Different versions of P2 firmware causing this?

[JJ4SDR]

I have two 7000 DLE MKII radios and I believe both have the same FPGA ( EP4CGX150F896 Cyclone IV series FPGA) in them as well although I am not 100% certain. One radio runs on FW v2.1.18 P2 ("radio #1" in the screen-capture below) and radio #2 runs on FW v2.2.2 P2. Note, I tried FW v2.2.2 on radio #1 a long time ago, but went back to FW v2.1.18 as From/To VAC underflows/overflows showed huge numbers acted pretty bad!

Observe 2 "spurs" (for lack of a more appropriate term) when using radio #2. They appear around 15kHz above and below my actual TX frequency. This, regardless of whether I am on CW or SSB or band. Granted that the strength of the "spurs" is fairly low and 99% of the time, they won't be audible. However, why would these rear up their ugly heads only on radio #2? The station set-up is identical as I have only swapped radios.

I suppose I could roll back to FW v.2.1.18 on radio #2 and see if that makes any difference!

Juha

Comps Radio 1 and 2.jpg (211.58 KiB) Viewed 10840 times

[w-u-2-o]

You are on the right track. Put the same firmware into both boxes and see if things look similar. You can put both on 2.2.2 or both on 2.1.18, etc.

[JJ4SDR]

Thanks Scott!

I already had FW v2.2.2 P2 on radio #1 and it didn't really work very well. Will roll radio #2 back to FW v2.1.18 P2 and see if that makes any difference. With regard to both radios looking "similar," I want them both looking as radio #1 does currently.

Update:
I guess radio #2 does not have the same FPGA that radio #1 has as after rolling back to FW 2.1.18 P2 the Thetis can't find the HW. Is there a later version of FW v.2.2.2 P2 that I might want to try instead?

Update2:
MAJOR PROBLEM cropped up! I tried FW 2.1.26 P2 as from another conversation with Doug (he wasn't a 100% certain which FPGA radio #2 had), surmised that by "going by the firmware v2.2.2 and the circuit board it most likely a EP4CE115F29C8N". Looking at the S/N of radio #2 and the thread at viewtopic.php?f=32&t=4462, FW v.2.1.18 is what radio #2 would like to see!

I then decided to go back to the good-ole FW v2.2.2 P2 and SAME result, Thetis no longer detects the radio! NO FW works on radio #2!

I HOPE that I didn't BRICK radio #2. When using Bootloader, it does find the board and before programming it says "Erasing device...." or something like that and that it might take up to 90 seconds......ALL seemed to go well, with no errors reported. Exited, switched OFF radio and put Bootloader switch to OFF position, restarted radio. Thetis can't find radio #2 no matter which P2 FW version I try. I was expecting that Thetis would find the radio with the original FW v2.2.2 P2 programmed into it.

Update3:
Tried even rolling back to Protocol 1, but no joy. THEN, it hit me, I always hear a RELAY click upon starting Thetis (clicking on the "Power" icon) and there is no click of any kind. Perhaps this is a STUCK relay and has absolutely nothing to do with firmware! Would sure value comments, to save my sanity, again!! Also, the right hand red LED stays lit, a solid RED instead of a blinking RED, which indicates the "heart is beating." Put in my radio #1 and all is well! So, I am thinking radio #2 has developed a hardware problem (relay?).

Juha

[w-u-2-o]

It seems very unlikely that a hardware problem occurred so coincidentally in the ANAN. Let's assume it has not, for now.

It also seems like there are no basic mistakes being made, i.e. forgetting to move the bootloader switch, things like that.

A surprisingly common issue is a bad Ethernet cable. More than once it has happened that manipulation of the ANAN, for instance to reach the bootloaders switch, has damaged a weak Ethernet cable, but only damaged it in such a way that it no longer will support GigE. Remember that only two pairs are necessary for 100Mbit/s operation (bootloader), but all four pairs must be good for GigE (Thetis). Swap cables, and check for good GigE indications at the PC or switch port the ANAN connects to.

If it's not the cable, then In the spirit of Sherlock Holmes, maybe, just maybe, there was some confusion about what firmware was, and what FPGA is, in #2. It might be worth rolling through the three different versions of firmware for the three different FPGAs just to confirm that this is not a problem.

viewtopic.php?f=32&t=4462
viewtopic.php?f=32&t=4972

The other thing to look at is to reprogram the IP address in #2 using bootloader. What is your setup for IP addresses on the PC, #1 and #2?

[JJ4SDR]

Scott,

I use CAT8 (overkill) cables. Put in radio #1 in line this morning and all works very well. Using an unmanaged Ethernet switch (have been for more than 3 years now) capable of 1 GBit/s speed.

With regard to P2 FW, I tried several of them. Radio #2 had been running well on FW v2.2.2 for quite a while so I don't think it has to with FW. There is no confusion about the FW at all. I even tried Protocol 1 FW, but no joy.

IP address: Why would the IP address be a problem all of a sudden? I guess it is possible. NOTE: I only have one radio on the network at the same time. I have made copies of AppData\Roaming\OpenHPSDR for both #1 and #2 in a safe place. Now, when I swap radios, I purge out the entire AppData\Roaming\OpenHPSDR for the radio I WAS using from the PC. Before I put the other radio in line, I move the entire AppData\OpenHPSDR Folder for that radio from that "safe place." Is this an incorrect way of doing it?

Thanks!
Juha

Below, see Hardware, Network Forms and Bootloader View:

[JJ4SDR]

I tried using Static IP address as well, just for grins, but no joy !

After having flashed a FW version on it and having switched the bootloader to the OFF position (included in the FW versions, v2.2.2 that worked beautifully just a few days ago) and powering up the ANAN, the right red LED stays solid, no blinking. It is blinking when I use the Bootloader to flash the FW as I believe it should.

As soon as I see that solid right red LED, with no "heart-beat," I know that Thetis won't detect the hardware. Also, there is no audible click when I am starting up Thetis. According to Doug: "That’s the firmware setting the relays to a known configuration. If the firmware is not running then no relay setup." (and, so no click/my add!). => So, the firmware(s) that I have flashed are not "running."

Update:
Bootloader switch is working. Switched it to OFF position, went into Bootloader and it didn't find the radio (board). Another good friend suggested that I check if all of the SMA connectors are snug. It is just such a bear doing that......

Juha

[w-u-2-o]

Does the Bootloader still find the board, or has that started failing, too?

[JJ4SDR]

Scott!

I am a happy camper!

Spoke with 2 other fellows about this as well. And, I must give Doug many thanks ! He was incredibly generous with his time and suggestions! It seems that the FPGA is a Cyclone V series one to boot which belies the S/N on the bottom of the radio. Some repair had been done in the radio so that threw me completely.

Your comment about the IP address did the trick! For a foolish reason - yes, I am not afraid to man up to that - when tried using Static IP at one juncture I likely reprogrammed radio #2 with the "wrong" address. I corrected that by putting all ZEROs in the IP address field and I am using Dynamic IP. Perhaps learning from my mistakes might benefit someone else.

Below, see results from tests I ran today. Now, I am hoping to settle on a working FW (not going to Protocol 1 though!) that will not generate those "spurs" that I reported in the first post.
Update: Made a mistake in the Excel table, but included the correct one below. The spurs are still there for all the working versions of FW though! No change at all. Also, CW is either "dirty" or "clean" depending on the FW used. "Surprisingly," I see "dirty" CW gets "cleaner" when PureSignal is engaged ! I was under the impression that PureSignal mainly improved SSB transmissions.

Juha

[JJ4SDR]

You cannot use readings from the panadapter to determine S-units except for CW:

https://apache-labs.com/community/viewt ... f=6&t=2463

My old Flex 3000, and now the ANAN, finally made me "comfortable", because RF system design is my day job. I despise the so-called metering on KenYaeCom equipment.

Scott,

Your comments regarding panadapter readings on CW vs S-units caught my eye! As you remember from my other post in another forum, I included a screen-capture (included here, for convenience), I am concerned over "spurs" I am seeing, both on CW and SSB. Radio #1 is MUCH cleaner than radio #2! Compounding the situation is that I suffer from OCD ! What is shown below is when I operate SSB.

Q1: See those 2 "spurs" around 15kHz above and below my TX frequency. From earlier threads having to do with PureSignal, I seem to recall that spurs like these will likely not be either seen or heard by anyone. Let's assume that I talk to a local fellow on 40 meters and I am S-9/+40dB at his station. Wouldn't he be able to hear/see those "spurs?" I will read the thread where you are discussing the difference between S meter readings and the levels in the Panadapter that you also posted in this thread.
Q2: When I am on CW, I can see these "spurs" as well and they are approximately -65dBc from my TX frequency. Wouldn't that mean those "spurs" can be easily heard by someone else in case I am putting out a very powerful signal. Say S-9/+40dB again!

I am thinking that especially on CW, I would be using a transceiver that will not pass FCC regulations......but, I think I need to revisit those!

Can you alleviate my concerns and worry?

Juha

[w-u-2-o]

Scott!

I am a happy camper!

Spoke with 2 other fellows about this as well. And, I must give Doug many thanks ! He was incredibly generous with his time and suggestions! It seems that the FPGA is a Cyclone V series one to boot which belies the S/N on the bottom of the radio. Some repair had been done in the radio so that threw me completely.

As I had suggested two days ago, when I wrote "then In the spirit of Sherlock Holmes, maybe, just maybe, there was some confusion about what firmware was, and what FPGA is, in #2. It might be worth rolling through the three different versions of firmware for the three different FPGAs just to confirm that this is not a problem".

Update: Made a mistake in the Excel table, but included the correct one below. The spurs are still there for all the working versions of FW though! No change at all. Also, CW is either "dirty" or "clean" depending on the FW used.

This is all a matter of how well the timing is closing in the FPGA as part of the physical gate floorplan, which will obviously be different in both FPGA variants, and a factor of minor differences between circuit boards. Hence it is not surprising that #1 and #2 provide different performance.

"Surprisingly," I see "dirty" CW gets "cleaner" when PureSignal is engaged ! I was under the impression that PureSignal mainly improved SSB transmissions.

RF chain linearity plays an important part in the spectral purity of the output regardless of what type of signal is being transmitted. However, as you note, any non-linearities become much more noticeable with a multi-tone waveform, and SSB phone is pretty much the worst case in that respect.

[JJ4SDR]

Scott,

I appreciate your comments. In another thread on "S meter calibration" - viewtopic.php?f=9&t=3981&p=31655&hilit=s+meter+calbration#p31655 - I also touched on the level of "spurs" (CW vs SSB) and my concerns. I am not sure if there is really anything else I can do in an effort to get rid of those "spurs."

73,
Juha

[w-u-2-o]

Q1: See those 2 "spurs" around 15kHz above and below my TX frequency. From earlier threads having to do with PureSignal, I seem to recall that spurs like these will likely not be either seen or heard by anyone. Let's assume that I talk to a local fellow on 40 meters and I am S-9/+40dB at his station. Wouldn't he be able to hear/see those "spurs?" I will read the thread where you are discussing the difference between S meter readings and the levels in the Panadapter that you also posted in this thread.

Some assumptions and approximations are necessary to make this calculation in a straightforward way.

- The other fellow is using a similar radio, e.g. a Flex or an openHPSDR type radio like Apache
- Let's assume a conservative 2700Hz receive passband
- Let's approximate your SSB phone signal as having a nominally flat spectrum across that 2700Hz

Obviously the other station will not hear them because they are well outside any reasonable receive passband width, even if it's set to 10KHz wide.

Will he see them? Let's do some math. Your S9+40 signal equates to -73 + 40 dBm, or -33 dBm in a 2700Hz passband. We have to scale that to the resolution bandwidth of the spectral display, let's say 3Hz. Rounding to the nearest dB, 2700Hz = 34dBHz, 3Hz = 5dBHz, thus your main signal in the passband will be peaking around -33dBm (2700Hz) - (34-5dBHz) = -62 dBm on the spectral display.

With those spurs being approximately -50dBc, that means they'll be around -112dBm on the spectral display. Without going through more math, he'll see those starting to poke above the noise floor with an S7 noise level (again, in a 2700Hz passband).

Q2: When I am on CW, I can see these "spurs" as well and they are approximately -65dBc from my TX frequency. Wouldn't that mean those "spurs" can be easily heard by someone else in case I am putting out a very powerful signal. Say S-9/+40dB again!

Maybe not "easily". Achieving +40 at another station is not typical.

I am thinking that especially on CW, I would be using a transceiver that will not pass FCC regulations......but, I think I need to revisit those!

Can you alleviate my concerns and worry?

Juha

From a US regulatory perspective there is no problem. See https://www.ecfr.gov/current/title-47/section-97.307, specifically sub-part (d).

[JJ4SDR]

Scott,

Your consistent feedback is very much appreciated. Will look at the math in detail.

When I am talking S-9/+40dB, I am referring mostly to local contacts on 40m where I am extremely strong, on receivers with calibrated S meters. On 20, I also do get very strong signal reports, +30dB every now and then and obviously, depending on the RX at the receiving end, my signal may be actually way less than that!

It is just that it "bothers" me that radio #2 is not as clean as radio #1 which is VERY clean in comparison!

Juha