7000DLE MKII Frequency Drift

w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

7000DLE MKII Frequency Drift

Postby w9mdb » Tue Apr 20, 2021 2:46 pm

I run a monthly frequency measurement test. Have been using an ANAN 100 with a Bodnar Mini GPS as the 10MHz sync and an Omni VII with another Bodnar Mini GPS as a reference frequency. FLDigi provides the means to use the reference frequency to very accurately measure a frequency. In this setup I can usually measure my transmit frequency to 0.001Hz standard deviation.

I just set up my new 7000DKE MKII and I'm seeing way too much frequency drift on the 7000 trying to measure a reference frequency. I've got the Bodnar on 10MHz fed into the 10MHz jack on the rig and the 2nd Bodnar on antenna 2.
In the best of worlds this line should be flat.

I also noticed a huge frequency drift occurring during FT8 decoding....totally unacceptable.
So what's the story with the 10MHz reference -- is there something else I need to do to enable it? Is the Bodnar square wave no good for it? The manual says nothing about it.
I'm hoping there's a fix for this as I don't want to be disappointed with my purchase. I do have a BG7TBL I can hook up too...so will try that next.

Mike W9MDB
Capture.PNG
Capture.PNG (23.54 KiB) Viewed 9458 times
Mike W9MDB
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Tue Apr 20, 2021 6:35 pm

There could be two different problems going on here:

1. The ANAN is not locking to the GPSDO.

- This could be a small matter of jumper settings. How are your clock related jumpers set up on you ANAN-100?

- It is better to test this by splitting the output of the GPSDO and using the same signal as both a frequency reference and the observed signal. If no drift is observed then you know for a fact that the lock is occurring. By bringing in another receiver and another GPSDO you cannot be certain the problem does not lay elsewhere.

- If you can do the splitter approach, use a totally trustworthy signal like WWV to check your calibration.

- In either case, before resorting to VAC and external software to test frequency stability, the first test is to use the built-in features of Thetis:

* Set General > Calibration > Correction Factor to 1.0000000 if it is not already.
* Tune to WWV; DSB mode; all receiver processing (NR, ANF, NB, EQ) turned OFF.
* Select Phase display mode (instead of Panafall, etc.)
* If you are properly locked to your GPSDO, then the little burst of phase noise should not be circling the display, it should be almost stationary.
* If you move 1Hz off of WWV, then the phase burst should circle the display at one revolution per second; i.e. how fast it is circling is how far you are off in frequency.

- Have you updated the GPSDO to the latest firmware, which is 1.16?

- Review this topic, which has a lot of good info on signal levels, jumper settings and stuff: https://apache-labs.com/community/viewtopic.php?f=17&t=2499

2. If you are using VAC to output audio your resampler might not be stable, and this can result in massive drifts even if your GPSDO is working perfectly and locked to the ANAN.

- There is a lengthy discussion on these problems and how to achieve a very high level of performance and stability be either a) disabling the resampler or b) adjusting your VAC settings to get stable resampler operation.
K1LSB
Posts: 639
Joined: Wed Feb 05, 2020 5:25 pm

Re: 7000DLE MKII Frequency Drift

Postby K1LSB » Tue Apr 20, 2021 8:40 pm

Scott, can you please clarify how to disable the resampler? is that done by simply checking the "Force" boxes? I think you told us once but I've forgotten..
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Tue Apr 20, 2021 9:37 pm

K1LSB wrote:Scott, can you please clarify how to disable the resampler? is that done by simply checking the "Force" boxes? I think you told us once but I've forgotten..

Yes, that's correct. Also take a look at the link to the resampler topic I posted above.
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: 7000DLE MKII Frequency Drift

Postby w9mdb » Wed Apr 21, 2021 3:59 am

Resampler was disabled
Capture.PNG
Capture.PNG (4.52 KiB) Viewed 9401 times


Is there some jumper I have to change?
Mike W9MDB
W4WMT
Posts: 325
Joined: Sun Apr 09, 2017 10:12 pm

Re: 7000DLE MKII Frequency Drift

Postby W4WMT » Wed Apr 21, 2021 9:33 am

Hi Mike,

Were those overflows & underflows static, or were they accumulating on you while you were measuring your drift?

Do you know how to click on the counters to reset them back to zero? That helps you to keep track of recent events.

You have to get these counters to zero, over long stretches of time (minutes), or all bets are off for the kind of freq stability you are looking for.

73, Bryan W4WMT
User avatar
Tony EI7BMB
Posts: 651
Joined: Sun Apr 09, 2017 2:31 pm
Location: Dublin
Contact:

Re: 7000DLE MKII Frequency Drift

Postby Tony EI7BMB » Wed Apr 21, 2021 9:59 am

Hi Mike, I have the 7000 mk 2 and did not have to change any jumper. Just connected the GPSDO to the input on the radio.
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Wed Apr 21, 2021 10:36 am

Tony--he has an ANAN-100 with a Hermes board. Switching is not automatic on that board. He will have to look at the schematic and move internal jumpers.

Schematics are here: viewforum.php?f=35
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Wed Apr 21, 2021 10:38 am

W4WMT wrote:Hi Mike,

Were those overflows & underflows static, or were they accumulating on you while you were measuring your drift?

Do you know how to click on the counters to reset them back to zero? That helps you to keep track of recent events.

You have to get these counters to zero, over long stretches of time (minutes), or all bets are off for the kind of freq stability you are looking for.

73, Bryan W4WMT


Bryan--his resampler is disabled. Those counts are therefore meaningless.
User avatar
Tony EI7BMB
Posts: 651
Joined: Sun Apr 09, 2017 2:31 pm
Location: Dublin
Contact:

Re: 7000DLE MKII Frequency Drift

Postby Tony EI7BMB » Wed Apr 21, 2021 10:53 am

I think that was Mike's previous rig unless I'm misreading Scott

w9mdb wrote:I just set up my new 7000DKE MKII and I'm seeing way too much frequency drift on the 7000 trying to measure a reference frequency.
I'm hoping there's a fix for this as I don't want to be disappointed with my purchase. I do have a BG7TBL I can hook up too...so will try that next.

Mike W9MDB


w-u-2-o wrote:Tony--he has an ANAN-100 with a Hermes board. Switching is not automatic on that board. He will have to look at the schematic and move internal jumpers.

Schematics are here: viewforum.php?f=35
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Wed Apr 21, 2021 11:11 am

My apologies to you, Mike, and to you, Tony. I did misread Mike's original post and did not catch that he had moved on to a 7000.

The 7000 should, of course, automatically switch to the external reference.

BREAK

Mike--I would encourage you to go back up to my post above and, leaving off the issue of jumpers, individually address all of the other points I made, starting with the firmware revision in your Bodnar unit, and moving on from there. I see you've already addressed turning off the resampler.

Have you looked at the calibration against WWV (or a split 10MHz signal) as I've already suggested?
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: 7000DLE MKII Frequency Drift

Postby w9mdb » Wed Apr 21, 2021 11:43 am

I have both an ANAN 100 and 7000DLE MKII now. Just got the MKII and setting it up and testing.
The Omni VII I've been using is able to measure the Bodnar to within 0.001Hz std deviation even with it's frequency drift. You may not be familiar with FLDigi's FMT capability....I worked with Dave W1HKJ to implement it. It's uses by most of the FMT participants now.

Check out the drift on the FT8 watefrall....this is thermally related even though I'm just putting out less than 2 watts (amp does the rest.
This can't be oscillator drift....it's like the frequency reference has no reference at all.

I'm on firmware 2.0.

Capture.PNG
Capture.PNG (24.55 KiB) Viewed 9369 times
Mike W9MDB
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Wed Apr 21, 2021 12:07 pm

Mike,

You know I am familiar with the FMT stuff, you and I previously corresponded on it here:

viewtopic.php?f=9&t=3400&p=10961

I've already made a number of suggestion on how to debug this earlier in the topic. I'll repeat them here:

The most probably issue is that the ANAN is not locking to the GPSDO.

- It is better to test this by splitting the output of the GPSDO and using the same signal as both a frequency reference and the observed signal. If no drift is observed then you know for a fact that the lock is occurring. By bringing in another receiver and another GPSDO you cannot be certain the problem does not lay elsewhere.

- If you can do the splitter approach, use a totally trustworthy signal like WWV to check your calibration.

- In either case, before resorting to VAC and external software to test frequency stability, the first test is to use the built-in features of Thetis:

* Set General > Calibration > Correction Factor to 1.0000000 if it is not already.
* Tune to WWV; DSB mode; all receiver processing (NR, ANF, NB, EQ) turned OFF.
* Select Phase display mode (instead of Panafall, etc.)
* If you are properly locked to your GPSDO, then the little burst of phase noise should not be circling the display, it should be almost stationary.
* If you move 1Hz off of WWV, then the phase burst should circle the display at one revolution per second; i.e. how fast it is circling is how far you are off in frequency.

- Have you updated the GPSDO to the latest firmware, which is 1.16?

- Review this topic, which has a lot of good info on signal levels, jumper settings and stuff: viewtopic.php?f=17&t=2499
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: 7000DLE MKII Frequency Drift

Postby w9mdb » Wed Apr 21, 2021 12:41 pm

Tuned to 5MHz -- correction set to 0
While doing this the rig drifted from about 20Hz too high to 20Hz too low.
And here's a video of the Phase
https://www.dropbox.com/s/0r5bs5ya26ep0 ... .webm?dl=1
Mike W9MDB
Capture.PNG
Capture.PNG (175.23 KiB) Viewed 9361 times
Mike W9MDB
W4WMT
Posts: 325
Joined: Sun Apr 09, 2017 10:12 pm

Re: 7000DLE MKII Frequency Drift

Postby W4WMT » Wed Apr 21, 2021 1:34 pm

w-u-2-o wrote:
W4WMT wrote:Bryan--his resampler is disabled. Those counts are therefore meaningless.


Even with the resampler disabled, we still want the ringbuffer to behave.
Each time the receive ringbuffer underflows, a VAC buffer's worth of samples gets dumped on the floor.
That's going to drive the drift measuring algorithm completly bonkers!

73, Bryan
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Wed Apr 21, 2021 2:31 pm

w9mdb wrote:Tuned to 5MHz -- correction set to 0
While doing this the rig drifted from about 20Hz too high to 20Hz too low.
And here's a video of the Phase
https://www.dropbox.com/s/0r5bs5ya26ep0 ... .webm?dl=1
Mike W9MDB

Mike--that measurement is no good. You have to be in DSB mode. Please make another video.
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: 7000DLE MKII Frequency Drift

Postby w9mdb » Wed Apr 21, 2021 3:27 pm

DSB mode makes no difference...but here 'tis...

https://www.dropbox.com/s/xakamn5qaprlb ... .webm?dl=1

Is there newer firmware than 2.0? Surely other people would be seeing this too from the recent shipments. I'm thinking I have a bum unit.

Mike W9MDB
Mike W9MDB
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Wed Apr 21, 2021 3:59 pm

Wow, that is bad. Obviously this problem needs to be fixed before you can start worrying about VAC, fine grain performance, and FMT measurements.

I'm going to ask for the third time: what firmware is in your Bodnar unit?

Also, what output level have you selected on your Bodnar unit, in dBm, please?

And, finally, what happens when you remove the external 10MHz input and rely solely on the internal reference? Can you use the adjustment provided in Setup > Calibration > Correction Factor to at least get close? Before I got a GPSDO I was able to get within 1Hz of WWV for at least short periods of time before thermal effects saw it drifting. If you can achieve that, then we can continue to narrow down where the problem is.
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: 7000DLE MKII Frequency Drift

Postby w9mdb » Wed Apr 21, 2021 4:09 pm

Bodnar has 1.16 and have been measuring this thing to 0.001Hz since I've had it.

32mA on the Bodnar and turning it on/off shows no change in the phasing.

Could something have popped loose inside?

Would opening it void the warranty?
Mike
Mike W9MDB
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Wed Apr 21, 2021 5:02 pm

Mike,

First see if you can get the unit reasonably close without the GPSDO. As I wrote above:

"What happens when you remove the external 10MHz input and rely solely on the internal reference? Can you use the adjustment provided in Setup > Calibration > Correction Factor to at least get close? Before I got a GPSDO I was able to get within 1Hz of WWV for at least short periods of time before thermal effects saw it drifting. If you can achieve that, then we can continue to narrow down where the problem is."
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Thu Apr 22, 2021 9:10 pm

Mike--did you ever try the test without the GPSDO?
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: 7000DLE MKII Frequency Drift

Postby w9mdb » Thu Apr 22, 2021 9:14 pm

Yes...the DSB phase display is ever so slightly worse without the GPSDO.

Same problem with drifting....I see 30Hz drift in the FT8 waterfall.

I've contacted Apache support for a warranty claim..has to be a bad rig....
Mike W9MDB
W4WMT
Posts: 325
Joined: Sun Apr 09, 2017 10:12 pm

Re: 7000DLE MKII Frequency Drift

Postby W4WMT » Fri Apr 23, 2021 9:20 am

FWIW, I run a dual ported Bodnar GPSDO on my new 7000. Works fine. The WWV DSB Phase display swims around two or three times per second with nothing connected to the 10Mhz ref input, then locks right in when the GPSDO is connected.

73, Bryan W4WMT
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Fri Apr 23, 2021 10:17 am

w9mdb wrote:Yes...the DSB phase display is ever so slightly worse without the GPSDO.

And you are unable to correct it with the calibration factor in setup?
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: 7000DLE MKII Frequency Drift

Postby w9mdb » Fri Apr 23, 2021 11:58 am

Doesn't matter what the calibration is -- it's drifting +/-30Hz with or without the GPSDO.
I'm 100% confident this is a bum rig.
Mike W9MDB
w9mdb
Posts: 446
Joined: Sun Apr 09, 2017 5:53 pm

Re: How to determine that an external reference is active?

Postby w9mdb » Fri Jun 11, 2021 3:25 am

I had my new 7000LE MKII repaired as it couldn't sync at all and was drifting horribly (50Hz or so).
Now that I have it back it's better...but still not good by any stretch.
Wild excursions happening. Never saw anything like this on the ANAN 100.
I sure can't be the frequency reference for the ppFMT I've been running with this behavior.
And spending another $80 to send this lemon in for repair again is not something I look forward to...especially since it seems it wasn't repaired correctly the 1st time.
The circular pattern in the phase plot just slowly tracks around the inner circle.
Capture.PNG
Capture.PNG (316.77 KiB) Viewed 8358 times
Mike W9MDB
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: 7000DLE MKII Frequency Drift

Postby w-u-2-o » Fri Jun 11, 2021 11:17 am

Slowly tracking around is normal. That's from ionospheric induced Doppler. Can you post a video of that?

Assuming the radio is now ok it may be that you have a virtual audio problem to solve.
User avatar
kc2rgw
Posts: 165
Joined: Mon Jun 22, 2020 5:44 pm

Re: 7000DLE MKII Frequency Drift

Postby kc2rgw » Fri Jun 11, 2021 11:57 am

That diagonal drift is something I ran into with fldigi early on when I was figuring out my Anan.

My “fix” was to ensure VAC and everything in the chain was 48kHz, buffer to 2048 and clicking “force” on the resampler. I left the buffers on auto and I was good to go at that point.

I have a 7000 but I don’t think this is a hardware thing, I think it’s a sampling rate issue. I’ve been rock solid ever since I set up this profile and saved it as “fldigi” which I toggle to for digital mode use with my VAC.

The fldigi WWV calibration mode is handy to run because you can see the line that develops go from angled to straight when you get the toggles toggled correctly.

Return to “Everything Else: Antennas, Relays, Switches, Power, Grounding, Cooling, etc.”