tonyowens_uk

Perhaps someone has a source of industrial quality USB cables they can pass along? Specifically I'm looking for USB 3 Type A to Type B plug cabling designed for industrial environments. This is the opposite of consumer and office-grade cabling. So I need either silicone or polyurethane (not PVC) sheaths, either flat or circular section, gold plated contacts, 3-5m length, choice of straight or right-angled plugs. This is for permanent outdoor use hooking up imaging equipment in Wicklow's cold wet windy climate. Many thanks Tony Owens

Graeme if you are still interested contact me off list to discuss or speak with Grant or Rob at FLO about the EQ8 modernisation service. It’s not a kit BTW. The service is collect and return as the modifications require a tool room or R&D lab to perform properly. Tony Owens

Thanks Dave, thats a good thought. Tony

Does anyone know whether Philip Keller the optical design engineer and supplier of professional Wynne correctors is still in business? If anyone has contact information for Hr Keller or his business I'd appreciate a referral, thanks. Tony Owens

No. Dignify the discussion by reading my essay. All that to try make a case for: more thoughtfulness about terms like quality in highly engineered systems the avoidance of loose language and the lazy equation of cost with quality and to discourage the casual dissing of the work products of other people (e.g. Onstep) without evidence! Best Tony Owens

Hi newbie! I dont think its meaningful to speak of the quality of a motor or any other technical component in the abstract. A 200 step/rev hybrid stepper motor cant be compared to a brushed DC servomotor without looking at the context. If that is running an axis of a telescope mount, we are looking at torque vs speed characteristics, electrical efficiency, noise emission, 'cogging' behaviour, spindle positioning elasticity (drive stiffness), robustness, wear life, ease of packaging, ease of commissioning and reliability. And in this application no meaningful comparison is possible without also including the motor drive electronics, cabling, connectors and firmware. On one side we have Maxon/Scitech and industrial sealed connectors and screened cabling. On the other, we might have a Chinese generic 400 step/rev NEMA17 hybrid stepper driven by a Trinamics TMC2209 driver and commanded by Onstep running on a modern ESP32 embedded controller, with unshielded and unsealed motor wiring. Or it might be a similar controller board operating a Clearpath SDSK stepper-servo with 6400 tic embedded encoder via Onstep's step/direction interface, and wired with silicone-sheathed shielded cable and metal IP65-rated M12 multipole connectors. In terms of measured following error during tracking (motion quality), the difference between these three solutions may well be negligible. Autoguiding compensates perfectly for the lack of a high-resolution axis encoder on all of these solutions. The torque capacity superiority of the stepper solutions is irrelevant as input torque capacity of Mesu's friction drive will limit how much motor torque can be delivered. The speed advantages of the servo are irrelevant as the steppers can 'pull' higher overall transmission gearing if necessary. As the friction transmission is (as I understand) backdriveable, the efficiency of the stepper-based drivetrain may be more efficient if the mount spends significant periods held stationary. Or not, as the case may be. There will be no difference in noise emission when tracking between these mounts except for the servo's faint 20 kHz commutation whine. None of these systems will exhibit cogging behavior visible in images (provided the servo is properly tuned with a small 'deadband' of about +/- 2 encoder counts and provided the Trinamics driver running the simple stepper has plenty of drive current available). One big difference between the simple stepper system and the other two (servo and servo-stepper) might be drive (motor rotor) positioning stiffness. Whether far superior rotor stiffness of the latter solutions provides any benefit will depend on the amount of friction in the Mesu's transmission. As this is far lower than worm drive mounts, for example, it may well be that there is no perceptible difference. But in a worm drive mount, especially of the antibacklash type where drag friction is significant, a difference is likely to be apparent in rms guiding error. Another difference emerges when looking at ease of commissioning. The Scitech servo system needs careful tuning to obtain stability without micro-oscillation at tracking speed or sluggish response to positioning moves, which makes building the systems more involved. None of these systems (so far as i'm aware?) employ speed reduction gearheads so all should be equally reliable. The only one that is environmentally-sealed to normal industrial standards for outdoor use (advisable for remote-controlled observatories for example) would be the servo-stepper one I mentioned with waterproof connectors. So there you have it. There is a very big difference between the cost of these three systems, the Scitech one being most costly by far. This is easily explained by: the manufacturing economics of the relevant component suppliers the avoidance of the need to recoup very substantial R&D costs on the Onstep system which is a GPL development that is designed to run on COTS hardware the fact that there are multiple parties involved in the supply chain for Scitech which leads to greater margin-taking For many end-users, performance differences between these systems will not be apparent, meaning that there is no perceptible diffrence in 'quality'. For a minority of users who need the greater simplicity of repairs and maintenance of the stepper system e.g. remote controlled imagers, that might be a better choice for them. For another minority who are interested in remote satellite tracking and orbital parameter calculations and need high dynamics, a servo-stepper system might be their choice. And perhaps a few (like me) admire the form and finish and heritage of historical highly-engineered assets - for us, the existing Maxon/Scitech solution is also appealing. Tony Owens

Quality of a motor or a motion controller has little to do with its cost. Stepper motor and servo motor drive systems have their own strengths and weaknesses, and in some applications the motor technologies can be blended (servo-stepper). What matters is the motion quality delivered, and the reliability and maintenance requirements. There is no reason I can think of why a microstepped Mesu friction mount should not track as well as the Maxon skewed-rotor/Scitech model, particularly if closed-loop steppers (servo-stepper technology with embedded encoder) are chosen. You seem to suggest that the Onstep control software is inferior to Scitech's. Do you have evidence for this?

The owner of the Skywatcher brand, and Celestron Corp and various other entities, and of the patent portfolio underpinning Skywatcher gear and manufacturing processes is Nantong Schmidt. David Chen is still executive chairman. I toured the manufacturing and engineering facilities in Dec 2018 and they looked healthy and financially successful to me. Tony Owens

Hi Mike. I agree with you about combining PPEC with properly-refined autoguiding - I've never seen a clear benefit myself in the case of EQ8's. What works for me extemely well (using the original motor board/motors used in the original EQ8 and antibacklasg gears) is very frequent small corrections, with a predictive guiding algo from PHD2. With the much improved motor board that appeared with the mount refresh, the picture is likely to be different because of the more refined microstepping behaviour, and I havent investigated that yet. Re references to the 'under the hood' EQ8-specific investigations and improvements built into GSS, I suggest you search for contributions from Andrew Johansen in particular. cheers Tony Owens

Hi Mike, Three examples of differences in favour of GSS v. EQMOD are: 1. If you use the integral PPEC on the EQ8 or EQ8R, I believe that incoming autoguider move commands that coincide with a PPEC 'bin' transition are ignored, which negatively affects guiding accuracy. This happens with EQMOD controlling EQ8, due to a design flaw in the EQ8's firmware design. My understanding is that this also happens with the newer EQ8 firmware versions (cross-compiled for the STM32 that runs the EQ8 motion controller). In GSS there is a workaround. 2. Very short autoguide corrections sent to either EQ8 or EQ8R are executed very rapidly in GSS using GoTo moves. EQMOD has nothing comparable. 3. EQMOD is not in active development. GSS has at least three intermittently active developers and a growing user base You can read about the first two of these issues on the GSS forum I hotlinked in my earlier post, if interested. Best Tony Owens

Nigella if you have not already looked at GSS as an alternative to EQMOD for controlling your new EQ8R, it would be worthwhile. There is a discussion list for GSS here: https://gss.groups.io/g/main GSS is in active development (it was originally developed to work around some of the EQ8's limitations, which it does rather well) with meaningful new functionality still appearing. You will no doubt have a large list of issues to work through with your 12" imaging Newt on your new EQ8R to get it operating reliably at the expected performance level, but when you get that under control, best have a look at GSS. Tony

Most pro-sumer mounts use stepper motors driven in microstepping mode. In the case of the older Skywatcher models, e.g. EQ8, there is a definite motor torque benefit to using the highest drive voltage possible. For safety, a regulated power supply set to 15Vdc maximum with at least 5A capacity and preferably 10A is my recommendation. Use sensibly short cabling from the supply to the mount, with 16AWG stranded wire, or more. That will hold cable loss to 0.3V or so. The reason for maximising torque is that in tracking mode when microstepping is active, significant friction arises in properly set up worm gear drives. This causes a 'following error' to arise - a lag beween the commanded stepper motor shaft position and the actual position, to the detriment of tracking accuracy. Fortunately the effect is small, but it is there. Boosting torque reduces following error. Tony Owens

I need to qualify my comment Ken. The comparison would be against quality antibacklash worm gears, not fixed-centre gears. Remember that many traditional mounts have neither ground worm gears, nor precision-grade worm bearings, nor antibacklash provisions. And in some cases not even proper environmental sealing of the oily bits. There is a huge quality spread across the hundreds of models of worm geared mounts available. Tony

From what I've seen, Skywatcher (Nantong Schmidt Instrument Company) have enormous manufacturing operations of their own, for optics, tubes, machined parts, worm wheels, finishes, assembly, test and many other things. Obviously they dont make their own PCB's, castings, small lenses and so forth. No OEM does. Such things come from specialist suppliers. But they do have an excellent design engineering team, who are very much involved in the latest developments, and plenty of resources. So far as harmonic gears are concerned, as used in the compact mounts referred to on this thread, these all come from specialist companies who serve the automation/robotics/semiconductor processing equipment markets. The technology has been around for decades, but it is difficult to master, as it involves materials, heat treatment, motion control, creep-grinding and structural analysis expertise of a high level all under one roof. At the hobby level,harmonic gears are 3D printable, but for ultraprecision uses such as imaging telescope mounts, neither Nantong Schmidt nor anyone else is going to be producing their own! It is worth mentioning, that for all their advantages (compactness, low backlash, low friction losses), harmonic drives have many disadvantages too (indifferent linearity (large periodic error), lack of torsional stiffness, cost). When autoguided by fast guide optics and star centroiding software with with good response, they should track fairly well. But they will not match the accuracy and low PE of a good set of machined and ground worm gears, never mind friction drives.

Hi Mike. Loved your account of the 12" Cass/RC build on your site BTW! The EQ8 is awkward as its neither fish nor fowl. The underlying mechanical precision and structural stiffness is superior enough to encourage seeing-limited imaging, but the omission of anti-backlash worm gear meshing and extensive use of thermally-reactive aluminium components (shafts, structure, worm wheels) frustrates being able to set a stable mesh clearance as ambient temperature changes. On top of that, many have run-out issues of the Dec worm wheel relative to the Dec worm, caused by tolerance stack effects. Consequently, it isnt possible to establish stable meshing friction around the Dec axis, never mind stable backlash. If run-out doesnt get you, temperature changes will! The best that can be done without surgery is to clamp the Dec clutch as lightly as possible, and set the worm meshing clearance at the 'tightest' position angle around the axis, judged empirically. The surgical fix is to replace the standard Dec gearing with an antibacklash one, which is very effective, and replace the standard clutch with a modified design. But this isnt particularly cheap.

Andy that isnt a great approach. The badly designed clutches on EQ8's do affect backlash on Dec (as well as cause image shift) on opening and closing them. But the effect is not predictable or controllable or desirable. Overlaid on this, is a seasonal variation in backlash due to differential thermal expansion of the mount components and varying ambient temperature. The mount is tight or binding in winter, whiel backlash grows in summer. If you are an imager, my advice would be not to waste your time messing around with a half-assed response, and either make the adjustment yourself or find someone qualified to do so.

There is a special interest group on Google for EQ8 mounts here: https://groups.io/g/SkywatcherEQ8/topics Backlash adjustment instruction and help are available. All worm gear drive mounts that dont have antibacklash gears have significant backlash (for an imager). Either adjust it yourself or get your dealer to do it. Its perfectly normal.

If you have to tote a tripod for non-permanent use, and need this to be capable, I'd highly recommend a heavy-duty well-designed wooden tripod. Berlebach is a German firm that offers models stiff and well-damped enough to work well, yet the weight for moving it around (in the deployed state) is manageable. The engineering is painstaking. I use one of their Planet models with an EQ8 occasionally to carry up to 30Kg imaging payloads and the stability is most impressive. Tony

Yes of course. The issue have less to do with the conical back than to do with the hassle factor of altering established polishing and testing setups to suit a large optic with a conical rear. You need to look out for a small highly capable firm used to jobbing precision optics and capable of making one that large. Know what profile of conical rear you need and have a proper enginering drawing ready before you make an approach, or you may look like a time-waster. If you dont have the FEA capability to design a hub-mounted blank with the right profile to give you deformations low enough to meet your use requirements, you should find someone with that experience too. The optician is unlikely to generate your required conical profile in-house. Diamond milling of glasses and optical ceramics is generally contracted to the optical glass stockist. I had a 14" F4 parabolic mirror to my own design produced by Oldham Optics UK years ago, later refigured by Es Reid. I believe others in the EU such as Toscan Optics in Italy can also supply the semi-pro as well as the professional market for custom optics, as can Norman Fullam in Quebec. Good luck!

The EQ8-R is functionally identical to the older EQ8 Pro. The improvements (through-the-mount cabling, improved microstepping quality, better static balance, better clutch design) are useful but not groundbreaking. The more fundamental upgrade, antibacklash gearing, was not implemented. The -Rh variant with an RA encoder apparently does a good job of reducing RA worm periodic error. On the other hand, autoguiding, especially with an off-axis guider (where possible) on the original EQ8-Pro can match the -Rh's guiding quality, provided the OEM clutches are modified, antibacklash gear modifications are installed, and Green Swamp Server is used to control the mount. This latter replaces EQMOD and brings better motor microsteppign performance and sutoguiding correction quality. High resolution axis encoders are not an answer to any problem in mounts of this class, by themselves. Sophisticated pointing model capability is needed in addition. That is not offered currently either by Skywatcher or EQMOD or GSS. And even when it appears, for most purposes it will offer no practical advantage over refined autoguiding and platesolving. Tony Owens

As a prominent Western prophet said "the truth will set you free". Objective eyepiece testing is far from straightforward and in any case is not the only criterion for value of an ocular. I had the chance to return the XW30 I had bought, following my own comparison (using my own preferred equipment). But though I intended to, I have not done so. 20 years ago I remember being blown away by the overall experience of using an XW10 compared with various commercially-available and privately developed high quality oculars - the ease of use, the lack of scatter, the ergonomics, even the smell. That memory was instrumental in my decision to buy a 6 element all-spherical design in 2020. The essential difference between the XW30 and the APM 30UFF (apart from superior tech performance and lower cost of the latter) is 'character'. We humans aren't the cold rational creatures assumed by most economic theory. As I've got older I've come to trust my non-rational judgements more and more...

Here is a summary of the testing I carried out on these two oculars a month and a half ago after I received the APM 30mm UFF which was briefly out of stock at APM: Test of Pentax XW30 v APM UFF30 Date: 25/05/20 Setup: Intes-Micro MN86 Deluxe, collimated, F5.9, flatfield EQ8, tracking Visual use with and without glasses My vision is within 1 diopter of normal, negligible astigmatism, reduced accommodation Seeing: 7/10, traces of high cloud occasionally, 12 C, not truly dark. Various sky targets Targets M27 Dumbbell nebula NGC6960 Veil Nebula area C1399 asterism (altitude 44 deg) M8 Lagoon (v low) Saturn (v low) Altair (for veiling glare and polish assessment) APM 30 UFF NOTES 545 gms No chromatic aberration anywhere Sharp to within a couple of degrees of the edge Eyecup not as effective as the Pentax at excluding ambient light, but can be folded over quickly to accommodate wearers of glasses. Little veiling glare apparent on Saturn or Altair Folded down eyecup a bit more comfortable and suits non-varifocal glasses wearers Pentax 30 XW NOTES 680 gms Seemed slightly brighter - a little more spectacle Head positioning to the exit pupil a little easier Sharpness starts to suffer noticeably from ~60% of the field on, not bad until close to the edge, but noticeable deterioration Obvious chromatic aberration off-axis on brighter stars A bit harder to locate the eyeball to the exit pupil perfectly, due to size of barrel end Little veiling glare apparent on Altair centre-field A little warmer in colour transmission than the UFF Conclusions Overall, the UFF performs better due to tighter stars across the entire field and no chromatic aberration. The Pentax is inferior, most likely due to an ageing optical design which does not suit modern astronomical telescopes with image surfaces of low curvature. I was also able to confirm broadly identical behaviours using a Skywatcher 80mm F5 APO without its focal corrector/flattener fitted, in some quick testing the following night. In view of the poor off-axis performance of the Pentax I didn't bother rigging the APO with the field flattener and retesting. As the direction of image plane curvature is opposite between APO refractors and mirror scopes like the Mak-Newt, it was clear that the Pentax was not going to match the APM ocular regardless of the image curvature sign or magnitude.

I have an APM 30mm UFF on backorder with APM. I plan to assess the Pentax XW30 v the APM UFF 30 when I get my hands on one. I'll do this with and without the field flattener fitted to the 80mm APO.

Hi. I bought an 80 mm Skywatcher triplet scope a couple of years ago with portable imaging and occasional visual use in mind. Until recently it hasnt had much use. Debugging the imaging side of things is being attended to currently, but I had hoped to come up with a choice of quality eyepieces that allowed widefield sweeping of the Milky Way at one extreme, down to around 160x or so for lunar views while travelling. I have the Skywatcher fleld flattener by the way. So recently I replaced a Televue Panoptic 27 with a Pentax XW30, as I hadnt been happy with the edge of field performance of the Pan with or without the FF fitted. I find the XW30 is not a lot better. I'm interested in recommendations for oculars that can do justice to the razor sharp imagery that this triplet delivers (evident in the star shapes while imaging). I have no astigmatism or other eyesight defects other than the usual lack of accommodation that is appropriate to a man in his late 50's! I prefer to avoid anything wider than 80 degrees or so (Nagler territory), and I's like to use the scope by day as well as night, so avoidance of 'kidney bean' is also an issue. Mainly, I'm after razor sharp images over 95% of the field and great comfort, with this fast, short FL scope. Thanks Tony Owens

I'm sure Bog Snorkelling is an important formative experience for Irish undergraduates: As for Green Swamp server (GSS) its been around for a while. Unlike EQMOD its in active development. Initially the target mount was just EQ8/HDI 110 but it now includes other Nantong Schmidt EQ mounts. There are several controls which extract better performance from EQ8's when autoguiding than EQMOD does. One, is homing that is far more reliable. But the key ones for imagers are motor current doubling during tracking and autoguiding, (which doubles stepper motor torque in small corrective moves to reduce following error) and use of short range GoTo commands while autoguiding using pulse guiding. These improve corrective move accuracy, reducing control loop dither around the target position. It is necessary to download a bugfix version for the EQ8 firmware to access the full feature set. There is a discussion groups here: groups.io/GSS/ Tony