symmetal

Clear for the moment and no wind. Older version of PHD2 installed and even changed power supply in desperation but just the same. The 5 mins unguided just shows the 10" RA periodic error having an effect on the star shape. Imaging at 2.74"/pixel that's 3.6 pixels movement which looks about right. The 5 mins guided shows about 8 pixel movement errors in RA and Dec which is 22" drift in both axis. The Dec trace hasn't changed in guided vs guided. What's going on! This is so frustrating! Alan

I've just remembered that PHD2 installed the latest 2.6.7dev1 just before this problem started. It also installed on my other rig and that has had no problems but it might be worth trying the previous 2.6.7 version. Can't see PHD2 deliberately driving the guide star off and not actually showing it on its display or graphs but I'm willing to try anything. 😁 CO is reporting clear up to midnight though it's rather windy at the moment. Alan

I tried all the balance options and the Dec only drifted 3" in 10 mins unguided so the PA was good. The elongated star images on some guided subs showed movement of over 20" while the guiding trace showed 2" p-p movement with occasional spike being 3" Alan

Thanks, I did check the guide settings and they were still at 0.5 as they were before. I also recalibrated and the calibration was fine. There is quite a bit of Dec backlash but it didn't enter the backlash zone while guiding. I also tried different balance settings but it made no difference. Previously I hadn't worried about even about getting good balance as with the relatively large image scale, fairly light setup and guiding it wouldn't be taxed very hard. The images always had small round stars. But not any more. Alan

During the last three days of good weather I can't get a decent image with my ASI071 and ZS61 on an HEQ5 without having oval or sausage stars. Up until then I had no problems. Running out of ideas I tried 5 min unguided just to see what happens and the image was much better. Here's two subs in SGP and the PHD2 trace. Just noticed I put ZS71 by mistake in the label bottom right. For the unguided section I just ran the guiding assistant. This is one of the better guided subs. Often they're much worse. RA is horiz and Dec is vertical and the camera and guide scope are pointing at the same area. PA was within 30" using Sharpcap. The Dec drift unguided is around 1.5" in 5 mins which is not bad. Everything's bolted down tight with no flexure. No wind either. The PHD2 movement indicated particularly in RA unguided is greater than guided. The elongated stars are generally on a diagonal not always the same one. My Atik1/FLT98 on AZ-EQ6 imaging at 1.9"/pixel with guide scope at 6.19"/pixel often has similar shape guiding graphs to the HEQ5 and sometimes worse when at low altitudes but I have no problem with misshapen stars and they're still tighter stars than on the ASI071. Anybody have any ideas on what the problem is? I'm a bit stumped. Alan

Active USB3 extensions do work but they significantly limit the maximum data rate. They were discussed here from a few posts down and the ones used were limiting the data rate to 60MB/s irrespective of the extension cable length. This is 1/10 of the max USB3 speed, although the user noted that didn't cause any issues in use and was happy with it. Different manufacturers may tailor the hardware to the cable length and give higher speeds with shorter cables but none of the specs seem to mention this and misleadingly say they work up to USB3 speeds of 5Gb/s which is 640MB/s. Alan

As long as the target you're imaging doesn't start getting clipped the signal to noise (S/N) ratio of the image will keep increasing the longer the exposure time so in that respect the longer the exposure the better. If you double the exposure the signal accumulated will double but the target noise will only increase by the square root of this, in this case 1.41 times. With normal RGB imaging the sky background level increases the longer you expose, so as the sky background peak in your camera histogram gets further to the right the images lose contrast and dynamic range. Clipped stars also get more bloated the longer the exposure. I think this increasing sky background level is what makes you think the images get more 'noisy' over 5 mins. However with narrowband imaging the bulk of the sky background 'signal' doesn't get through the Ha filter so you can expose for a lot longer before the sky background peak on the camera histogram gets too far to the right. In fact you may have to expose for 30 to 60 mins or more, depending on your light pollution level, before the sky background level is excessive. This is longer than most practical exposures so for narrowband expose for as long as you can if your guiding allows it. Very long exposures have a greater chance of being ruined by aircraft passing through etc. so it's often better to use a series of shorter subs. Star bloating is also less of an issue with narrowband as the bulk of the starlight is blocked. the 12nm astronomik filter is a wider pass bandwidth than the 7nm or 3nm ones used with dedicated astro cameras, so the sky background level will increase faster compared to using the narrower ones but I would have thought 10 to 20 min exposures would be worth trying if you can manage it. I don't do DSLR imaging so someone who does narrowband with DSLRs can give a more experienced answer. For narrowband however it's generally the longer exposure the better within practical limits. Alan Edit: I forgot about the dark current which is a significant source of noise in DSLRs. I did say I don't do DSLR imaging. The noise from the dark current is the square root of the dark current * exposure time, and the total noise from dark current is the same whether you take 1 eight minute exposure or 4 two min exposures. I believe the read noise quickly becomes insignificant compared to the noise from dark current so an advantage of taking long exposures compared to more short ones to reduce read noise is not really important. vlaiv needs to confirm or deny the last paragraph so don't take my word for it yet. 😁

Hi smr, Having perfect polar alignment does not mean your guiding will necessarily be improved. Perfect PA is not actually necessary if you're guiding and depending on the mount performance can even be detrimental. If you are imaging unguided, then getting PA as close as possible is what you want. If guiding, small PA errors will be corrected by guiding, unless the PA is well out when you will end up getting the image rotating around the guide star during the exposure leading to elongated stars in a circular pattern, increasing the further the distance from the guide star. For mounts with Dec backlash as most mid price mounts have, it can be beneficial for the Dec PA to be slightly out by a few arcminutes so that the Dec is always guiding in one direction to compensate and so the Dec gears stay meshed. With perfect PA, the Dec axis will end up sitting in the backlash area and Dec corrections may seem to be ignored until the backlash is taken up, leading to the Dec cycling between fairly large +ve and -ve errors every 10 seconds or more due to the actual corrections being many seconds behind the command corrections. Dithering in Dec can upset this beneficial constant Dec meshing if it dithers to the other side of the backlash area. The mount can then take a while before the Dec settles down again. Getting a good guide star is important. I find at times the guiding is good and then half an hour later is all over the place and changing the guide star helps. The sudden changes in RA error you're seeing on your graph is due to the guide star shape changing between successive guide frames due to seeing, and PHD2 trying to correct for this. Getting the right guide exposure and star shape selection is a bit of a balancing act and will probably need changing throughout the session. I find stars below 100 SNR generally give worse guiding while those much higher are often clipped with flat tops. However, I find that clipped stars can sometimes give better guiding than the normally preferable sharp spike star profiles, as their overall profile remains more constant. It may be worth running PHD2's 'Guiding Assistant' for a few minutes to see if that recommends any changes. It will also measure Dec backlash if you tick the box. Alan

Yes, as the other Alan says it's very easy to use. The guidescope doesn't even have to have the NCP or Polaris in view though it needs to be reasonably close so that its plate solving and adjustment indicating can work. A wideish FOV is also needed on the guidescope again to enable platesolving. Your Starwave should be fine. You need your Sharpcap display screen next to the scope so you can see the results as you make adjustments. It will select a star (not necessarily Polaris) in your FOV and indicate on the display where the star needs to be moved to by altering the mount alt and azi bolts. That's it. Worth the small cost just for this feature alone. Alan

mumble = 4 😁 Alan

I don't recall seeing any actual imaging evidence regarding switch mode vs linear. Most CMOS cameras are powered via the 5V USB from PC or hub anyway, which would have substantial noise on it, so noise on the 12V which in these, mainly power just the cooling doesn't matter. I imagine the cameras have good filtering on their supply inputs to mitigate power noise anyway. The 071 which you have is powered via USB without cooling if there is no 12V. With the 12V connected it will take power from that for the camera as well as the cooling. When I did my own test a couple of years ago using an Atik CCD camera which is only powered by 12V I could detect no difference in noise on the output of the camera when powered from the Mean Well switch mode compared to a 12V leisure battery so from then on I've used switch mode supplies. On my imaging rigs with integral mini PCs, hubs and dew controllers etc., I incorporate LC filters on the 12V distribution PCB I made just to avoid noise generated by the PC etc. getting back on to the main 12V supply. Probably overkill but costs little extra. I also have a Ham radio switch mode supply as mentioned by stash_old but have used Mean Well on my rigs. If you can get the Mean Well supply cheaply, then as you say it would be fine. Famous last words. 😁 Alan

Just to satisfy my curiosity but is the Fusion PS-101 as mentioned several times in the thread actually a linear supply. It seems physically very small for a linear, and reviews on ebay state than it's 'light weight' and 'causes interference on AM and SSB', neither of which would normally be attributes applied to a linear supply. I tried searching on the web but couldn't find any reference to what type of supply it is. Its low price implies switch mode but I'm happy to be proved wrong. 😁 Cheaper switch modes of a more basic design will generally cause more interference than well designed, more expensive ones. In msacco's defence, he did state a 10-12A power supply in his first post. Alan

That's a pity. Running out of ideas I'm afraid. Alan

I use switch mode supplies and have no issues with any more noise being seen on camera outputs compared to using 12V batteries. It's a long standing debate about switch mode vs linear but it's up to you. Many people here on SGL do use switch mode supplies for their imaging rigs. I buy known reputable brand power supplies like this one which has the specification you want. The data sheet for this one does say the adjustment range is from 10.2V - 13.8V which would be fine for you. The load regulation is +/- 0.5% from 0 to 100% rated current so the output voltage will only vary by +/- 60mV from 0 to 12.5A load so no problems there. They are still made in China, so the one you listed is most likely a similar specification but no guarantee of course. Alan

Actually, in the specification it states it uses a DC servo motor which implies feedback via optical encoders or similar, in order to give more precise control of angular position, which is what you need. Nothing in the description mentions this, which I thought would be a good selling point, so am not sure what actual system it uses. The reviews state that people are happy with it and that it works well for automated focusing, so it may be worth a try. Hopefully someone here uses one and can give a more informed answer. Alan

I assume the EQMOD cable you plugged in was connected to your PC USB port at the other end. If it wasn't I couldn't see it causing any harm. The USB Eqmod cable only uses pins 4, 5, and 6 on the RJ45, with pin 4 being Gnd (may also be pin 1 instead of, or as well) and pins 5 being RXD and pin 6 TXD, both 5V TTL levels. Assuming I've got my mental gyrations correct on orientating the RJ45 on your board Pin 4 (Gnd) is already connected to pin 1 on your PCB which is fine, and then appears to connect to L4 Pin 5 (RXD) being an input (high impedence) shouldn't upset your board though it is connected to L5 Pin 6 ( TXD) being an output (TXD) doesn't appear to be connected to anything on your board. It's worth checking continuity on L4 in particular, and the other inductors going to the RJ45 section of your PCB for good measure, to confirm they still measure short circuit. As malc-c suggests it's quick to check the continuity of all the L components on the board. The 'crack' noise you heard could easily be the ferrite core of one of these inductors cracking as the thin wire in the inductor got hot and melted due to excessive current. L4 being shorted to ground by your Eqmod cable is a prime candidate.

This uses a DC motor rather than a stepper motor so accurate positioning would be more difficult. It has an Ascom driver but I don't know if SGP will work with it or if SGP just works with stepper based focusers. Hopefully someone can say how well they perform. Alan

The 'grinding' noise may be the mirror weight and mechanism pushing back against the autofocuser once the stepper has stopped moving. The residual torque of the stepper once de-energized may not be enough to hold it in place. As the lakeside motors are heavily geared down this should not normally be a problem but I don't know how much force the mirror mechanism is exerting. Just a thought. 🙂 Spending more money is the best cure. 😁 Alan

Thanks for the pictures. It makes more sense now. 😀 As Dave and Stuart have said it looks like trying to move the mirror directly with the autofocuser is more trouble than its worth, and using a separate focus tube on the rear of the scope with the autofocuser is the way to go. You can check whether you have sufficient backlash compensation by checking the focuser knob moves slightly 'inward' after having completed a focus out command, (assuming you have selected the 'IN' direction in the SGP focuser setup). This means the compensation has taken out all the backlash and your first autofocus V curve segment should then not be flatter than the rest. Setting the backlash compensation too high is not a problem. It just means the focuser moves a greater total distance when focusing out so takes more time. Alan

The first 4 points being horizontal means the backlash compensation value used is not enough or things are not tighted up in the focuser. The stepper is rotating but the scope focus tube ( or mirror if connected directly to Edge HD focus control) is not yet moving until the gear backlash is taken up. This also explains why your final focus position is incorrect as the final move is in the opposite direction to get back to the centre position and so backlash is being taken up rather than the focuser tube actually moving. The excessive backlash is possibly the grub screws on the lakeside shaft adapter not being aligned on the flats of the shaft. Not sure how your two brackets are fitted. A picture may help. The lakeside probably uses a 16 bit value to store the focuser position, hence the limits being 0 - 65535 steps. Your setup having 112,000 steps means you cant have full travel with the lakeside but that shouldn't be a problem if you set it in the range you normally use it. Alan

Have you installed the AscomPad software? The Eqmod page doesn't mention Ascompad is needed for gamepad use with Eqmod. I last tried it a few years ago and I'm pretty sure Ascompad was needed then, but whether it still is I'm not sure. Alan

You need to keep the shed well ventilated to avoid condensation forming during temperature changes. If the interior was fully sealed you would likelyneed a dehumidifier running to stop condensation forming. Being well ventilated the air inside will be the same humidity as that outside so problems with damp or mould would need to be addressed. For my small roll-off shed which is well ventilated as the doors have a large gap down the hinge sides, I just have a 60W tube heater on the floor by the scope. Before using the heater the wooden shelves would show signs of mould after a while but since using the heater they are fine. It's thermostatically controlled so only comes on when the temperature drops so doesn't use too much power. It just keeps the interior a few degrees or so above the outside temperature so lower humidity. Alan

Yes, EQ mounts just track in RA. If accurately polar aligned there would be no Dec movement when imaging DSOs so there is no need for Dec tracking. Solar system objects will drift in Dec but would be very slow for the planets but faster for the Moon as seen. Selecting Lunar or Solar tracking on the EQ mount just changes the speed of the RA tracking to match the speed of the Moon or Sun's momement in RA. Alan

Though it seems they don't keep good focus. 😁 Alan

Very nice Dave. Haven't had my lunt out for months. Must get it out again now the weather is more friendly. 😁 Alan