New mount suggestions

[Rainer]

28 minutes ago, vlaiv said:

I would say that it's quite questionable - since CEM120EC2 is in Mesu 200 territory price wise. I'm not entirely sure I would go for CEM120EC2 over Mesu 200

 

Correct, everybody can spend their money as they want just one point which makes me a bit nervous about MESU. It is a one man show ... and everything is located in Netherlands ...

For the people living in Europe and everything is 8 hours away by car ? it is s good alternative but worldwide ? That is the question ...

OK, I am in Mexico and so even with iOptron it is not that near but I have a good vendor in Houston, Texas who is 16 hours away by car ?

One more thing, and this is my very personal opinion and nothing more then my very personal opinion, That friction drive inflicts me quite some respect ... ?

regards Rainer

IMHO ?

MESU mount PE image taken from his website

 

The PE of my 2 CEM 120EC2 mounts

 

[vlaiv]

2 minutes ago, Rainer said:

The PE of my 2 CEM 120EC2 mounts

That is one major deciding factor for me. Although cem120EC2 has very low RMS error - it just looks too jittery compared to mesu. I like the smoothness - that just means excellent guiding capability. Add in good stiffness, no backlash and thru the mount cable management in new version - what's not to like

I do however get your point about road connectivity and gasoline consumption

[vlaiv]

7 minutes ago, Rainer said:

The PE of my 2 CEM 120EC2 mounts

Btw, what is the time scale on those graphs for CEM120?

[RolandKol]

24 minutes ago, vlaiv said:

Btw, what is the time scale on those graphs for CEM120?

1sec :))))))))))))))))))))))))))))))))))))

joking

[cotak]

4 hours ago, vlaiv said:

Btw, what is the time scale on those graphs for CEM120?

One worm cycle typically.

[cotak]

4 hours ago, vlaiv said:

That is one major deciding factor for me. Although cem120EC2 has very low RMS error - it just looks too jittery compared to mesu. I like the smoothness - that just means excellent guiding capability. Add in good stiffness, no backlash and thru the mount cable management in new version - what's not to like

I do however get your point about road connectivity and gasoline consumption

Different graphs for different reasons.

iOptron does regular PE graphs too for non-EC mounts.

The point of the EC graph is to show how tightly the encoder helps keep overall excursions. It's a active control so it's going to look quite dynamic. Not that different from a guide graph at fast intervals.

[Rainer]

5 hours ago, vlaiv said:

That is one major deciding factor for me. Although cem120EC2 has very low RMS error - it just looks too jittery compared to mesu. I like the smoothness - that just means excellent guiding capability. Add in good stiffness, no backlash and thru the mount cable management in new version - what's not to like

I do however get your point about road connectivity and gasoline consumption

 

5 hours ago, vlaiv said:

Btw, what is the time scale on those graphs for CEM120?

240 seconds = 360 teeth on the worm gear

[vlaiv]

12 hours ago, cotak said:

Different graphs for different reasons.

iOptron does regular PE graphs too for non-EC mounts.

The point of the EC graph is to show how tightly the encoder helps keep overall excursions. It's a active control so it's going to look quite dynamic. Not that different from a guide graph at fast intervals.

 

11 hours ago, Rainer said:

 

240 seconds = 360 teeth on the worm gear

Ok, that is a bit worrying to be honest. I counted roughly 90 local peaks in 240 seconds, if whole graph is indeed 240s. Let's say on average p2p of those excursions is about 0.75" - looking at the second mount (first one is a bit better) - labeled GB20002.

240s / 90s = 2.6666s. 0.75" p2p / 2.6666s = 0.28125"/s

That is seriously rough ride in my book. Let's do comparison with mesu, just to put things into perspective. Let's read from above posted graph - to my eye one of the "roughest" places is at about 1020s mark (peak between 990 and 1035). Again eyeballing it - in about 30-40s (two / two and a bit tick marks), there is a change of about 0.5" (roughly 1/4 of 2" interval). Let's take lower time to get higher change speed value (it will put the mount at "disadvantage" - but I would rather think it's performance is poorer for this comparison than better).

0.5" / 30s = 0.016666"/s

That is about x20 less "jittery" / more smooth than CEM120EC2.

It also shows one important aspect - you can use for example 10s guide exposures and still have only 10s x 0.016666"/s = ~0.167" of movement from base position. In 10s you will certainly smooth out any seeing effect.

That is far better behavior than "dynamic active control" of CEM120EC2 that jumps around the place in those 10s making about one whole 0.7" p2p trip and then some ...

My above analysis might be flawed, so I urge everyone to check if I made a mistake somewhere, but as it stands, two mounts in same price bracket and to me there is no comparison in smoothness.

Wonder what PE of CEM120 / CEM60 without encoders looks like measured with absolute encoder with enough resolution on the shaft (same technique as mesu).

[cotak]

35 minutes ago, vlaiv said:

 

Ok, that is a bit worrying to be honest. I counted roughly 90 local peaks in 240 seconds, if whole graph is indeed 240s. Let's say on average p2p of those excursions is about 0.75" - looking at the second mount (first one is a bit better) - labeled GB20002.

240s / 90s = 2.6666s. 0.75" p2p / 2.6666s = 0.28125"/s

That is seriously rough ride in my book. Let's do comparison with mesu, just to put things into perspective. Let's read from above posted graph - to my eye one of the "roughest" places is at about 1020s mark (peak between 990 and 1035). Again eyeballing it - in about 30-40s (two / two and a bit tick marks), there is a change of about 0.5" (roughly 1/4 of 2" interval). Let's take lower time to get higher change speed value (it will put the mount at "disadvantage" - but I would rather think it's performance is poorer for this comparison than better).

0.5" / 30s = 0.016666"/s

That is about x20 less "jittery" / more smooth than CEM120EC2.

It also shows one important aspect - you can use for example 10s guide exposures and still have only 10s x 0.016666"/s = ~0.167" of movement from base position. In 10s you will certainly smooth out any seeing effect.

That is far better behavior than "dynamic active control" of CEM120EC2 that jumps around the place in those 10s making about one whole 0.7" p2p trip and then some ...

My above analysis might be flawed, so I urge everyone to check if I made a mistake somewhere, but as it stands, two mounts in same price bracket and to me there is no comparison in smoothness.

Wonder what PE of CEM120 / CEM60 without encoders looks like measured with absolute encoder with enough resolution on the shaft (same technique as mesu).

I suggest you read over this:

Yes the iOptron plot looks rough but the Mesu site graph is likely not raw data.

 

[vlaiv]

12 minutes ago, cotak said:

I suggest you read over this:

Yes the iOptron plot looks rough but the Mesu site graph is likely not raw data.

 

Even so, if you look at graph with replaced encoder, max error rate is 0.05"/s - which is x4 larger than my calculation above, but still x5 smaller than CEM120EC2

Don't know if all mesu mounts suffer from this and will it be corrected in MK2 - I've seen mention of new more precise encoders on motors.

[andrew s]

Gven the mass/inertia of the mount and scope, how could they respond that fast? What is being measured here? I suspect noise e.g. star scintillation.

Regards Andrew 

[vlaiv]

19 minutes ago, andrew s said:

Gven the mass/inertia of the mount and scope, how could they respond that fast? What is being measured here? I suspect noise e.g. star scintillation.

Regards Andrew 

I think that CEM120EC2 graphs are absolute encoder corrections - very unlikely related to actual seeing / star profile.

[andrew s]

9 minutes ago, vlaiv said:

I think that CEM120EC2 graphs are absolute encoder corrections - very unlikely related to actual seeing / star profile.

So most possibly digital noise in the control loop and unrelated to real motion of the axis?

Regards Andrew 

[cotak]

2 hours ago, vlaiv said:

Even so, if you look at graph with replaced encoder, max error rate is 0.05"/s - which is x4 larger than my calculation above, but still x5 smaller than CEM120EC2

Don't know if all mesu mounts suffer from this and will it be corrected in MK2 - I've seen mention of new more precise encoders on motors.

Only true after low pass filtering. The iOptron graphs as far as we know have no filtering.

Also look at the graph from 10u.

I think you are looking at a tree and missing the forest. 

Edited April 26, 2019 by cotak

[Rainer]

Hi,

I am just ROFLMAO many times ...

Again a simple graph is now taken to analize something and immediately it is assumed that that is huge problem ?

?

@vlaiv MY best advice to you is DO NOT buy an iOptron CEM 120 nor EC nor EC2, You will NOT be happy at all ... Spare yourself in future bad moments ?

... and BTW the answers are drifting in the same way as a bad Polar Alignment

Edited April 26, 2019 by Rainer

[Rainer]

3 hours ago, andrew s said:

Gven the mass/inertia of the mount and scope, how could they respond that fast? What is being measured here? I suspect noise e.g. star scintillation.

Regards Andrew 

Hi Andrew,

This graphs are made in the factory using a high resolution encoder on the end of the shafts and then they let the mount run for a cycle or more but only print one ccyle and in this case 240 seconds and not as assumed by vlaiv who wrote in doubt " ...  if whole graph is indeed 240s. " 

@vlaiv It is 240 seconds, believe me ? ?

Now another thing, why nobody asks how can DEC have an RMS if it is static while pointing on and object, well iOptron does measure that too for PE comparisons of the worms installed in the DEC axis ...

WOW, one never stops to wonder day by day ...

OK, souls33k3r from my side I will not be mentioning the CEM 120EC2 again as it is distorting the whole idea of your original question. Apologies ... ?

 

Edited April 26, 2019 by Rainer

[vlaiv]

4 hours ago, cotak said:

Only true after low pass filtering. The iOptron graphs as far as we know have no filtering.

Also look at the graph from 10u.

I think you are looking at a tree and missing the forest. 

Might be that I'm missing the forest by looking at a tree. Just trying to asses performance of the mount I don't have and go by what is available to me.

This is why I find yours and input by @Rainer so valuable.

On the matter of 10u graph - note time scale on that graph, it is at least x6 "denser" than CEM120EC2 (and I add apology here since we are still talking about CEM120, although main mount of interest is CEM60, but I guess similar can apply to that mount as well?). Note also that p2p in 10u graph is 0.4" on average (between 0.2 and -0.2 initially and stays roughly the same). Just for fun, I'll compare that to CEM120 above.

There we had ~ 0.75 / 2.6666 = 0.28125"/s

For 10u we have something like ~0.4 / 16 = 0.025/s

Andrew indeed pointed out to very interesting thing. Is external absolute encoder precise enough to measure these sort of graphs? Just for fun, let's check what would be resolution of very good absolute encoder in arc seconds.

Best that I could find online are 22bit absolute encoders, meaning they have 4194304 steps. There are 360 degrees in a turn, 60 minutes per degree, 60 seconds per minute totaling 1296000 arc seconds in revolution.

This means that such encoder will resolve ~ 0.309". Two interesting facts - high resolution absolute encoders don't even have precision to produce above graphs.

iOptron on their website under specification says they use encoders with 0.035" resolution. That would require 25 bit absolute encoder.

10u on the other hand says something like: "Ultra-high-precision RA & Dec on axis encoders, featuring more than 10 million increments (interpolated), fully encapsulated and calibrated". This is closer to 22bit encoders being interpolated between single ticks (time interpolation?).

All of that is quite confusing, I must say ...

 

[souls33k3r]

Can we kindly keep the discussion within my budget please

[Jkulin]

Problem is that no one knows how the new Mesu will perform, and like all new things will take time to get rid of all the bugs?

[vlaiv]

23 minutes ago, souls33k3r said:

Can we kindly keep the discussion within my budget please

Yes, best do that.

In the light of current findings, this is my opinion at the moment (subject to change, caveats included - like have not seen or operated one).

I think that CEM60 is decent upgrade in comparison to for example heq5/eq6 family of mounts. There are bunch of good features like belt driven, magnetic tension on worm / no backlash, cable management, lightweight, payload capacity, stepper precision to justify 50% price increase over said competition (EQ6-R for example).

Only reservation that I have is the issue with PPEC, but if manufacturer specs are to be trusted, PE is 5" or less p2p, so there is a question if PEC is worth using at all.

With regards to EC version - after above analysis, I don't think it's worth the money. Specs on it for example for discussed big brother - CEM120 claim 0.035" encoder accuracy - that simply can't be true (no single rev encoder with such precision), and if it were, I would expect much much smoother graph of tracking - which calls into question other claims by manufacturer.

Bottom line, if heq5 can be guided down to 0.4-0.5" RMS in right conditions, then I would expect CEM60 standard version to be able to be guided down to 0.3" RMS in right conditions, and certainly 0.5" RMS most of the time. That is as good as can be expected in that price range.

[Freddie]

This has been a very interesting thread. The one thing I do find odd though is that the claimed max imaging load for the eq6R is greater than that claimed for the CEM60.

[Whirlwind]

On ‎26‎/‎04‎/‎2019 at 23:17, vlaiv said:

In the light of current findings, this is my opinion at the moment (subject to change, caveats included - like have not seen or operated one).

Unfortunately there are a lot of unknowns in both plots to really understand what the differences are.  To have a completely scientific comparison they would need to be measured in the same way to ensure that the data is comparable.  Statistically assuming a perfect test you should get a distribution of errors over time.  If the CEM data is almost an instantaneous sampling of the data and the MESU was averaged (say 5/10 second exposures) then from a simple statistical analysis you would expect more variance in the CEM data -really it should be a median or average over the same timeframe.  I hence don't think we can make much of the comparison of the data without some real errors being shown on the data points.

[vlaiv]

1 hour ago, Freddie said:

This has been a very interesting thread. The one thing I do find odd though is that the claimed max imaging load for the eq6R is greater than that claimed for the CEM60.

I've seen that EQ6-R has 44lbs quoted AP payload capacity (according to SW manual for EQ6-R), while I have not seen actual figure quoted for CEM60 for AP, but it says it has 60lbs carry capacity.

If we go by 2/3 - 3/4 for AP, that would place CEM60 in the same capacity category? 40-45lbs?

[Freddie]

We are in the same ball park, I took imaging as being 1/2 to 2/3 of total, so payload is roughly the same/ a bit lower for the CEM but most people talk of the CEM as being a definite improvement over the eq6r. I guess there is more to it than just payload but I would have thought it would only be a marginal improvement.

[Anne S]

2 hours ago, Freddie said:

This has been a very interesting thread. The one thing I do find odd though is that the claimed max imaging load for the eq6R is greater than that claimed for the CEM60.

I checked with Altair Astro who are the distributor. They say the capacity for the Cem60 is imaging capacity ie 27kg. 7kg more than the Eq6r. I think that's a fair increase in capacity, and it's lighter so I can handle it on my own. I couldn't get the eq6 up from the floor!