, achieving an additional 1.3 dB represents a challenge.

Good to see in your chart what the tradeoffs are.

[WHJ]

For distant reception, if one may already be using a large antenna, achieving an additional 1.3 dB represents a challenge.

You were taking my extreme on gain reduction--I would not recommend that. For HDHR tuners the limiting factor for weak signal reception is not the noise figure or minimum dBuV, it's how strong is the loudest other signal from the antenna between 50 - 650 MHz being fed to the tuner. If it causes tuner overload due to the economy input amplifier used, its weak signal sensitivity will suffer.

I use tower mounted preamplifiers having a much greater dynamic range the HDHR's with all of my antennas. There are RF filters and attenuators inside which I adjust so that the strongest signal being received does not overload my tuners.

I made similar tests to those above on the TQP3M9009 to the GALI-52 used in the Flex 4K. SiliconDust uses 1,020 feedback resistance with it. That decreased the gain by 2.5 dB while increasing the dynamic range by 2.1 dB. Naturally, the minimum dBuV level increased by 0.4 dB.

The HDHR5-2US uses a GALI-51 input amplifier. It has a 930 ohm negative feedback resistor along with a series inductor which limits the gain reduction on the higher frequencies.

[tripelo]

You were taking my extreme on gain reduction--I would not recommend that.

That is correct.

..For HDHR tuners the limiting factor for weak signal reception is not the noise figure or minimum dBuV, it's how strong is the loudest other signal from the antenna between 50 - 650 MHz being fed to the tuner. If it causes tuner overload due to the economy input amplifier used, its weak signal sensitivity will suffer.

Yes that is often the case, more likely than not.

Exceptions may be found at receiving locations where no strong locals exist, especially if no external preamp is used. In these cases the noise figure of the HDHR could make a difference.

...I use tower mounted preamplifiers having a much greater dynamic range the HDHR's with all of my antennas. There are RF filters and attenuators inside which I adjust so that the strongest signal being received does not overload my tuners.

Yes, your work with multiple high dynamic range and low noise preamps (previously reported in this forum) is much appreciated.

Over several years, have designed and built several versions of preamps using the PGA103+ (same as SPF5189z) and PSA4-5043+ (same as SPF5043Z). Earlier you indicated your results with these devices. Also, built a few preamps using the ATF54143 GAASFET, but generally prefer the block amplifier versions over discrete transistors. My interest was mainly lowest noise figures, but certainly appreciate high OIP3 intercepts. Still plan to build a TQP3M9036 version that had apparently lowest noise figure (MER-15 Min your chart).

Certainly, in your situation with the external preamps, the noise figure of the HDHR receiver becomes insignificant.

..I made similar tests to those above on the TQP3M9009 to the GALI-52 used in the Flex 4K. SiliconDust uses 1,020 feedback resistance with it. That decreased the gain by 2.5 dB while increasing the dynamic range by 2.1 dB. Naturally, the minimum dBuV level increased by 0.4 dB.

The HDHR5-2US uses a GALI-51 input amplifier. It has a 930 ohm negative feedback resistor along with a series inductor which limits the gain reduction on the higher frequencies.

Interesting that the HDHR5-2US uses GALI-51 and the Flex 4K used the GALI-52.

GALI-51+ (Approx at 600 MHz)

OIP3: 31 dBm
Gain: 17.7 dB
Noise Figure: 3.6 dB

[WHJ]

Over several years, have designed and built several versions of preamps using the PGA103+ (same as SPF5189z) and PSA4-5043+ (same as SPF5043Z). Earlier you indicated your results with these devices. Also, built a few preamps using the ATF54143 GAASFET, but generally prefer the block amplifier versions over discrete transistors. My interest was mainly lowest noise figures, but certainly appreciate high OIP3 intercepts. Still plan to build a TQP3M9036 version that had apparently lowest noise figure (MER-15 Min your chart).

My tests with the SPF5189Z have shown it to not be very durable.
Due to TV band compacting, new power supply and line noise, atmospheric noise, interference, etc., I have not found that a low noise figure is not as important as I initially expected.

The TQP3M9028 amplifier is the one that I have used most often. After my feedback resistor tests, I have built some alternatives:

I think that I will put the above TQP3M9008 amplifier to use outside.

[tripelo]

My tests with the SPF5189Z have shown it to not be very durable.

Had a couple of them fail (think they were are all PGA103+ versions). One failure was due to an error during testing. The other failed on tower top due to lightning strike nearby.

Have a PGA103+ based preamp on tower for a few years with no failure. Modified input of this PGA103+ preamp to better withstand input Voltage spikes.

According to the specifications; the PSA4-5043+ is better protected against static Voltages that the PGA103+. Have had one of these in service on tower for maybe 10 years without failure.

...Due to TV band compacting, new power supply and line noise, atmospheric noise, interference, etc., I have not found that a low noise figure is not as important as I initially expected.

Agreed.

Have a rural location with several hills or mountains between transmitter and receiver. Because of the repack, noise figure is not as important these days . All commercial TV stations are 75 to 100+ miles distant, but for almost every channel there is a duplicate in another direction (75 to 100+ miles distant). At times, This duplication of transmitted channel tends to raise the effective noise floor.

The TQP3M9028 amplifier is the one that I have used most often. After my feedback resistor tests, I have built some alternatives:
I think that I will put the above TQP3M9008 amplifier to use outside.

Interesting new data.

[WHJ]

According to the specifications; the PSA4-5043+ is better protected against static Voltages that the PGA103+. Have had one of these in service on tower for maybe 10 years without failure.

I do not like having to deal with the increased gain on the lower frequencies when using a PSA4/SPF4053 or PGA103. I have found that a 332-ohm resistor, 68 nH inductor and a DC blocking capacitor in series providing feedback for the SPF5043Z will come close to balancing the gain between 50 - 610 MHz.

All commercial TV stations are 75 to 100+ miles distant...

Unless you have them filtered out, FM broadcast transmitters or any other strong VHF or UHF station can overload your preamplifier and/or HD HomeRun tuner. Without compensation, you will have significantly more gain on the FM vs. UHF frequencies.

[tripelo]

I do not like having to deal with the increased gain on the lower frequencies when using a PSA4/SPF4053 or PGA103.

...Unless you have them filtered out, FM broadcast transmitters or any other strong VHF or UHF station can overload your preamplifier and/or HD HomeRun tuner. Without compensation, you will have significantly more gain on the FM vs. UHF frequencies.

Yes, we are in agreement.

These devices without feedback have relatively high gain at lower frequencies tapering down as frequency increase. And, that is opposite of what is usually desired.

As you mentioned FM stations, both by being more prevalent and being relatively lower in frequency, their signals can be strong enough to desentisize a tuner.

Additionally, the spectral energy content of lightning is very strong at very low frequencies tapering downward to the TV bands.

...I have found that a 332-ohm resistor, 68 nH inductor and a DC blocking capacitor in series providing feedback for the SPF5043Z will come close to balancing the gain between 50 - 610 MHz.

That most certainly helps prevent strong signal overload.

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

Realizing you have, previously in this forum, described your preference for high-pass filtering in front of mast mounted preamplifiers.

In my case, there is only one relatively strong FM station (PBS).

All preamps are outfitted with an input high-pass filter, designed not to affect whatever band is of interest (VHF or UHF depending on antenna).

The filter consist of first element being an inductor to ground followed series capacitor. All components in front of active device are chosen to have highest practical Q. The inducctor is usually hand wound with silver plated wire and capacitors selected to be highest Q.

As you know, the inductor to ground is quite important, it constantly drains charge (static or otherwise) that can build up in the antenna/coax combination. Being low pass it shunts low frequency energy to ground before it gets into other components.

The input filter's main task is to reject lightning energy, but it is effective against FM signals as well.

Another precaution against lightning is to place low capacitance fast transient protector in front of everything.

Similar to my understanding of your setups, after the preamps have various notch, bandpass, and band reject filters before combining signals for input to tuner.