Rob Sherwood, NCOB's presentation at Contest University 2018

The word from the top!
Forum rules
Only admin's and moderators can start a topic in this forum, but anyone can reply to an existing topic. Off topic replies will be deleted.
Posts: 44
Joined: Wed Aug 24, 2016 5:20 am

Rob Sherwood, NCOB's presentation at Contest University 2018

Postby administrator » Sun Jun 03, 2018 4:14 am

User avatar
Posts: 2330
Joined: Fri Mar 10, 2017 1:47 pm

Re: Rob Sherwood, NCOB's presentation at Contest University 2018

Postby w-u-2-o » Sun Jun 03, 2018 7:00 pm

My take-aways:

1. From a pure receiver performance perspective, there is very little difference between the top 15 or 20 radios out there. Obtaining Rob's "Holy Grail" performance levels no longer requires rocket science or a ton of money.

2. When arguing about (1) above, make certain that no signal processing other than attenuation or pre-amplification is being used, the same filter passband is being used, and that both radios are sending audio to the same exact speakers at the same exact volume level. Even something as simple as having one radio louder than the other will cause one to be chosen over another.

3. Since the top 15 or 20 (some are not on Rob's list) are effectively the same when attached to an antenna, one should choose from them not based on pure receiver performance but rather on what else they bring to the table. Things such as number of receivers, filtering, signal processing features (RX and TX), and cost, are all much more important discriminators than 2dB here or there.

4. I disagree that receiver noise floor is not a significant performance parameter for some people. I regularly obtain better than "quiet rural" conditions. Thus it will be in my radio selection criteria as discussed in (3) above.

5. The IC-7300 remains the greatest value of all time in ham radios. You can't beat it in terms of dollars per decibel. Nobody (Flex or Apache) can compete with it at the low end of the direct-sampling SDR market, much less the conventional radios, that's why they are not trying to. If I needed two receivers, I'd consider buying two 7300's before I bought a single 7610. And you CW types can use an external APF.

Apache specific comments on Rob's presentation:

- Rob has only tested an ANAN-200D (99dB dynamic range) and an ANAN-7000DLE (103dB dynamic range), but certainly every Apache Labs radio will demonstrate better than 95dB dynamic range as they are all essentially identical in terms of their receivers, whether it's an ANAN-10 or an ANAN-7000. The later models have better preselection and phase noise, certainly, but they all should be better than Rob's 95dB number, with the possible exception of the 10E and 100B (14 bit ADCs on those radios).

- There is some work being done to see if the 6M LNA can be switched in for 15M and up. If successful, that should address some of Rob's concerns.

- Since 90% of the "radio" is in PowerSDR and not in the hardware unit (what many of us call the "radio"), all Apache Labs radios using the latest version of PowerSDR will all perform similarly. That includes latency, which no longer needs to be anywhere close to the very old value Rob obtained in 2016 when he tested the 200D. This also holds true for nearly all other features.

- It should not take messing with 7 settings to obtain on the order of 20mS latency for CW receive. That should be obtainable, or close to it, by merely selecting the low latency filter option. However that is not the "out of the box" default for PowerSDR.

- PowerSDR "out of the box" defaults definitely have major room for improvement. For someone with zero clue about the myriad features of PowerSDR it could take months to discover the necessary and desirable optimizations. For someone who is very familiar, it is still a chore to set up.

- Apache setups don't require dual monitors, although certainly monitor real estate is very worthwhile. A single 4K monitor is *plenty*, as are dual monitors of nearly any resolution except 640x480.

- Fully solid-state T/R switching is definitely a requirement for CW, and very desirable otherwise.

- When determining how much attenuation to use to optimize dynamic range for existing band noise, use the spectral display and avoid using old school methods such as your ear or S-meter. Add attenuation until you see the signal-to-noise floor ratios drop on the largest signals, then take out a little. With PowerSDR and Apache Labs radios, this can be done in 1dB increments if you wish by switching the receiver into step attenuator (S-ATT) mode--just double click the "ATT" above the setting window.


Posts: 6
Joined: Tue Jun 26, 2018 8:52 am

Re: Rob Sherwood, NCOB's presentation at Contest University 2018

Postby JA2GXU » Wed Jun 27, 2018 1:16 pm

Unfortunately this presentation don't have any minimum readable weak signal strength such as assumed MRS. And only have MDS that don't mean any MRS.
And also RMDR and DRNS can't meet actual reciprocal mixing test.
I posted to ARRL due to RMDR and actual reciprocal mixing test with test procedure and recomend ultra low phase noise SG E8663D Option HY2 and Option 1E1.
Ham radio requires kind of MRS or MRWS that very useful for DX etc. However MDS don't have any weak signal's readability.
I am thinking about assumed MRS and reciprocal mixing test that depend on Oscillators Phase Noise at 1kHz offset etc.
ANAN series have -138dBc/Hz Phase Noise at 1kHz offset.
Assuming MRS :-138+6+(-3)+0=-135 ➞-135dBm. (at 14MHz)

Double conversion use 2DDS and DSP that example are Elecraft K3 and Kenwood TS-590 .
If those rigs use good DDS then Phase Noise at 1kHz offset almost -125dBc/Hz.
Assuming MRS :--125+6+(-3)+6=-116➞-116dBm.
I assumed and thought Old version of K3 that DDS phase noise around -115dBc/Hz at 1kHz offset, thus asumed MRS -106dBm.



Return to “Apache Labs News”