EQMOD Control Failure

[BlueAstra (Graham)]

I have been unable to resolve my communications problems with my EQ6 mount using the ASCOM/EQMOD/EQDIR configuration. I have tried the following equipment in almost every possible combination:

ZBOX HD-1D11 Nettop (Atom D510 @ 1.66GHz, 2GB ram, 64GB SSD, W7 32bit HProf)

ACER 3950 Desktop PC (Intel core i3 550, 2G ram, 500G HDD, W7 64bit H Prem)

Hitec USB-EQDIR Module
Shoestring Serial-EQDIR Module
Prolific based USB-Serial adapter (PL2303 driver, 12.3.10, 3.3.11.152)
FTDI based USB-TTL cable (32/64bit driver CDM20802)

ASCOM v5.5.1, EQMOD v1.23e
MAXIM DL v5.1.4, Starry Night Pro Plus v6.0
I normally have other equipment connected (EG, focusser, powered hub, cameras) but these have all been disconnected for this test. Only the EQDIR module or FTDI cable is connected.

The basic problem is that I can setup almost any combination of the above equipment and connect successfully to the mount. However, during a slew to a target at some point EQMOD will indicate a communication loss and try to send stop commands. The mount will continue to the target (the stop commands are ignored) and stop, but no further comms is possible unless I exit MAXIM or SN and restart the programs. The one exception to this is the FTDI USB-TTL cable which connects, but after a short pause EQMOD shows connection errors without any further action. I have tried changing port parameters without success.

The only successful means of communication is with a USB-Serial adapter through the handset in PC Direct mode. This works without a problem, but I do notice single comms loss events very occasionally. This is less than the 5 consecutive errors which triggers a stop, so it has shown to be effectively 100% reliable.

The only other thing I’ve noticed is that the EQ6 red power light flashes on/off during a slew (handset or EQDIR), and sometimes flickers. The PSU is the one supplied with the EQ6 (12V, 6A). Not sure if this is significant or not.

I have now run out of ideas trying to find a solution to this. I am trying to find a setup that will suit a DIY observatory, and having to use the handset during setup is inconvenient and does not follow the EQASCOM guidelines. However, it seems I haven’t a choice. I would appreciate if anyone has any further suggestions.

[danielsun]

Just a thought. Are you using non powered USB cables longer than 5 mtrs?

Cheers Daniel.

[BlueAstra (Graham)]

No, they are all 1.5m USB with the PC close to the mount.

[BlueAstra (Graham)]

The voltage at the mount (measured with the handset) is 11.7 V. An RA slew at rate 9 drops this to 11.1 V, and a RA+DEC slew drops it to 10.4 V, with a flashing red LED. I found the primary cause was a power extension cable that I was using since the power lead from the PSU was fairly short. If I connect the PSU directly to the mount I get 12.2 V at the mount and 11.7 – 11.9V with a RA+DEC slew. The red LED is now solid throughout the slew, that’s one problem solved.

However, unfortunately it doesn’t seem to have solved the communications issue. Is it possible the low voltage operation has somehow damaged the mount circuit board, causing communication issues? It still seems to work normally using the handset alone, or in PC Direct mode from the PC.

[RobF (Rob)]

Hi Graham

My EQMOD homemade cable for the HEQ5Pro (TTL/USB cable with RF45 connection on the end) start to misbehave a couple of months ago. The mount was getting lost or dropping connection during slews as you describe. I'm pretty sure it was the stretching and moving of cables upsetting it. In anger one night I pulled out my spare TTL/USB cable, cut the end off and spliced it on to a a commercial Cat 5 cable with reinforced RJ45 plug. Has been fine since.

The flashing light is supposed to signal low voltage, so probably worth looking for a high amp power supply. I use a cheapish 3A from local electronics stores and behaves ok.

The fact it all works through the handset sounds like a connection problem to me, not the mount controllers.

[BlueAstra (Graham)]

Well, eventually I may have found a smoking gun. Today I tested the system with an old desktop PC with a built in serial port. Connecting this just to the Serial EQDIR adapter led to the usual effect: connects OK but looses comms during a slew to target. I regressed the software to earlier versions of Ascom/EQMOD without effect. However, I then put an in-line capacitive filter between the serial cable and the EQDIR module, and would you believe it everything worked fine. I removed the filter and the problems returned.

So the issue may be noise pick up in the serial cables, and the 'working' PCs I had earlir may have been just because the cables lay differently. I've set the mount to track now for a few hours to check if any comms errors appear, so fingers crossed!

Graham

[RobF (Rob)]

Thanks for reporting back Graham. MIght be another mystery of the EQ mounts deciphered. Nothing worse than your mount heading off wherever - great value for money platform one you get it figured though.

[BlueAstra (Graham)]

Another false lead! Just to recap, I’m trying to control my EQ6 mount locally with a small PC which is controlled via remote desktop from the house PC. The local PC is a ZBOX nettop and the house PC is the same (see earlier posts for spec). My problem with this layout is that with just about any combination of Serial EQDIR / USB EQDIR / Keyspan USB-Serial adapter on the local PC I get a communication loss either just after connecting or during a slew to a target. The control software is Maxim DL 5.14, Ascom 5.5.1, EQMOD 1.23e. However, if I use the handset in PC direct mode, the system works effectively, with just the occasional single comms error that doesn’t affect operation. It’s essential that I get the system working reliably if I am to progress to an observatory build. I have taken Maxim out without any improvement.

I’m sure the problem is in the local PC, but I can’t find it. The old desktop with a serial port does work with the in-line filter, and not reliably without it, but the filter doesn’t improve the other setups. I did actually get a single comms error event during parking with the desktop. I borrowed/have some other computers that work:

HP TX-2 laptop (AMD Turon x2dual core RM-74 @ 2.2GHz, 3G RAM, W Vista SP2 32bit) running Maxim 5.14, Ascom 5.0a, EQMOD 1.19g.

Acer Aspire Netbook (N270 @1.6GHz, 1G RAM, W7 Starter) running Maxim 5.1.4, Ascom 5.5.1, EQMOD 1.23e

I’ve noticed with the working PCs that the slew graphic in Maxim moves in regular jumps to the target. With the local PC, the jumps are very irregular in the failure setup, and slightly less irregular in the PC direct setup. Clicking on the Ascom logo in EQMOD for the command window shows the comms hesitates in synchronism with the irregular graphic jumps.

The Keyspan adapter includes a useful port monitor. This is an excerpt close to the comms error. ‘Buffer overrun’ doesn’t sound good, but I’ve no idea where the buffer is it’s talking about. You do get them on the good PCs, but maybe not as many.

18711: 17896 (80486 39.436.869): writeCompl (0/0)
18712: 17897 (80487 39.436.869): unknown 71 port (5/-2032336724) code 0 len 14 p1 0
18713: 17898 (80488 39.436.869): serdGetBuf (0/0) len 0 In 0 Out 0
18714: 17899 (80489 39.436.869): read/partial (0/0) req 1 act 0 (0)
18715: Events Request returned 5600 bytes (100 Records)
18716: *** Buffer Overrun ***
18717: 17900 (80495 39.452.469): read/complete (0/0) len 1 (00000033)
18718: 17901 (80496 39.452.469): serdStatus (0/0) MSR 00 cumErrs 00

So the investigation continues. Will keep you posted.

[Visionoz (Bill)]

Hi Graham

In EQMod I believe that there is a setting that you can use to prioritise its activities over everything else that the PC/lappie needs - set it to "High Priority" so that EQMod would get greater time-slices

Perhaps posing the question of your problem in the EQMOD Yahoogroups might help too

HTH
Cheers
Bill

[BlueAstra (Graham)]

Investigation update. I think I’ve found a significant issue with the mount power supply. I’d previously measured the output of the supply as 12.2 (static) – 11.7 (during slew). The red light stayed on continuously, so I thought everything was OK. However, measuring the PSU directly showed it was oscillating between 11.3 – 12V over a period of about 1-2s. This didn’t seem to register on the handset, but was clearly visible on a multimeter. I obtained a stabilised OEM power supply (TDK NV-175, 12V 15A) as a replacement and set it to 13V.

The typical fault in EQDIR mode was the occasional comms loss just after connection, and an almost guaranteed comms loss during a slew. This reduced to an occasional single event comms loss when using the handset in PC direct mode, which didn’t provoke a stop command. With the 13V OEM supply powering the mount in EQDIR mode I didn’t have a single comms loss during the entire session, with plenty of slews and 3 hours of tracking.

It’s not quite clear cut because I re-connected the original PSU and got hardly any comms losses. However, I did get one or two single events, and the mount did pause during a slew a couple of times. Maybe I’d connected the plugs more reliably. So I think the root cause may be cyclic low voltage from the original PSU triggering comms drop-outs, aggravated by plug connection issues. I’ll monitor the system over the next few nights to confirm if this solution works.

Graham

[RobF (Rob)]

Great to hear Graham. I usually wedge my 12V power connection in tight against my RJ45 connection, but recently ran off batteries for 2 nights and had a couple of dropouts myself until I taped down the power lead to maintain some tension on the 12V plug on the mount. Any sign of your red power light blinking if you wiggle the 12V connection is a sure sign of these types of problems.

[BlueAstra (Graham)]

I thought it would be useful to summarise my experiences in trying to track down and cure communications problems with my EQ6 Pro mount when used under ASCOM/EQMOD control. It might help other users who have similar issues, and point to areas for investigation. The help from the EQMOD, ASCOM and EQ6 Yahoo forums was also invaluable.

The equipment setup is a Nettop ZBOX HD-1D11 (Atom D510 @ 1.66GHz, 2GB ram, 64GB SSD) running XP, Ascom v5.5.1, EQMOD v1.23e, connected via USB to a Hitec EQDIR module, which connects direct to the EQ6 input. All other equipment was disconnected during the tests.

The typical fault in EQDIR mode was the occasional comms loss just after connection, and an almost guaranteed comms loss during a slew. This reduced to an occasional single event comms loss when using the handset in PC direct mode (USB-Serial adapter – Handset port), which didn’t provoke a stop command. The communication status can be seen if you click on the ‘display +’ button (top left) on the EQMOD panel, and/or the Ascom symbol which opens a comms terminal window. The mount worked perfectly normally if you just used the handset.

The first suspect was the prolific chipset used in the Hitec EQDIR adapter. Prolific chipsets are not noted as the most reliable by the EQMOD designers, and XP did not allow access to the latest drivers (3.3.11.152, which are for Vista onwards). I upgraded to W7 on the nettop and downloaded the latest drivers from prolific, but it made no difference. I then tried a Shoestring Serial EQDIR module with an Astronomiser USB-Serial adapter, but this also used prolific and had the same problems. Next up was a FTDI based USB- 5V TTL cable, same problem. Finally, the gold standard, a Keyspan USB-Serial adapter, which unfortunately also gave the same communication issues.

So the problem was probably not associated with the computer – EQDIR link, so I started to look elsewhere. I had noticed that the red LED on the mount flashed occasionally during slew. In my case I was using a free ‘12V 6A’ PSU supplied with the mount, through an extension cable to the mount. The handset said I was getting 11.7V, dropping to 10.4V during a RA+DEC fast slew. This gave the flashing LED during the slew. When I removed the extension cable I got 12.2V static and 11.7V during a slew, and a solid LED. However, I still got the communication failures. I removed the power connector at the mount and measured the voltage with a multimeter held directly against the jack plug. This showed the voltage varying from 12.1 – 11.3V over a period of about a second or so, even with nothing connected.

Many people use this mount and EQDIR/EQMOD with no issues, and many people use a power tank to power it. Lots of people also use laptop style power ‘bricks’ to power the mount. Power tanks/batteries run at about 13.8V, so I wondered if the low voltage, or the cyclic nature of it, was contributing to the comms error. I replaced the PSU with a stabilized supply set at 13V and the comms errors have gone away! If I were to guess I would say the cyclic power fluctuation triggered a comms error during the voltage ‘dip’. This would be aggravated during a slew since the overall voltage dropped when the motors ran. So finally I believe the problems were caused by a faulty power supply, and cured with a good regulated supply.

The supplied power brick ran the mount perfectly with just the handset, so you would never notice a problem under these conditions. Also, if the mounts were not tested with EQMOD/EQDIR prior to delivery, the supplier would never notice it either. The problem only appeared during direct EQMOD/EQDIR control. The problem would also not occur if you used a power tank or battery, since they tend to run at about 13.8V. Of the USB-Serial adapters, they all subsequently proved to work. The Keyspan is probably best, and it comes with lots of useful diagnostic software. However, I now used the FTDI based USB- 5V TTL cable as my EQDIR interface, very cheap and reliable.

So if you get comms issues with EQMOD, check your power supply first, then move on to the various bits of hardware/software. Post queries in the various forums. The cable based USB- 5V TTL EQDIR solutions are a good method for linking your computer to the mount. To have a reliable EQMOD setup allows you to realise just how good this software is for mount control.

[RobF (Rob)]

Great work Graham. Hope you've got it licked now.
EQMOD really is an amazing piece of work isn't it, and for free!

I've only recently noticed the "Pierside Only" Point Filter for Alignment/Sync - has made a huge difference to the quality of my pointing for those of us that can't afford Tak, AP, etc

[Peter.M]

Hm, I have been having issues with my EQ MOD disconecting after an indetermined number of slews. I sure hope that this is the problem with mine. Where did you end up getting your regulated power supply from?

[bmitchell82 (Brendan)]

Ill second that Rob, I have a velcro cable tie that i put around the RS 232 plug and the power. then on the second time around all the power that goes up to the cameras goes in there.

[Peter.M]

I watched the mount during the connection issues today, and the red light is solid the entire time. I have been using an old usb to serial through the synscan controler and at the moment I am suspect that it works properly, so I just ordered the shoestring astronomy heq5 to usb inteface. When I get that I should be able too cut out the middle man, and the dodgey usb to serial convertor.

Have to see how i go

[Peter.M]

Got my eq5 usb2 inteface today, and it has fixed all my conection issues. Now im set up with eqmod and am not getting drop outs I dont think i will ever plug the handset in again.

[RobF (Rob)]

Great news Peter!
Aweful feeling when you don't have confidence in your mount.