Hello,
This is my first post, so be gentle.
I have been having trouble over the last couple of months with sequences crashing. The crash appears to occur when SGP attempts to communicate with my Lakeside USB Focuser. I have spent a lot of time with my configuration to ensure I have all the latest drivers/firmware etc, I have also upgraded/rerouted USB cables in an effort to minimise any noise or interference. All with no improvement. “Crash” is perhaps the wrong word, SGP continues to operate but incorrect focus operation.
The problem occurs in all the SGP Beta V2.6 range of builds and V2.5.1.17.
During an imaging run SGP appears to lose comms or have partial comms with the focuser, this can happen at either a filter change or at a temperature compensation point between frames. I am not using autofocus. I manually focus with a Bhatinov at the start of an imaging run using the In/Out Fine & Coarse buttons in the focus tab of the control panel and let the temperature compensation do its stuff at 5 steps/degree with auto adjust per filter set.
The In/Out Fine & Coarse buttons always function correctly and the temperature is always read and displayed in the Focus Information field correctly during the manual focusing stage. At apparently random filter changes or between frame temperature compensation points during the imaging run, the focuser can end up set with a value that maybe several hundred steps in error. I have also noticed that the temperature will may longer be reported but shown as NA. Sometimes the focus values can be seen slowly decrementing/incrementing to the desired value in bursts with large delays between. Sometimes the focuser appears never to reach its desired position.
I have tried adding up to 10s delays after a filter change and between images without any noticeable improvement.
Another thing I have also tried, is running the Lakeside focus control utility at the same time as SGP (but without SGP connected to the focuser), this has work fine with no comms issues. I have also tried swapping scopes to try a different stepper motor, also with no improvement.
In an effort to get to the bottom of the problem. I have run some test sequences with minimum equipment and with simulators selected ie not really imaging at all. The focuser and filter are directly connected to the PC and not through a Hub. Please see Dropbox - Error for the log files and sequence. Time 11:04:26 is when things start to go wrong.
I have sought assistance from Peter of Lakeside Astro, who has been extremely helpful. Peter has verified that Windows/Firmware/Ascom drivers are all correct. Peter has managed to recreate the problem with his set up and would be glad to get involved with resolving the issue.
Just a few comments to add the George’s post above regarding this issue. I did do some testing on this and like George did find some problems when switching between filters with the auto focus change check box enabled. What I was getting though, was that frequently when switching between filters SGP was crashing completely, whereas in George’s case SGP did not crash.
I was doing my tests with everything except the focuser running in simulator mode, so that may have had some bearing on the issue. What I did find when SGP crashed was that nothing had been sent to the focuser according to the focuser log and that the focuser was still up and running.
Just to complete the tests I then set everything, including the focuser, in simulator mode and was still able to make SGP crash when switching between filters as above.
That is really bizarre - I have been using Lakeside for the last three years with SGP. One thing though, I did have an issue with the temperature readout - the sensor is at the motor connector. I soldered a 1nF capacitor across the two pins of the sensor and it stabilized the measurement. Peter was going to introduce it in production. My filter wheel is part of the QSI camera and takes power from the 12V supply. Is your filter wheel powered by the USB cable? I’m wondering if the power drain is causing an issue. Before moving over to QSI I had a SX filter wheel and it and MDL had frequent crashes.
Buzz,
The focuser consists of a control box and a separate stepper motor for each 'scope. The control box is powered from a 12v psu. I have tried powering the control unit from my mount psu and a charger type psu(fully encapsulated module) with no improvement. So I am ruling out the power supply at this time. The control box interfaces with the PC via USB or the user can control the focuser manually with some buttons and view the current focuser position on a LED display.
I’ve been wondering whether SGP blows up because it initializes with a valid temperature reading but doesn’t receive one at a later read, then subsequently uses an erroneous value to calculate a temperature compensation value with a massive error.
Eweng - my setup is the same - I’m using the lakeside control module. The bit about the temp sensor - it is a three pin sensor which has been simplified to two pin. I was getting the N/A a few times on the readout on one of my motors and when I looked up the data sheet, I realised that while it should theoretically work, it could probably do with some local filtering. I slit open the heatshrink at the DB9 connector, popped in a capacitor (I say 1nF but it could have been 10nF) across the sensor pins and put a fresh bit of heatsink over it. The N/A went away, never to return. Depending on how old your unit is, there was a firmware update that fixed a power-up problem that sometimes didn’t initialize the USB. My unit is buried in a box, so I never use the buttons and need to take it apart to check the firmware.
Your addition of a small capacitor across the temperature sensor has been implemented on all units for some time now. We actually put a 100nf across the outer two pins of the temperature sensor. All three pins are used with the centre pin providing the data. It may be worth checking the one on George’s motor, but I still don’t think it would cause SGP to do what it is doing. The original problem was seen mostly where users had coiled up the spare motor ribbon cable to tidy up the wiring. Coiling up the excess motor cable made it more prone to interference and the capacitor sorted that out.
One of the symptoms of the interference was to reset the temperature sensor, and the other causing it not to be read. When the temperature sensor has been reset and is then read for the first time it sends -85 degrees C. This is deliberate so that the user can see that the device has just come on line. In the old days this -85 degrees C was used as if it was a valid reading, but once we saw what was happening we stopped it and sent the last known good value. The -85 only happened once so it didn’t cause a major problem. It also didn’t cause an issue generally other than causing a blip in the focus because the imaging program made an adjustment on it. It was also clearly visible in the Lakeside log which we are not seeing with this problem.
If the imaging program requests a temperature reading and the Lakeside driver gets an N/A (which means no sensor present) it ignores it and sends the last good reading because it knows the sensor is there from it’s initial checks. The driver only sends an N/A if it gets repeated responses from the Lakeside saying that there isn’t a sensor there which would mean that the sensor has failed totally or somebody has unplugged the motor cable. Once again, we would see the N/A in the Lakeside log file, and we are not seeing this.
When I did my testing I didn’t get as far as checking the temperature issue as I was seeing issues with simple switching of filters and where the focus was changed on filter change, as I said above. And in addition I seeing SGP crash at this time although I was using some simulated devices.
Regarding the other issue you mentioned, i.e. the problem with the Lakeside controller at startup, George has the latest firmware which fixes that issue.
Hi Peter - I was not sure how old Eweng’s system was - I have never had SGP crash when focusing with your ASCOM driver (as far as I know) over the last few years (Win7/10). I use a mixture of autofocus after changing filters and filter focus offsets. I’m using your 2.02 ASCOM driver.
Scratching my head to understand why Eweng has issues. My sequences typically have a 1 or 2 second delay between events. Could that be it?
Can I check if there was a resolution to this problem with SGP and Lakeside focusers? Run into trouble with the temp reading disappearing and then the focuser hanging at the start of auto focus after moving out to the limit. Does not happen all the time, but did last night I lost a few hours of capture.
Richie
No, the problem was never resolved. I have since moved over to auto-focussing where the problem does not occur. I ran the problem passed Peter at Lakeside (he’s very helpful) and he managed to replicate the problem. I quite like the thought of temp’ comp’, especially after an initial focus is achieved either manually or automatically.
Some nights temp’ comp’ would work fine but mostly it would hang after a random number of frames. The problem would also occur when only running the focuser (with everything else set up as a Simulator).
The problem also occurred when I tried an external USB Temp’ sensor. I’m pretty sure its a SGP issue.
Thanks, George. That’s a bit discouraging - spent all summer getting on top of SGP and the AF routine. Ken + Jared is there any light on the horizon to resolve this issue?
I have just tried switching filters rapidly and with ‘filter offsets’ enabled in the AF settings, while waiting for the start time to come up. I could not make SGP hang. I know I used to have many issues with filter changes on my SX filter wheel, in combination with MDL. I’m using the QSI filter wheel - what are you using.
Well - it is a bit early to declare what is wrong - it may well be that the wheel system is unreliable and SGP is OK. It is very difficult to tell, especially when there is a load of software in-between. The SGP interface to the focuser is using the ASCOM commands. How those are implemented in the ASCOM driver is not under SGP control. SGP issues regardless of the make. Is the Lakeside any different to others, is the Xagly different too?
If you run SGP with only simulators for filter wheel and focuser. Does SGP fall over then?
Not run simulators - something I could try. At the moment have a Win 10 laptop out int he garden running a test sequence to see if it falls over like my main laptop on auto focus…
