Community Forums

Share:

Prism constant question  

Page 3 / 5

squowse
Posts: 968
Member
(@squowse)
500+ posts
Joined: 6 years ago

> > I know they are both 0mm constant, so should be okay, but I am not 100% sure.
>
> They're not the same. I put together this graphic to help me manage my prisms:
>
>

used another company's GMP101 the other day and found prism constant errors of about 2mm.
we were using it on 10no. spigots mounted on the walls of a tunnel. When I free-stationed from a different position, the prism positions that were being observed from the opposite side to the original observation gave errors 3-4mm. (basically the same test as described above). I was using a constant of -17 on a trimble instrument. (17.5-34.4).
Swapped back to using leica circular prisms and no problems - tight freestations from all positions.

As the GMP101 is meant to be the most accurate mini prism in the range I was surprised and wonder what I was doing wrong.

Reply
James Johnston
Posts: 632
Member
(@james-johnston)
500+ posts
Joined: 5 years ago

> I suspect the mini-prism setting on your instrument may match the second type.

We already checked that on TPS1200 and confirmed in LGO raw data. The correct constant was used. My unresolved question was more in regards of the ATR and how different sizes prisms affect its automation.

Reply
squowse
Posts: 968
Member
(@squowse)
500+ posts
Joined: 6 years ago

Unlike the Leica round prism the GMP111-0 isn't a nodal point prism so you would get small pointing errors if the miniprism wasn't perfectly facing the instrument. At steep vertical angles there will be a (tiny) level error with prisms that aren't nodal even if they are pointed correctly.

I don't see that it could affect the ATR centring.

Reply
Jim Frame
Posts: 5961
Member
(@jim-frame)
5,000+ posts
Joined: 9 years ago

> In this particular case, the mini prism used was a GMP111-0.

I wasn't aware of that model. Good to know.

I never did care for Leica's decision to hide the true offset from the user. It's been a source of confusion for anyone who steps outside of Leica's walled garden.

Reply
James Johnston
Posts: 632
Member
(@james-johnston)
500+ posts
Joined: 5 years ago

I was thinking along the line of the ATR using the radius of the prism in its calculation to adjust to true center. If the instrument is configured for a large prism / large radius and a small one is used, any effects on the computation?

Anyway, mm stuff which will not create a problem on this here job, just good to know for future reference.

Reply
2 Replies
JMH4825
Member
(@jmh4825)
Joined: 3 years ago

20+ posts
Posts: 44

I'm fairly new to robotics, with that said, is it poor practice to use a circular prism (large or small) while using a robot?  I'm not concerned about distance offsets,  I'm concerned with angular error due to pointing at different prisms.  I have used a circular prism as a BS and a 360 on a FS to turn sets just because I only own one 360 prism.  The BS prism is always pointing directly at the robot.  Will the robot not center in on the prism correctly even if the correct prism is set? 

Reply
Plumb Bill
Member
(@plumb-bill)
Joined: 9 years ago

1,000+ posts
Posts: 1429

I can only speak to Trimble or Leica.  Leica has a very accurate 360 prism if you take care to orient the yellow triangle back towards the instrument (it's looking at you, but used to align like a sight).  The larger circular prisms are better for all robotics as long as they're good quality prisms & pointed correctly.  Some even have aiming sights (like on some traverse sets).  The Trimble 360 prism is the least accurate of the bunch, but is till capable of tight work if you take effort to point on entire peice of glass back at the robot.  The passive 360 has an odd number of prisms, the active has an even number - so you can use a divider between the prisms as a "sight" on the passive one, and on of the prisms themselves as a "sight" on the active one.  The new MT1000 prism (disco ball thing) is supposed to be more accurate by default.  If you're dealing with a lot of slope none of the 360's are all that great.  If pointing accuracy becomes a major concern, say for tight control or layout, invest in a nodal prism.  They usually have a weird constant, but there is no error intruduced by pointing.

I haven't used a brand new Leica much, I know they actually don't turn to center - only apply a calculated offset.  I think prism pointing would still affect performance, though.

Reply
Page 3 / 5