36 replies to this topic

[Knight Sky]

Hello everyone!! I am back again here with another mirror making project and hoping to avail the good counsels of esteemed members just like the last time which was not too long back. I was helped by quite a few very knowledgeable members and I would forever be grateful for the knowledge instilled by them. Here's that mirror making thread.

 

When I had bought glass blanks (plate glass), I had bought 2 8" blanks just in case I damaged the first one badly - the first one turned out not too bad except that I still haven't been able to judge the minor astigmatism that's present there. Barring that I Have been able to have decently clear views upto 280x. At 280x though it's a little soft but that's partly also due to my mirror cell not having been constructed properly and the silicone that's holding the mirror, so I surmise, is distorting the ultra thin (standing 12mm at the center) mirror.

 

I started working on Mark II on 7th March or so, and the progress has been slow due to not being able to find optical grade CeO2. I spent about close to 20-25 hours polishing it with substandard CeO2 and then found a place (optician) in my town who gave me a little of good quality CeO2 a few days back. Progress has been expedited since then. For lucky me, grinding and polishing went without a single significant scratch this time.

 

After last nights session I realize that I have a decent parabola on the mirror except for a bad TDE and perhaps not quite a smooth curve. WIth Ronchi my guess was that I should be about 1/10 or so and then in the morning I took just one set of reading to confirm that. Since the room I work cannot be made a darkroom I just took a single reading to gauge where the mirror stood. More sets of reading will follow later at evening/night.

So the question I have is - how to deal with the edge now without messing up with the figure too much? I will add 70% null zone image along with more readings later. Can I use thumb or some very small star shaped felt pad stuck to thumb to bring down TDE? Thumb or a finger would give better control working the edge is what I am thinking.

 

Thanks a ton in advance and good to be back here!  

 

Mirror details:

Dia: 198mm (clear aperture) 1mm bevel

glass type: plate glass

thickness: 19mm (now standing about 15-16mm at the center)

FL: 2343.15 (I need to measure it more accurately but should be off only by 8mm or so)

Intended F ratio: f/6

 

Ronchi grating is 152 LPI (a huge thanks to Lucas for sending these lovely Ronchi screens!!). Both the Ronchis are outside RoC.

 

 

 

 

EDIT: I have been cautious this time to make sure that I have taken all the precautions to avoid astigmatism this time and having run Ronchi through multiple diameters (separated by about 45 degrees or so) I have not been able to see any significant signs of it. But I still need to check it again as it may still come as a surprise.

Edited by Knight Sky, 08 April 2022 - 06:58 AM.

[Knight Sky]

There seems to be a TUE/zone/kink at about .85 mark, that does look a little unpalatable/ugly. 

[davidc135]

I would take Ronchigrams with the same number of fringes inside focus and compare the edge kinks. Some of whats shown above may be diffraction effects rather than t.d.e. But you took a Foucault image? That should show it.

 

It looks to me that the miror is under-corrected in the outer third or so zone by maybe 0.2-0.25 waves pv. I had a look at Mel Bartels' Ronchi matching site at www. bbastrodesigns.com.

 

The centre may be a touch over-corrected.

 

David

Edited by davidc135, 08 April 2022 - 12:44 PM.

[Scott E]

I agree with David. It's easy to interpret what you're seeing at the edge as turned down when it may be mostly or wholly diffraction. And Figure is showing what you need to do; deepen the whole center to correct the zone. Pay close attention to the transverse error. This reflects the effects of the steep slope you see in the surface profile graph. Despite the 1/14 wave amplitude it's the slope that kills. It's sending light reflected from the outer zone outside the Airy disc. You need to get that number much lower.

 

In the Ronchigram the lines should be fairly straight in the center and gradually get more curved towards the edge. Yours are nearly straight outside and curved near the center, opposite what you want.

 

And I've found that astigmatism shows best when you can see the whole mirror at once. That is, just use a knife edge and Foucault test (no mask). Look for any sings of a yin-yang shape.. But I don't see any signs of it in what I can see here.

 

Scott

[Knight Sky]

Seems like I was overly optimistic in my judgement and I misinterpreted Ronchi corrections in the center and the edges. As both of you have pointed out center infact is overcorreted and edges are undercorrected.

 

I just took another two sets of readings at a different diameter to gauge the parabolization status. A big caveat here is that I am still using the older Couder mask constructed for 197mm fl. 2324. Whilst this one is 198mm fl. 2343.15. So this will have a shift in zone centers and overall correction required should lower even further, throwing these readings further off. I will be building the new Couder mask soonish. I am also itching to try suggestion of Ccassie to use Everest pinstick style marker (tape as he suggested).

 

z1 z2 z3 z4 z5

 

0 0.49 0.95 1.41 1.75
0 0.5 1.05 1.35 1.7

 

 

Seems like my Foucault tester improvement has made it worse, I just cannot move it smoothly anymore and it suffers from backlash. In anycase the TDE appears to be massive, but I think a large part of that is due to the fact that I use 4x zoom on my mobile?? DSLR with 85-100mm would show it a little less prominent I think.

Edited by Knight Sky, 09 April 2022 - 01:21 AM.

[chantepierre]

Hello Knight Sky !

Fantastic to see you working on another mirror. If I may, maybe consider building a Bath ?

For me, learning alone since I started with covid, and there's noone nearby to learn with, the Bath lifts a lot of uncertainty and subjectivity from my shoulders.

But, when I first started to learn to use it, it agreed with the results I got with Foucault Unmasked. That also can help you assess the mirror !

Have a nice day,
Lucas

[Knight Sky]

The course of action I am imagining here is that I try a star pressed lap and give TDE a shot. If the figure goes out of control, I get back to sphere and start from the beginning. If the TDE improves without impacting the rest of the inner zones then all I need to do is to work on reducing the correction just a bit, that should take perhaps about 1 session of max 5-6 minutes.

[Knight Sky]

Fantastic to see you working on another mirror. If I may, maybe consider building a Bath ?

Thanks Lucas! I see you have started working on another mirror as well...great to see that!!   Coincidence that we almost started on our second mirrors about the same time. 

 

I have sent Dale's design to a friend to have the 3D stage (+ Bath) printed, it's been a while and I think he may have encountered some issues. I havent' checked on it with him in a while. But I hope to test this mirror on Bath and also on Foucault Unmasked.

[davidc135]

Yes, plenty of tde. One approach is to use very short and quick centre over centre strokes which may also lift the mirror centre.  David

[Knight Sky]

Yes, plenty of tde. One approach is to use very short and quick centre over centre strokes which may also lift the mirror centre.  David

David, that stroke is what I used earlier to fix TDE, but in my case that stroke tended to create ripples in the inner zones. If I can find a way to deal with those ripples then I think this stroke is, as they say, pooping on two cars by a single dropping (of a bird). Not surprisingly though, anytime I have worked ToT, I have managed to create a TDE even if the stroke length is less than 1/2D. Only way I can think of avoiding those ripples is to not touch those inner zones. Maybe slightly slower 1/4D CoC stroke would work better?

[ccaissie]

There's a sharp narrow TDE, and because your edge is undercorrected, I would use a technique to lower that edge and expect the TDE to resolve...you've got to smoothly remove all the interior glass down to the level of the desired edge.

 

That being obvious, you need to plane it down carefully, especially if the rest of the mirror is nice.  Full lap TOT, accent over outer zones, etc.  My theory..that the edge of the lap does a lot of work where you stop and go, avoid a stroke that lets the lap edge reverse over that 85% zone, where you have a "trough".  I hardly see it in the Ronchi, but your measurments reveal it.

 

OR, once everything is really just right, increase the bevel.  How wide is this TDE?   The Ronchi images show just a slim thread of light band invading the black fringe at the extreme edge.  1mm?  Increase the bevel or face grind it against a flat glass plate with 5-12 micron to put a .5mm face and watch the diffraction ring show nicely.  Heresy, I know.

 

 I am also itching to try suggestion of Ccassie to use Everest pinstick style marker (tape as he suggested).

 

Or put sharpie marks at the zone centers, set the KE at the ideal ROC settings and see if the shadow crests hit the marks.

If they are inside the pair of marks, you're overcorrected.  If this method appeals to you then make a stick with pins or painters tape with notches for repeat tests.  The diffraction effect on any of these markers lights them up nicely, even in the deep shadows.  

 

I use a lathe apron from an 18" Hendy that I mark the dial on for whatever mirror I'm working on. I set the KE to each zone and read the crests/pins/marks.  Because I can easily 'Zero' the dial, I will often pick a zone that is quite clear re: shadow crest on mark, a favorite is the 70.7%, then reset the dial to that mark and then measure to other zones..a self-check going in and out to eliminate errors in backlash and errors "between the head phones".

 

The original Everest article (ATM series Vol 2)  described making a barr scale to measure the KE fwd and back.   A dial indicator is the preferred instrument these days.

 

 

 

 

 

Attached Thumbnails

• 

Edited by ccaissie, 09 April 2022 - 06:30 AM.

[dogbiscuit]

That mask will work for that mirror.

The only thing different is you target knife movement from each zone to the next is .4 mm.

 

Looks pretty good except the edge.  Is the lap diameter also 198mm?

[Knight Sky]

 

There's a sharp narrow TDE, and because your edge is undercorrected, I would use a technique to lower that edge and expect the TDE to resolve...you've got to smoothly remove all the interior glass down to the level of the desired edge.

 

That being obvious, you need to plane it down carefully, especially if the rest of the mirror is nice.  Full lap TOT, accent over outer zones, etc.  My theory..that the edge of the lap does a lot of work where you stop and go, avoid a stroke that lets the lap edge reverse over that 85% zone, where you have a "trough".  I hardly see it in the Ronchi, but your measurments reveal it.

 

OR, once everything is really just right, increase the bevel.  How wide is this TDE?   The Ronchi images show just a slim thread of light band invading the black fringe at the extreme edge.  1mm?  Increase the bevel or face grind it against a flat glass plate with 5-12 micron to put a .5mm face and watch the diffraction ring show nicely.  Heresy, I know.

These are all great suggestions Ccaissie! Let me ponder over a little more and hear dogbiscuit's response and weigh in which one I can execute easily. David's response also sounds pretty tempting, something that I know works, except with the caveat I mentioned (ripples).

 

I'll definitely try pinstick once again, last time I tried I failed to judge where crest began and ended properly. But I think I have slightly more trained eyes now than I had back then.

 

That mask will work for that mirror.

The only thing different is you target knife movement from each zone to the next is .4 mm.

 

Looks pretty good except the edge.  Is the lap diameter also 198mm?

Good to know dogbiscuit that I can use the same Couder. I might just try Ccaissie's method to see if it works for me, that method seems a little less complicated than making a Couder mask.

 

My work bench

 

 

Lap

 

 

Padding is a 5-6mm rubber mat. Padded by a 3mm thick foam sheet.

 

 

EDIT: dogbiscuit, no the lap dia is 200mm (original mirror dia). This thought did cross my mind but I wasn't sure that 1mm extra would be harmful.

Edited by Knight Sky, 09 April 2022 - 03:25 PM.

[dogbiscuit]

To me it appears the outer 10 mm of the lap is not in full contact.

Due to the micro faceting there is some contact, but smaller parts of the microfacets contact the lap nearing the edge. This is seen by more white cerium in the micro channels nearing the edge.

 

Don't know for sure if that is the cause of the tde.  Maybe. 

I see the lap isn't beveled so no reduction of the lap diameter for bevel, that's good.

Is the diameter of the lap less than the mirror?  For me that would make getting a good edge more difficult.

I think more pressing will help the edge.

 

The figure...

 

If I had that but with a good edge I would use accented pressure TOT on the outer zones, 5  4  and 3

Lots of details to exactly how I would do that.

But the edge isn't good so I would press better and use some 1/4 and 1/3 Ws to reduce correction and see if the edge gets better.

 

I have more stuff going on here than I can handle so I won't have much time to comment but I'll be watching.

 

[Knight Sky]

Unfortunately the last session didn't go my way. I went daredevil and used roughly about 1/8D (trying to avoid 85% trench) fast CoC MoT strokes (2 turns around the barrel 7 strokes at each clock position) but created a trench at 45-50% mark. There's a slight improvement in the TDE in that I could see the ring of fire sneak a peek sporadically at points where, perhaps, the strokes were more dominant. Afterthought is that I should have pressed a star extending up to 80% zone.

I think that the damage is not too bad and should be salvageable. I will work on it later at night if possible.

 

 

 

[dogbiscuit]

OK.  I said I was just going to watch,  but I got a minute.  Waiting for glue to dry. 

 

There is not much chance you will avoid a hyperbola trying to bend that into a smooth parabola, unless you go back toward a sphere and come at it again.

 

Maybe I'm wrong... what you did that last session just might have reduced correction enough to allow correcting the outer zone without making a hyperbola.

 

The problem is not enough correction in outer zones and is or was too much in the central zones.  Common problem.

Add some correction in the outer zones. You could try these and see how they works.

 

Press.

MOT 1 turn each

About the number of strokes shown in the drawings.

Don't do too much and risk going far into hyperbola land.

 

On occasions I've seen the edge improve with strokes like this... but more often not.

Your edge isn't so good so you wouldn't lose much trying these strokes now.

 

Overall correction will probably end up over corrected.  But at least this would be some practice at getting correction in the outer zones so next time you have an easier time. Just see if you can make the outer zones correction more like central zone correction.

 

If you do go back to a sphere, some of this sort of stroke to correct the outer zones along with the normal parabolizing stroke can make more uniform correction center to edge.  Don't let the center correction get so far ahead of outer zone correction.  When correction gets near 100% it is too late to add correction in the outer zones unless you back up toward a sphere and it isn't so easy to do that smoothly.

 

There is Texereau's edge correction parabolizing stroke with higher density of strokes near center of the W, diminishing density toward the sides of the W that work the outer zones more, but some people have trouble modulating the stroke density the same way at every clock position.  So maybe you can do better working only the outer zones (strokes near center correct outer zones). And in your current situation you don't want any additional correction in the central zones.

 

The long strokes will probably smooth the kinks a little bit.

 

I would prefer TOT accented pressure for this, but it's more complicated to explain pressure and stroke.  My glue would be dry long before I finished the details.

[Knight Sky]

The suggested strokes seem logical (with my limited understanding) to me looking at the figure on the mirror. I just had this passing thought though, since I can see that the correction in the center has reduced a bit and that the edge seems to have improved ever so slightly, would it make more sense to try these stroke after another session of the 1/8D strokes that I last tried to improve the edge further? Why I am thinking this way is, if with these strokes the edge improves then all is fine and dandy, but if only the figure improves then we are sill stuck with edge problem. Well, we can always deal with that by chamfering off a mm or two if required in that case given it's not exorbitantly wide as an alternative. So, I see these two paths that could be taken at this point. Let me know what you think.

 

Due to my work room not temperature controlled and the temperatures in the town already hitting upwards of 40 C, I may not be able to work until the next morning, when the temperatures are at its lowest at about 24-25 C. I was thinking of keeping the lap in a cold water bucket (about 15-20 C) and working since sessions are not going to be longer than 5-8 minutes, lap could maintain a relatively consistent temperature for that long given that I am using 1:100 watery CeO2 slurry. Would it work? Has anyone tried it? My assessment of the ideal working temperature for the lap is about 20-25 C.

[dogbiscuit]

I Don't think another round of 1/8 D strokes are going to help at this time.

Let's see what it looks like after the two strokes to correct the outer zones.

 

What kind of pitch are you using?

 

What are temperatures 4  6 or 8 hours before morning when it's 24 or 25 degrees?

 

Pressing is important.  My opinion is it would be better to press several hours until the mirror and lap, and workroom cool to good working temperature 24° or 25° if that is what you think is good. I think this is you best chance of making a good edge, good press and not making sudden temperature changes of lap and mirror prior to starting stroking.

 

Soaking or dipping the mirror or lap in warm or cool water hasn't been part of my methodology since many years ago. It didn't work for me.  I get better results starting stroking immediately after pressing.  I rarely experience 40°C temperatures and never in the workroom. Completely outside my experience so I can't say it wouldn't be helpful in your circumstance.

[Knight Sky]

I am using Rosin mixed with half a teaspoonful of turpentine. Pitch is on slightly softer side. The channels were taking about 2-3 hours (roughly) of polishing time before requiring rechanneling. Temperatures are in the range of 30-25 C starting from 11PM or so until 5:30-6am or so but then rise steeply after the sunrise.

I have not worked on the mirror as yet but took a set of readings. There is a little bit inconsistency in the readings and it seems that either I need to up my game with ke readings or that the tester is misbehaving. I'll try to improve on both.

 

z1 z2 z3 z4 z5

 

0 0.18 0.82 1.21 1.46
0 0.27 0.86 1.19 1.41

 

I'll try and work when the temperatures are under 30°C, as getting up that early would be almost impossible.

[dogbiscuit]

There are sharper kinks in the curve and that can cause difficulty getting a reading if the kink is in one of the mask zones.

 

OK on the temperature.  I recommend checking the channels and trim if necessary, apply thick slurry and begin pressing at about 11pm or so and it will be ready to begin polishing in the morning when temperature is coolest. The cerium oxide applied before pressing will probably be enough to do the entire session.

 

Thanks  for the information on the pitch.  How much Rosin are you mixing with half teaspoon of turpentine. There is another atm in India that is making a mirror and his pitch is too hard. We know he needs more turpentine but we have no idea how much. Your surface looks nice and smooth so seems you have a good mix.  He is dealing with temperatures similar to yours.

[Knight Sky]

dogbiscuit, I did finally execute the strokes suggested in your post #16 in sweltering heat (ok, the temp in the room might have been between 30-35°C). I did press the tool for about 1 hour. My initial assessment with the Ronchi is that the corrections have gone in well except that there is a slight raised mound in the center and the TDE has not improved at all or has worsened. Some zones may have been slightly overcorrected. These Ronchis were taken immediately after the session so the mirror quite possibly has not reached equilibrium. I will take better images later in the evening/night along with ke readings.

As for the quantity of Rosin, I did not weigh it. I just poured turpentine in the amount I felt would give a decent stiffness (testing it by cooling it in water and then doing the thumb nail test). In my estimate it wasn't more than 350gms. My pitch lap came about 6-7mm thick with tool conforming to the mirror curve. Just to mention here that I am using a different type of turpentine (artist quality double distilled turpentine used by oil painters) which is slightly more viscous (due to being slightly old) than the one used in woodwork and house painting jobs. So the quantities I am using most likely won't apply to others.

 

 

 

 

 

If the ATM you are talking about is with the screen name PRS then I know him and will let him know how I am making my pitch lap.

[dogbiscuit]

Looks better. Much as was expected, more correction outer zones 4 and 5.

Some more correction in zone 3 could blend  from zone 5 all the way to zone 2 and then only zone 1 would need a little correction added.

Question is will it be over corrected and if so more than ~10% over? 

Even correction between 90% and 110% would be a pretty good mirror.

 

Will wait to see what the zone numbers say.

[PRS]

dogbiscuit, I did finally execute the strokes suggested in your post #16 in sweltering heat (ok, the temp in the room might have been between 30-35°C). I did press the tool for about 1 hour. My initial assessment with the Ronchi is that the corrections have gone in well except that there is a slight raised mound in the center and the TDE has not improved at all or has worsened. Some zones may have been slightly overcorrected. These Ronchis were taken immediately after the session so the mirror quite possibly has not reached equilibrium. I will take better images later in the evening/night along with ke readings.

As for the quantity of Rosin, I did not weigh it. I just poured turpentine in the amount I felt would give a decent stiffness (testing it by cooling it in water and then doing the thumb nail test). In my estimate it wasn't more than 350gms. My pitch lap came about 6-7mm thick with tool conforming to the mirror curve. Just to mention here that I am using a different type of turpentine (artist quality double distilled turpentine used by oil painters) which is slightly more viscous (due to being slightly old) than the one used in woodwork and house painting jobs. So the quantities I am using most likely won't apply to others.

 

 

 

 

 

 

 

 

 

If the ATM you are talking about is with the screen name PRS then I know him and will let him know how I am making my pitch lap.

Hi all, here I am. I was facing problem with the rising temperatures but I have poured a new lap and it's now working fine. Stuck with taking readings with ke so gathering more info about it. Meanwhile I have returned to a sphere and start parabolizing afresh.

Edited by PRS, 11 April 2022 - 12:50 PM.

[Knight Sky]

Readings aren't looking too bad. Only if the central mound/plateau could be lowered, zones smoothed and a crisp ring of fire visible, this mirror could be called done. Readings are across one diameter only.

 

z1 z2 z3 z4 z5

 

0 0.17 0.66 1.08 1.41
0 0.09 0.5 1.02 1.38
0 0.13 0.59 1.01 1.35

 

avg
0 0.13 0.583 1.037 1.38

 

 

So close yet so far.

[dogbiscuit]

The video shows TOT accented pressure on zone 3 of a 4 zone Foucault tested mirror.

https://www.youtube....h?v=7PdE2wqUJZo

Same thing for you except you would be working zone 4 of your 5 zone tested mirror.

Working zone 4 will increase correction from zone 4 to 5.

It will probably also reduce correction from 3 to 4

This stroke is worked to only the right side at each step around the mirror.  Left side is ok if you like that better. The zig zag in this case represents the path of the lap's right edge over the mirror.

Apply about 1.5 kg of pressure at the lap's edge.  Left hand only guides the mirror, no downward pressure on the left.  I press down on a kitchen or postal scale to get the feel of the desired pressure. 

See that I have made a "bullseye" of alternating black and white circles representing the mask zones.

I watch the right edge of the lap and stroke the edge in the W pattern across the zone.

 

1 trip around the mirror might be all you need.  Don't do more without testing to see if and how fast it is working

 

Correction percentages zone to zone.

1 to 2    33%

2 to 3  113%

3 to 4  113%

4 to 5    86%

 

 

If you decide to try this, watch that video a few times to get familiar with what you will be doing.  Do it the same at each step around the mirror. Same zone, same number of strokes, same length of strokes, and same pressure.

 

If that stroke works good you will only need to work a small area in the center.

I'll send you something on that a little later.