What's new
What's new

Is a Brother Speedio the right machine for these Tolerances?

Trevor360

Aluminum
Joined
Jul 25, 2021
I currently am running a lot of small 3 axis work on a Hermle 5-axis and it is time to purchase a 3-axis mill to offload this work onto a cheaper machine. Our work is High mix and medium to low volume (20-1000 pieces runs). I REALLY like the idea of a Brother Speedio, I like how simple they are to work on and how well supported they are, as well as the small form factor. They seem like an excellent platform to build a shop around and everyone on the threads here seems to love them (have read a lot of these threads).

Rigidity will be fine for our application making smaller parts. However, my concern is with the quality of the parts:

Our parts regularly require: Interpolating small holes, slots, and bosses to +-0.0002, 0.001 true position on small parts (between the size of a bic lighter and a house brick), Z drops of +-0.0005 or less between multiple different tools or setups. Mostly stainless and some aluminum. We also want to maintain the nice surface finishes and general aesthetic of the parts we get off the hermle and don’t want to spend all day dialing in offsets. For jobs like this on the Hermle there are almost 0 messing with offsets, we dial it in once and it just works, never have to worry about the machine warming up and changing size etc..

What will our experience with the Brother be like? Will it do work with these requirements all day long or will it be a struggle? Alternatives would be going with something like a Okuma Genos with its thermally stable modified gantry design.
 
A Speedio shouldn't have any problems hitting those tolerances, so long as you have decent temperature control in the shop and the right strategy.

Speedios are so fast in tool changes and linking moves that the best strategy is single piece flow, i.e. run each part to completion before moving on to the next. This is irrespective of how many parts you have loaded on the table. The idea here is to isolate thermal drift on the machine from the work (mainly in Z). For example, if the machine grows .0015" in Z over a given period of time, your Z tolerance is blown if you're machining 100 parts at the same time, with one feature machined at 8am and another feature at 11am. But if you bang out a complete part every few minutes, the relative Z values from one feature to the next on the same part are much less likely to change much.

As for surface finishes, Speedios handle aluminum like any other high end machine. With stainless, you can get great finishes on endmills 1/2" and smaller, but don't expect facemill finishes to be nearly as good. In fact, avoid hogging with a facemill altogether in anything other than aluminum. Your Hermle has a polymer concrete base that weighs more than an entire Speedio.

Lastly, consider a multiaxis U500 over an S-series 3-axis. If you're going to offload work from an expensive high-end 5-axis, it could greatly benefit you to mirror most of its capabilities with at least 3+2 positioning. If you have parts that can't be offloaded completely, that doesn't mean you can't partially offload them. Run the parts halfway on the Speedio and then move them onto the Hermle, like a transfer line. It'll just require that you have two identical ZPS bases in both machines.
 
I got an R650 last May and make smaller aluminum parts, nothing too tricky but I often use a mic to measure my parts and see if I can hold +/-.0001" since I am still learning this machine. While I don't see any dimensional differences on 1" measurements I do see drift in my12"-15" long aluminum fixtures as temps rise some during the day, mainly coolant. I haven't seen any Z axis shifts between tools but my longest cycle so far has been around 10 minutes on 4 parts. Any issues I have seen have only been a couple of tenths at most and due to temps changing or I think less than perfect foundation, the machine is tight. In short, once set I don't see any drift, other than temps I should be controlling better, and I adjust both tool and fixture offsets .0001" or .0002" at a time when I do make adjustments.

While fast, chip-to-chip tool changes are not free, so I like to make a few parts at a time and don't see any issues doing it. Dividing say 24 seconds of tool and part changes over 4 or 8 parts is an easy way to significantly reduce cycle times.

If going Brother I really think going rotating table is at least a level higher, maybe more, at increasing production. Having a consistent part swap time of around 2 seconds, if you subtract the tool change time you normally make anyway when changing parts, is a big deal to how long it takes to make parts. Plus I don't have to be at the machine to swap parts when it is ready, I can do it when I want, which makes the machine much more productive. These are only some of the benefits of a rotating table, there are many, many more.

The only thing I have found that chokes it is holding .0001"-.0002" on profiles with many features, like radiuses and short lines. The accuracy settings to hold those tolerances on those features really throttle the feeds back.
 
A Speedio shouldn't have any problems hitting those tolerances, so long as you have decent temperature control in the shop and the right strategy.

Speedios are so fast in tool changes and linking moves that the best strategy is single piece flow, i.e. run each part to completion before moving on to the next. This is irrespective of how many parts you have loaded on the table. The idea here is to isolate thermal drift on the machine from the work (mainly in Z). For example, if the machine grows .0015" in Z over a given period of time, your Z tolerance is blown if you're machining 100 parts at the same time, with one feature machined at 8am and another feature at 11am. But if you bang out a complete part every few minutes, the relative Z values from one feature to the next on the same part are much less likely to change much.

As for surface finishes, Speedios handle aluminum like any other high end machine. With stainless, you can get great finishes on endmills 1/2" and smaller, but don't expect facemill finishes to be nearly as good. In fact, avoid hogging with a facemill altogether in anything other than aluminum. Your Hermle has a polymer concrete base that weighs more than an entire Speedio.

Lastly, consider a multiaxis U500 over an S-series 3-axis. If you're going to offload work from an expensive high-end 5-axis, it could greatly benefit you to mirror most of its capabilities with at least 3+2 positioning. If you have parts that can't be offloaded completely, that doesn't mean you can't partially offload them. Run the parts halfway on the Speedio and then move them onto the Hermle, like a transfer line. It'll just require that you have two identical ZPS bases in both machines.
Hi Eric, thanks for all the great info here! Also, love your products.

I am actually considering either an R pallet-changing machine or an S model. Would start it out as a 3 axis but could see adding a 4th and tombstone.... or..... 5th and robot/spindle gripper to the machine if the right work came in the future.

Obviously, the R450 and R650 are much more productive for volume work. But how livable do you find setups on them? Am I going to hate myself for using an R machine for a run of 50 pieces where I am proofing out the program on the first part?

The U500 looks very cool but not a good fit for what we are trying to do with this machine right now.
 
I got an R650 last May and make smaller aluminum parts, nothing too tricky but I often use a mic to measure my parts and see if I can hold +/-.0001" since I am still learning this machine. While I don't see any dimensional differences on 1" measurements I do see drift in my12"-15" long aluminum fixtures as temps rise some during the day, mainly coolant. I haven't seen any Z axis shifts between tools but my longest cycle so far has been around 10 minutes on 4 parts. Any issues I have seen have only been a couple of tenths at most and due to temps changing or I think less than perfect foundation, the machine is tight. In short, once set I don't see any drift, other than temps I should be controlling better, and I adjust both tool and fixture offsets .0001" or .0002" at a time when I do make adjustments.

While fast, chip-to-chip tool changes are not free, so I like to make a few parts at a time and don't see any issues doing it. Dividing say 24 seconds of tool and part changes over 4 or 8 parts is an easy way to significantly reduce cycle times.

If going Brother I really think going rotating table is at least a level higher, maybe more, at increasing production. Having a consistent part swap time of around 2 seconds, if you subtract the tool change time you normally make anyway when changing parts, is a big deal to how long it takes to make parts. Plus I don't have to be at the machine to swap parts when it is ready, I can do it when I want, which makes the machine much more productive. These are only some of the benefits of a rotating table, there are many, many more.

The only thing I have found that chokes it is holding .0001"-.0002" on profiles with many features, like radiuses and short lines. The accuracy settings to hold those tolerances on those features really throttle the feeds back.
Thanks, DavidScott, super helpful info. I may have seen some videos you have on Youtube (if they are yours?). Honestly, some of the better content on the walkaround and running of the R650 that's available right now! How awkward or easy it is to set up the R machine ergonomics-wise. Sounds like you are doing much higher volumes than us but trying to get a sense for how difficult the setups and prooving out programs will be.
 
I do a lot of small parts using speedios and I am able to hold .0002" on slots and bores consistently. Z height will change with temperature about .002" throughout the day for me and I just keep an eye on it when I'm doing second ops. If the run time of the part is short enough, it shouldn't make a difference within the part.
 
Hi Eric, thanks for all the great info here! Also, love your products.

I am actually considering either an R pallet-changing machine or an S model. Would start it out as a 3 axis but could see adding a 4th and tombstone.... or..... 5th and robot/spindle gripper to the machine if the right work came in the future.

Obviously, the R450 and R650 are much more productive for volume work. But how livable do you find setups on them? Am I going to hate myself for using an R machine for a run of 50 pieces where I am proofing out the program on the first part?

The U500 looks very cool but not a good fit for what we are trying to do with this machine right now.
The R650 is great for running small quantities. Visibility is a nonissue. I'm used to running horizontals, so the layout of the door/window/control was very natural for me.

The R450 is a different animal. Not horrible, but it is weird having to basically peek over the non-working pallet to see what you're cutting.

If you think you might add a 5th at a later date, I would rule out the R machines and go for an S700 or W1000. The R650 has limited headroom when pallet changing, and a good sized tilting rotary will not fit.
 
Thanks, DavidScott, super helpful info. I may have seen some videos you have on Youtube (if they are yours?). Honestly, some of the better content on the walkaround and running of the R650 that's available right now! How awkward or easy it is to set up the R machine ergonomics-wise. Sounds like you are doing much higher volumes than us but trying to get a sense for how difficult the setups and prooving out programs will be.
Yes, CGPM. I don't find my Brother to be awkward or difficult to program, setup, or prove programs on other than it is a different control and my table only has threaded holes, no t-slots. For reference, I am used to Fanuc and Yasnac. When talking about Brothers you must understand they are 24/7 PRODUCTION machines, PERIOD! Yes, you can do onesy twosy things on them but for that stuff I would rather use my Kitamura. But as soon as I am doing 5 or more parts I am loving that rotating table, and everything else about the Brother that is so focused on production. I think the rotating table alone is good for a 20-30% increase in parts a day, and I had my old system down to swapping pallets of parts with 15-20 seconds of spindle down time. Everything that irritates me about my Brother is because it is focused solely on production, which is why I bought it and not something else. There are no compromises to make it general purpose.

F35Machinist, you are seeing .002" of Z axis movement? I can't say I have seen any noticeable movement after warming it up, and .002" would be very noticeable. Perhaps your thermal comp needs adjusting.

Orange Vise, you can get a lowered table version for more room.
 
Orange Vise, you can get a lowered table version for more room.
They also have different "roomier" sheet metal available for the partition if you plan on using a rotary on the R series machines.

Look up the Brother T200A rotary table brochure. It has all the specs for the different sheet metal and table height options.
 
I find that my S700X2 changes Z height by 0.02-0.03mm at some point in the first 30 minutos of warm-up. After that I don't see it move the rest of the day (despite reasonably significantly changing temperatures)

I speculate that the thermal temp compensation only kicks in after the axis has moved XX meters (not sure what that number would be?!) ie I don't think it's time based??

So my warm up is to whirl the table and Z around for a bit (with the spindle on low speed)

I also plan to copy what I saw someone else here do and create a fixture of known height with a calibration ring stuck on and integrate that with the tool probe fixture. Then you can have your warm up routine recalibrate everything as well... (I mean you could write a macro to reprobe all 21 tools in not many seconds! It's that fast!)
 








 
Back
Top