"Sticking" Relay on ADC 1 and 2 Receive

User avatar
NJ2US
Posts: 106
Joined: Sun Oct 29, 2017 9:39 pm

"Sticking" Relay on ADC 1 and 2 Receive

Postby NJ2US » Sat Dec 30, 2017 2:50 pm

I have Anan 7000DLE Serial # 19, starting to notice that occasionally, my Receive sensitivity drastically drops during a PTT cycle, almost like you expect from dirty relay contacts somewhere in the front end. A quick PTT cycle clears the issue , and sensitivity returns to normal. Happened several times in the past 24 hours. It is happening on both ADC's but I was given advice not to use "Gnd BPF2 on Tx" possibly due to this problem.
Last time it happened moments ago, I didn't cycle the PTT, I closed the OpenHPSDR application and restarted it, only to find the receiver on ADC 1was still in a reduced sensitivity state. A PTT cycle cleared the problem.

Do we have a hardware relay issue?

Jeff
NJ2US
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Sat Dec 30, 2017 3:01 pm

Could be, Jeff. Or it could be a problem on the control bus that commands the relays, or even possibly a firmware timing problem on your specific serial number.

As none of these problems are "user serviceable", I'd recommend you drop the factory a line directly. They will likely put you into touch with the US service center.

73,

Scott
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Sat Dec 30, 2017 8:22 pm

I've seen this myself on RX2 - a good number of times. One of these days I'll not be in QSO and I'll try to track it down leaving it that way... but definitely it loses the antenna and cycling the PTT corrects it.

Gary
K9RX

8000DLE
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Thu Apr 19, 2018 1:06 am

this issue CAN be quite problematic. I was working a new one - rather rare one on 160 meters - 7Q7EI on FT8 DXp mode the other day and the stupid relay HUNG on me during HIS attempt/call to me ... fortunately I was able to work him 10 minutes later or so ... I HAVE to figure out which relay is sticking on. This was using ADC2 through the separate BNC antenna connection connected to a vertical RX array. The signal is gone - nothing on the pan (well nothing other than the noise floor). I have to hit PTT to get it back and of course on FT8 that means a complete cycle MISSED.

If anyone else has found the answer please post here. I find it hard to believe this is hardware unless its a bug in the FPGA code.

Gary
K9RX
dj1yr
Posts: 52
Joined: Mon Apr 10, 2017 2:32 am

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby dj1yr » Thu Apr 19, 2018 7:43 am

Hello,

it's an SPI problem, I could control that with a Logicanalyzer.
I've programmed a simple sniffer for atxmega for such purposes, which sends the SPI data words per uart.
TX and RX load strobes are doing right, but the bits are not set anymore.
There is also a bug, if you activate XVTR HF, you have to change the band shortly before it works.

There is still a lot of work to do, Orion MK2 is still beta status for me, I hope it will continue soon.

regards René
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Thu Apr 19, 2018 12:50 pm

This has been a problem on every rig since the ANAN-100. The problem is multi-fold:

1. The method used does not demonstrate a high level of signal integrity. Current firmware writes the registers more than once in order to minimize the chance of missed bits. However this is a faulty approach because each write sequence is independent of the others, i.e. each write is not additive.

2. The method used is not true SPI, merely that some SPI lines are used to write into some serial to parallel registers. If it were true, 2-way SPI, whereby the registers could be read and confirmed, then any write-errors corrected.

3. There have been some problems with actual stuck relays, i.e. not serial-to-parallel write errors.
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Thu Apr 19, 2018 3:00 pm

As a developer that uses SPI on my products this doesn't sound .... well I guess I wonder if it is definitive or a hypothesis? There should be no reason why an SPI interface isn't 100%. If there were issues with the interface and more than one bit were being set as you indicate then one would speculate that it would also happen randomly to any other bits in that register... is that the case?

This to me is a show stopper. One doesn't spend $4000 for a radio where the RX goes DEAD part of the time. And if its been there for a long time it makes me seriously wonder if this is truly a "hobby radio", one for playing around and not taken wholy seriously or one intended for the main stream - guys NEEDING a radio to be reliable all the time. Being Apache Labs is the maker of the radio - whether they control the software development or not, I would hold them responsible for getting this fixed - and pronto. Not years.

Gary
K9RX
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Thu Apr 19, 2018 3:13 pm

Gary,

Go to the "schematics" sub-forum and download one of the PA/RF board schematics to understand the implementation in hardware. It doesn't matter which one, they are all similar.

73,

Scott
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Thu Apr 19, 2018 4:19 pm

Well I'm looking at it .... there are 4 serially connected 8 bit registers ... all are immediate transfer to the output it looks like (i.e. as its being loaded these bits are toggling as the data stream propagates for 32 bits ... odd when there is a transfer line). So RX1 has no issues. I don't know which relay in RX2 is getting left on(off) ... but the clock is the same - the serial data other than being cascaded is one line for both RX's ... I don't see how ONE bit can be affected by this design. If it were that the first two in line are uneffected and its something in the last 2 then that would imply other relays are being incorrectly set/cleared ... other than getting a meter on it I don't know how externally I can check that. Connecting a different antenna other than the RX array designed for 160 - in order to confirm if a different band/filter has been selected, would require changing bands to test which would in itself send another data stream.

Alternatively the output is latching up - I'll check to be sure the flyback diode is in place (although I believe the TPIC has this built in so not likely an issue - yes, they have clamp diodes).

Gary
User avatar
n1gp
Posts: 175
Joined: Sun Apr 09, 2017 6:34 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby n1gp » Thu Apr 19, 2018 4:53 pm

Hey Guys,

I saw what looked to be a bug a while back in the SPI.v module for the Protocol 2 firmware
when sending the 2nd retry and had fixed that.

When I saw this thread today I checked the latest Protocol 1 FW and found the same bug.
It attempts to send the Alex data twice, but didn't set the data pointer back to the start of
the Alex data. Instead I believe on the 2nd time it was sending random data but eventually
the data pointer would get to it's proper starting position and continue on with the SPI
transfer.

I made a new FW image with a fix for that.

I'm not sure this will fix your problem but worth a try. I'm assuming your using P1 FW. If so you can download
the test image here:

https://www.dropbox.com/s/ymvebbndfg6mz ... 9.rbf?dl=0

-Rick / N1GP
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Thu Apr 19, 2018 5:02 pm

Thanks, Rick! :)
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Thu Apr 19, 2018 8:24 pm

Super - I'll give it a try.

g.
dj1yr
Posts: 52
Joined: Mon Apr 10, 2017 2:32 am

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby dj1yr » Sat Apr 21, 2018 9:05 am

Hello,

with Orion MK2 48bits (6x shift registers) are pushed in series, the normal protocol knows only 32bit.
The problems do not concern Anan100 / 10 / Hermes, because they use the same SPI protocol as Mercury. The mistakes were also there and the transmission was slowed down.
So this probably only affects Anan 100D / 200D / 7000DLE / 8000DLE.
Also, the SPI DWord is not sent twice, as it is claimed here, only when changing one of the 48 / 32bits.

regards René
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Sat Apr 21, 2018 12:14 pm

I tried it for about 5 minutes (0 power) on 160 where I use the RX array .... it did not hang up.... I don't think that is necessarily a complete test as sometimes it would take a while (pretty random). I'll try again as time permits. Fingers crossed... I'll report back with more results. Of course its impossible to be 100% - that it will ever fail again - but if I can get in say an hour without a lock up that should be pretty definitive. If anyone else was seeing this - it takes using RX2/ADC2 - and has results I'd appreciate it if they posted here.

Gary
K9RX
User avatar
NJ2US
Posts: 106
Joined: Sun Oct 29, 2017 9:39 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby NJ2US » Mon Apr 23, 2018 5:40 pm

Applied the update and firmware on my 7000DLE is now reporting version 2.4

My ADC2 was “sticking” frequently, I’ll watch it for a couple days and see what happens.

Thanks Rick!

Jeff
NJ2US
User avatar
NJ2US
Posts: 106
Joined: Sun Oct 29, 2017 9:39 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby NJ2US » Mon Apr 23, 2018 6:36 pm

OK, so after about a half hour of operation, I notice a new anomaly. After several seconds of transmitting, I notice that my power output suddenly drops to ZERO. I release the PTT and key up, and normal power output resumes. It is very repeatable, happening about 10 times in a 20 minute period. This is with Rick's Version 2.4 loaded on my 7000DLE.

Jeff
NJ2US
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Mon Apr 23, 2018 8:58 pm

Does it do this with the "normal" firmware, Jeff?
User avatar
NJ2US
Posts: 106
Joined: Sun Oct 29, 2017 9:39 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby NJ2US » Mon Apr 23, 2018 9:14 pm

Ricks latest version is 2.4, the current generally available release is 2.3.

I went back to 2.3, and so far, the Transmit power to zero problem is not occurring. I’ll continue to test through this evening.
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Tue Apr 24, 2018 12:44 pm

I'm also using Rick's beta (which does show as 2.4 even though the embedded version in the file name says 2.3) ... I've not had a chance to go back and test for the sticking relay issue. I will say I've been running it since Saturday - 3+ days albeit with conditions not a great deal but i've made probably 20 - 30 contacts ... no issues with power dropping off here.

Gary
K9RX
User avatar
NJ2US
Posts: 106
Joined: Sun Oct 29, 2017 9:39 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby NJ2US » Tue Apr 24, 2018 1:00 pm

I went back to the production version 2.3, and the power drop off issue stopped, but I did have two ADC receiver drop off events during 3 hours of operation last evening (using 2.3) . Let me know if there is anything else I can help with.

Jeff
NJ2US
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Mon Apr 30, 2018 2:20 pm

Just a note since this thread is where i decided to upgrade the radios firmware. I had been using 1.7 given to me by Doug way back in September of 2017. The relay sticking problem might be fixed - I haven't revisited it since my post earlier in this thread.

However I note one serious issue that has come up. I'm going to start a new thread called: Version 2.3 firmware upgrade and CW issues on Anan 8000.

Gary
K9RX
User avatar
NJ2US
Posts: 106
Joined: Sun Oct 29, 2017 9:39 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby NJ2US » Mon Apr 30, 2018 2:37 pm

I bought my 7000DLE in November, I believe it was delivered with FW 2.1 Noticed the sticking relay immediately.
I'm back to 2.3, the power drop issue that arrived with Beta 2.4 stopped when I went back to 2.3, but the "sticking" issue returns. Hard to characterize how often it happens, but it is infrequent. Perhaps twice during an average 2-3 hour "Session".
It only seems to happen on ADC2.

I am convinced many operators are not noticing this problem, because not many are running continuous simultaneous ADC1 and ADC2 operations. I operate exclusively with "cans" ADC1 in left ear and ADC2 in right ear. 570 foot loop on ADC1 and a gang of four 600 FT Beverages on ADC2.

Or, I have a mechanically bad relay on the board.

One other data point:

When ADC2 fails, I have shut the software completely down, followed by a software restart, and the problem persists, and only clears when I either: 1) Initiate a PTT, or 2) Cycle ALEX

I have had the Sticking occur with ALEX both on and off, didn't seem to matter.

Jeff
NJ2US
K1wpo
Posts: 1
Joined: Sun Apr 09, 2017 3:13 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K1wpo » Thu May 10, 2018 4:49 pm

I have the exact same problem on my 7000DLE
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Thu May 10, 2018 9:36 pm

I'm now back on version 2.3 again ... and will soon reload 2.3(N1GP) to again test for the issue with ADC2 relay sticking. When I did that the last time and did limited testing it seemed to fix the sticking relay - however I found that CW TX wasn't working very well - I was unable to bring up RBN spots that normally would fill the screen ... I went back to 1.7 and lo-n-behold it was back again. Literally went from no RBN spots one minute and 4 minutes later after having loaded 1.7 a screen page (40 selected and filled) of spots. The issue on CW persisted for days by the way and impacted CQ's as well as responses to calls. I.e. It was real!

Since then because no one else reported a CW issue I went back and tried 2.3 again, this time deleting the file I had downloaded previously and re-downloading it... it has worked fine ever since on CW. Go figure. A frustrating thing indeed.

so I'm going to try to reload Rick's beta and retest. If any one else tests please let us know yeah OR neah on this. Works or doesn't at least regarding the sticking relay.

Gary
K9RX
User avatar
Conrad_PA5Y
Posts: 101
Joined: Fri Jan 19, 2018 8:11 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby Conrad_PA5Y » Fri May 11, 2018 8:41 am

I have the same issue here with v3.4.9. I bought the second hand ANAN-100D specifically for the 6m Es season. This thread does not indicate that there is an actual fix. Last night I could not use the radio in a contest because of the problem. I am amazed that this only happens with some radios and I do not feel inclined to do diagnostics on digital command lines, I do not have a logic analyser and I should not need one.

I think that for me I will have to use an external relay solution and some hard wiring.

Is there a reliable way of doing this?

Regards

Conrad PA5Y
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Fri May 11, 2018 11:27 am

Let's see if I have this straight:

Only affects ADC2 transmit-to-receive transitions, i.e. is only a factor for those using ADC2 (and only for radios equipped with dual ADCs).

Is fixed in the 7000 by Rick's experimental Orion MkII firmware, but the experimental firmware breaks some CW stuff.

Is that a good summary?

If so, this points to a firmware problem and not a hardware problem. Which also means that it is not repairable by sending radios back for service.

Thanks,

Scott
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Fri May 11, 2018 11:28 am

Conrad_PA5Y wrote:I think that for me I will have to use an external relay solution and some hard wiring.

Is there a reliable way of doing this?
Almost certainly there is. However, please describe exactly the switching that you need to occur, first?

Thanks,

Scott
K9RX
Posts: 414
Joined: Fri Apr 14, 2017 3:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby K9RX » Fri May 11, 2018 12:59 pm

Hi Conrad.... I feel for you. Indeed although it gets pooh-poohed by those that aren't using the same feature/functionality it IS important - and it simply should work.

That stated - I believe it is not PowerSDR ... rather it might well be the radio firmware. Perhaps the same hardware and methods are used in the 100 as are used in the 7K/8K and the fix is the same - it would just mean that firmware base would need to be changed similarly. I don't think this is hardware per se, hardware that is somehow "broken/intermittent".

stick with it.... (no pun intended)

Gary
User avatar
Conrad_PA5Y
Posts: 101
Joined: Fri Jan 19, 2018 8:11 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby Conrad_PA5Y » Fri May 11, 2018 1:00 pm

Hello Scott my setup on 6m is relatively simple. I use a mast mounted LNA with a very good quality RX/TX relay with 100dB + isolation on 6m. This really only needed on EME of course.

I then have 2 downleads from the mast, one for TX and one for RX. So the ANAN Ant 1 output goes to the PA and the RX lead goes to Ext 1 or 2, both suffer the same problem. Sticking when going from TX to RX. I never have a problem with TX output, it is always there.

I will have a look at the schematic and see if I can organise a dedicated RX input. I need to do the PS mod anyway so I will have a few SMAs on the back panel. I will use my IC7300 for the 6m Es season as I have very little spare time now.

As I have repeatedly said I am not particularly interested in SDR for the sake of it. To me it is a tool that should enable me to have very good TX quality and extremely good dynamic range. My hobby is not about tinkering with the innards of a radio that I thought was quite mature. I am very disappointed to have to put so much effort into it. However I will not abandon the platform and I will make it work for me. It is also extremely important for my microwave set up.

73

Conrad PA5Y
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: "Sticking" Relay on ADC 1 and 2 Receive

Postby w-u-2-o » Fri May 11, 2018 3:11 pm

Conrad,

Why use EXT1 or EXT2? Why not just use ANT1 and ANT2? TX on one and RX on the other. Or do you need full duplex operation?

If you don't need full duplex, then if you look at the schematic for the PA/RF board for your radio (you can find it in the schematics section of this forum), you'll see that there is a different set of chips and a different strobe line used to latch the registers for the ANT connections vs the EXT connections. It may be that you can get the ANT1 and ANT2 combination working where you can't get EXT to work.

Also, this is a completely different kettle of fish compared to the problems others are experiencing that is specific and peculiar to ADC2 operations on the 7000 and 8000. It may be that you do indeed have a hardware problem of some sort that is repairable and needs to be repaired. That would be my guess because I've had both a 100D (still have it) and a 200D and have put many, many thousands of cycles on the Bypass and EXT relays with no problems, as have many others.

I'd be very surprised if the ANT1/ANT2 combination does not work for you. This is the same chip that runs the TX/RX relay so we know that it must be working OK.

73,

Scott

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