[dhardt]

If a 0i control could do what a 31i could do...that would be making the 0i described by FANUC as "4 Axis simultaneous control, simple cutting applications" do what the 31i does "20 Servo Axis, up to 5 Simultaneous (B5), 6 Paths, 8 Spindles, High precision."

Thank you for not asking it to do what the 30i does..."32 Servo Axis, 24 Simultaneous, 10 Paths, 8 Spindles, Extreme precision."

If you want to be realistic, you might request that it act like one of the nearest two in cost and abilities: either the 32i or the 35i
32i:"12 Servo Axis, 4 Simultaneous, 2 Paths, 8 Spindles, General cutting."
35i:"16 Servo Axis, 4 Simultaneous, 4 Paths, 4 Spindles, dedicated processes like transfer"

Seriously. All of these are options available in most configurations; it just takes money at the time of purchase.
Like your local car dealership, we provide the machines, you provide the cash...But unlike most cars, this vehicle prints money for its owner (in the right hands)
You wouldn't buy a stock street car to win at a racetrack, because not everyone wants to pay for their car to race.

So we have improvement options. Buy the improvements you want at the cost it takes to include them.

It seems like simple economics to me, but I dropped out of high school...so what do I know.

[Green0]

We ran into cross path communication in this last program in tool load monitoring, so we dropped that one tool but lost 4 hours on tool load monitoring settup working all the issues- because the machine doesn't teach synchronous, and will not take recognize manual input of untaught tools when program organization doesn't allow complete teaching in one shot per path- it needs the upper taught, the lower taught seperately and that is tough to execute when you are crossing and requires making some scrap parts. 5 hours later on a second shift, M35 milling lower main, M135 milling upper sub, the machine locked up like the timing put it into a thought process it was looking for a M136 synchronous milling command. More cross path communication. No M136 was commanded. We didn't have a settup machinist there on second, so we lost 12 hours on second and third shifts to that one.

That's 16 hours out of that 24 hour block lost on those issues. We are putting a two second dwell on the upper to theoretically change the machine's thought process and get it to stop thinking it is watching the runup to synchronous milling. That issue happened 4 hours into running production today before our theory and hypothetical work around attempt. So it is rare- timing must be tight is our theory, but cross path communication is happening.

These issues represent a machine that is being sold and is essentially in an incomplete state of development. Korea should not have stopped selling 31I TT's without the I series being ready for prime time. It was a machine marketed as saving money we didn't want to save, and as an upgrade. It is a serious downgrade. I don't think all shops are even equipped to deal with these kinds of issues. Some of them would just be handicapped forever. We aren't unconcerned about being severely handicapped for a long period of time on these issues.

We wanted to scale an operation on these machines. This is a pretty messed up situation. If I need a working twin turret machine, this might be time to consider Nakamura, or Okuma. It's not ideal to have to start thinking about which other brand machine can work, because the existing machines were changed from functional to disfunctional. And being told to hold on and we'll get it for 8 months and continuing is really hard.

[Green0]

If a 0i control could do what a 31i could do...that would be making the 0i described by FANUC as "4 Axis simultaneous control, simple cutting applications" do what the 31i does "20 Servo Axis, up to 5 Simultaneous (B5), 6 Paths, 8 Spindles, High precision."

Thank you for not asking it to do what the 30i does..."32 Servo Axis, 24 Simultaneous, 10 Paths, 8 Spindles, Extreme precision."

If you want to be realistic, you might request that it act like one of the nearest two in cost and abilities: either the 32i or the 35i
32i:"12 Servo Axis, 4 Simultaneous, 2 Paths, 8 Spindles, General cutting."
35i:"16 Servo Axis, 4 Simultaneous, 4 Paths, 4 Spindles, dedicated processes like transfer"

Seriously. All of these are options available in most configurations; it just takes money at the time of purchase.
Like your local car dealership, we provide the machines, you provide the cash...But unlike most cars, this vehicle prints money for its owner (in the right hands)
You wouldn't buy a stock street car to win at a racetrack, because not everyone wants to pay for their car to race.

So we have improvement options. Buy the improvements you want at the cost it takes to include them.

It seems like simple economics to me, but I dropped out of high school...so what do I know.

That's not true. I sat a table with Sales and applications, and literally told them I didn't want a different machine. I said I was concerned about Tool load monitoring and cross path communication because we experienced these issues in a 2013 machine that Doosan had a solution for in 2017. I don't blame them- sales is their job. The factory's job is to give them quality machines to sell. They had no personal experience so they were speaking to the brand quality. I was hoping and wanting them to completely prove my fears totally un-valid. I like the brand. I just don't know why this isn't priority one for Doosan to have machines that actually work right.

We were told Doosan was no longer importing the 31I TT1800SY, and that purchasing one was not an option. We were also explicitly verbally guaranteed that this I series machine had every capability of the 31I series machine, and more- they went through brochures line by line (there are a few more supported options thermal comp, MCS, Look Ahead, but possibly missing CPU power), and would work as well, or better or they would work with us to make it right.

The second machine was verbally supposed to hit our floor with the solution to the first and was paid to be fully integrated before it hit the floor. So that didn't happen either. $11,000? I couldn't give a **** about $11,000 on this. I needed the machines to work and not give us problems. I asked to pay more for 31I before the first I series was purchased. Korea is assuming shops won't pay, and they are changing specs that box the customers in like us here.

This CPU issue is beginning to look like a serious hit to capabilities.

[Green0]

They said look you're going to save $11,000. I remember saying, I'd rather pay $14,000 for a 31I and peace of mind. Like all new machine customers, the reason to buy is a production requirement. The only way to fulfill is to drop in a solution that rapidly functions to fulfill capacity issues. 8 months isn't rapid. And the machine isn't able to operate to purchased and expressed capacity.

The crazy thing is that somewhere in a room, intelligent people liked this as a solution.

I called the Fanuc regional manager, got voicemail and walked out on the floor. The $264,000 machine is washing the window- stalled on m35. M135. Green light doing nothing but running 50% recommended duty cycle live spindles in the air because it doesn't want to read two lines of code at the same time consistently.
Attachment 459678

[Green0]

That picture says a lot. This isn't a race-car. This isn't even a car. No car just stops with the gas pedal down that I'm aware of. That's a BUG- they call that a lemon. It's JUNK. The machine is taking random UNAUTHORIZED BREAKS. If it was an employee companies would be firing it.

This should be a race car though. I wanted to pay for a race car. They just said, we can't sell you that much horsepower. The machine wearing the badging of a racecar breaks down now. Say hello to technological advancement.

I couldn't reach the Fanuc regional manager. We couldn't reach Jaimie Ochs. I can't seem to reach anyone.

Cross path communication is a big problem for both operation and TLM. TLM not allowing sychronous path teaching is kind of a disaster. TLM not allowing teaching of an individual problem tool is a problem. TLM not allowing a zero value to kill monitoring of an axis or spindle is a problem (especially with erroneous cross path communication). No save out button is an issue. No recognition of manual data entered is a problem. These issues together are making this machine fail to succeed and advance the Doosan brand on the floor. The twin is producing more like a single turret machine. It's losing a cost basis. That's over-complex requiring extra expensive guys to operate for no gain in productivity.

This situation needs total and final conclusion. We can't talk to Korean people on the floor. We don't speak Korean. Those guys need real English interpreters.

[dhardt]

Jaime Ochs was out today (it happens).
I contacted him at home.
He is awaiting your email at [email protected]
He will assist you, after hours
I also contacted Douglas Rizzo, a former Doosan Applications trainer. He is at dinner, but will be checking in here in about an hour to see if he can assist.

As stated before: My cell number is in the footer of every post that I have here. If you can't reach people...CALL ME. I have personal phone numbers.

[Green0]

It would be cool if he was able to work with us. We are ok working during business hours if that's better for him. We put this into service tonight. The big deal is knowing this is a high priority and that it matters for Doosan that they have machines that actually work to specification. Knowing someone is actually full time on it till it's right is important at this point. I think it's been more of a back burner slow development thing. I guess they have no actual machine testing in Korea- they use simulators according to Doosan, and maybe that's part of why it's not resolving.

I'll give him an e-mail. If they could ever work this one out, it would be a really good idea to freeze that machine and let people buy it with confidence for a while. We are going to need more machines, but we need to know they are going to be correct or we would have to be forced to find a more serious turning brand.

[DouglasR]

I'm here..
I think.

OK, humor aside - are you using M264/M265?
M264 is "other pulse coder" - meaning whatever path it is on, it will look at the OPPOSITE path - Upper will look at sub, lower will look at main. It MUST be cancelled with an M265 or the machine may lock as it can't see what the other path is doing.

My cell is 862-203-9352. Call up to 9 pm

[dhardt]

Use his email if you wish to read him, as he carries a version of it everywhere he is. His office phone is only in his office. (AGAIN, call me if you have trouble reaching people, especially during the Lockdown Era, when everything is in flux as to who can enter a building and who is being 14 day quarantined)
Regarding the concept of testing machines with simulators, it sure as heck wasn't that way on either of my two visits. I got a tour each time (of different sections of many massive assembly lines) and at the end of each line there are many multiple machines at the QC area being FULLY checked. Some are then shipped whole, while others are disassembled into containers for the protection of their components.

...but all are fully tested.

- - - Updated - - -

I'm here..
I think.

OK, humor aside - are you using M264/M265?
M264 is "other pulse coder" - meaning whatever path it is on, it will look at the OPPOSITE path - Upper will look at sub, lower will look at main. It MUST be cancelled with an M265 or the machine may lock as it can't see what the other path is doing.

My cell is 862-203-9352. Call up to 9 pm

Eastern

[DouglasR]

I will email him immediately.

[dhardt]

Green0,

Douglas is trying to reach you, but your PM que is not accepting.

Jaime just called and has received no email, so he is going back to finish his evening. If you need him, please email him tomorrow, he has confirmed that he will be in the office.
[email protected]

[DouglasR]

My email is [email protected]
My phone is 862-203-9352

[Green0]

I e-mailed Jaimie Ochs last night a reply to his e-mail. He had a voicemail from me with my number. I got delivery of a laser machine today, and last night one of the guys indexed a main approach live head into the way covers on a Lynx 220 LYSC, and we had Ellison here today for an alignment for that abnormal and stupid issue so we were a little distracted.

Jaimie had mentioned he had a TT1800SY on the showroom floor and would try to replicate the M35/M135 issue. We didn't hear anything from him. I got no feedback from Ellison Service on the warranty ticket we opened for Tool Load monitoring problems and this cross path communication issue. They knew we were working with Doosan applications on the TLM for the last 7.5 months though already.

The machinist who had the brainfart and indexed the turret into the way covers after mounting the wrong tool in the Lynx saw this thread this morning and the M264 / M265 stuff. He told us about that. I and the twin turret machinist talked briefly with Doug Rizzo on it, and we appreciated his help. We implemented that strategy. It was weird in that we were using a 31I TT1800SY program that was proven and ran and were now needing to add codes we never once used on the 31I TT's. We still had the M35/M135 hang one time today in the last 8 hours with the M264 codes on top of all the cross machining live tools and M265 at the last tool before the re-cross. It appears to have reduced about 90% of the system hang issues we were having but they are still happening less frequently.

The logic of that is odd.
If the upper sends M135 (this means hey Upper activate milling mode on subspindle)
M264 (No really I want to mill something on the subspindle like I said)
G code for milling
M134 (Upper path put the subspindle back into turning mode)
M265 (We're leaving cross path mode)
G28 reference
M01

G28 Reference
G54 (upper returns to main)
M34(Upper puts main into main turning mode)

The information was already there. I feel like Korea manufactures complexity into the ladders over time, and complexity is why we have issues like cross path communication issues. I believe it would be true to say Korea made a more complicated ladder in this machine.

We're still hoping to get some dramatically improved TLM update and a ladder solution.

[Green0]

Korea doesn't get that we want uniformity. The whole floor is blue so it can be simple. We have 6 different types of load monitoring on 3 models of machines 2015-2000 production right now. We need simplicity to keep these staffed- as it is HR is tough. Added complexity can only make it tougher.

They just pushed more parts at my one good TT1800SY today. It sucks those other two just don't produce as quick as that good old one. If Doosan could solve this TLM and Communication stuff, we could use another one or two. It just would be insane to think buying another without seeing this resolved was going to bring us a resolution and make these right.

[DouglasR]

It's my pleasure to help.
Again, please call if you need.

862-203-9352
[email protected]

[DouglasR]

Curious as to how you are doing..

[Green0]

Curious as to how you are doing..

Sorry whenever we get sidetracked nothing stops so we end up buried on all the things we were supposed to get done in the middle, and have to unbury ourselves before we can come up for air.

The Ladder modification went in about 4-5 hours before the machine changed over to another part. We were having the incidence of the hangs every 30-60 minutes before M264 implementation, and every 55- 240 minutes after M264 implementation so that turning function was helping eliminate milling hangs about 50-75% of the time. We were told that the M264 was intended for Upper sub or lower main side turning, and wasn't intended to support milling but it appeared to have some cross-over functionality in this case.

The ladder edit ran about 4-5 hours with no hangs with the M264's and M265's in the code. So it appeared we either resolved the cross path hangs, or at least didn't hurt that situation. It was unfortunate we only had 4-5 hours to work with that before a program turnover but I feel like the ladder improved the machine's control function. I think it would be ideal to get the I series plus running like the 31I (not needing the M264/M265's to run code), and I think the ladder edit, either achieved that, or helped.

The Tool load monitoring issue has not been resolved. I think the situation there is that the reduced computing power of I series plus cannot handle the amount of data that legacy 31I tool load monitoring is looking at, so I think the software has to be modified to natively reduce the bit rate of data received by 50-60% in order to derive the ability to function (and teach) simultaneously on both paths, but also some edits are required to get the TLM table export and import buttons to appear, and to allow zeros to be recognized as "ignore this item", and to best case add the ability to re-teach single tools without clearing the tables. Objective should be to also recognize manual entries of numbers. Right now TLM ignores manually entered numbers and will not monitor a manually entered tool so there is no flexibility when one tool doesn't get taught- or is stubborn to teach.

Until Doosan develops Tool load monitoring to the state that it should have been when these machines first reached the market, end users will be losing time on every changeover in ways that reduce the number of Doosan machines the floor staff at customers can support. So TLM is getting in the way of machine sales and customer growth, hurting Doosan, Ellison, and Ellison customers.

I feel like the $11,000 cost reduction per machine has cost us ~110K on the first machine, ~$43K on the second, and also has cost us probably 150,000 in lost business we could have supported if we could buy another machine with confidence over this time. Right now the TLM wet-nurse time is just too great to leave us enough room to realistically babysit another machine on the floor. So the $22K savings has probably reached $303K cost, overlapping loss is $281K. The larger that number gets, the more obviously I'm a fool for believing this will go away. Economics say this control change was colossal mistake. I don't know how difficult the TLM software development is, so it's tough to say how high a priority it has been with Doosan. I guess I'd be in a lot happier place if Doosan could have sold me 31I machines, and waited to offer the I series plus after developing that product to a state of market readiness. The estimations are conservative- for example, I'm not accounting when one TT is promoting down time on two others. I feel like the 31I's during proveout would promote one other TT down for a short time like 1 day if the timing was right, but the I series plus machines I've watched proveout complications promote down time on a pair of TT's for more than one day several times. That's where this lost time doesn't work in this machine category- this is where companies like Nakamura have bragged about ability to promote greater profit on the floor. This I series plus twin path is selling that concept a lot better than it classically sold against Doosan.

If anyone reading can speed up the TLM solution I'd appreciate it.

[Green0]

I was contemplating the meaning of life with a $13422 repair bill on a coolant damaged $26,000 UR Robot unloader for a Puma 2100SY II when I walked out to tell my production manager we had paid the UR robot ~$78 an hour for service rendered. I guess we should have paid a person from a different department to do that- hindsight being 20-20.

I looked up and the first of the two "new and improved" TT1800SY's was washing the window running a spindle, hung on M200 (Tool load monitoring OFF).

If it's not going to see the probably erroneous load until M200, it might as well let the process continue. After asking for information I was informed we're losing 33% of time on one of these two newer machines still. The other I'm hearing is less problematic, but 33% is very severe. We don't have a spare (miscellaneous) qualified machinist, who can stand in front of it all day to tend it like a formula one pit crew. We are renting a freeway billboard to try to hire one, and we have indeed listings but we haven't found one yet.

I need to increase capacity but I can't make a purchase. Doosan hasn't fixed the first two that are under warranty. The latest guidance was May for the new Tool load monitoring release. This problem has no priority to Doosan. We're at 7 months with that machine. They don't seem to care how much time we lose. They have to know what's wrong, and how to fix it and just aren't applying the time.

Attachment 460988

Attachment 460990

[Jim Dawson]

Sounds like you are having a really bad day. I don't know much about production job shops, but it seems like you have a lot of crazy problems. Maybe this is the norm in the industry, but I can't imagine having to put up with these kind of problems. I'm a manufacturer so I do some production work and my machines just run as needed, and run dark half the time. If I were running into the problems that you have had I would seriously look at retiring and go sit on a beach somewhere.

[Green0]

Sounds like you are having a really bad day. I don't know much about production job shops, but it seems like you have a lot of crazy problems. Maybe this is the norm in the industry, but I can't imagine having to put up with these kind of problems. I'm a manufacturer so I do some production work and my machines just run as needed, and run dark half the time. If I were running into the problems that you have had I would seriously look at retiring and go sit on a beach somewhere.

Yeah if Doosan never solves it, I start to look crazy advocating for solutions effecting the success of my 45 employees and their families don't I?

Quitting is really what you are saying. Quitting isn't an option. The only option is the direction of a path to success.

Early I think people thought this was a ladder issue. If there are smart people at Doosan (and I believe there are) they probably knew many months ago there was a major software issue here and that's where I look stupid because it took me a long time to understand that. Granted you have to realize I don't speak Korean so I got bits and pieces of information from our production manager third hand and have to patchwork quilt them together to try on my not solution end of the problem to try to figure out how to influence resolution.

I'm really surprised that Doosan was OK with these issues to ship a machine, and I keep waiting for a resolution as a warranty customer on $500,000 worth of the effected equipment and I'm not seeing light in the tunnel, and I need to order more equipment ideally from the same company and also have that equipment help us operate it with less hand holding, and everyone here is frustrated at this point. The known issues are:

1. No ability to import (put in old saved data)

2. No ability to export (save out proved data)

3. No ability to manually enter known or theoretical data (work around populate a single tool, or
re-load manually from pictures an entire known proven and unable to be exported table)

4. No ability to functionally operate "TEACH" function in dual path mode

5. No ability to recognize a zero to cancel monitoring of a single axis or spindle

6. No ability to teach a single tool that was skipped (or that skipped an individual desired axis population) in a teach process

7 (machine ladder problem) when the machine hangs on account of tool load monitoring, the spindle and coolant should shut off on that path only. (I feel like 1/2 the hangs are erroneous and probably also related to software issues).

8 (TOTALLY UNRELATED) The need of the machine to be given at odd times M264 and M265 to accomplish some mystery function involving the machine ladders gremlin issues.