Hi Hood,
you're right again, I didn't realize that!

I really got to get some of these suckers before they all disappear. I bought a house a while back and that has shot my CNC budget to hell!

Craig

Hi,
found this....seems to suggest that the DSD-030 may struggle to power servos in excess of 2.2kW, recalling the DSD-030 has
a rated continuous output current of 15A.

Have noted also that the high power MPL-A4********* and MPL-A5********* are very thin on the ground on EBay whereas the smaller MPL-A3********* are dime a dozen.

Craig

You better tell that to the Z axis on my lathe, it has been using an H4075 motor for 12 years, also the X axis uses an H or possibly and F 4050  which if I recall are over 2.5Kw.

MPL motors are different beasts from H and F series.

What you have to remember is the Ultra 3000's are old drives and they were designed for H and F series motors, MPL motors were not about at that time, the MPL motors are really meant for the newer 6000? drives.

Hi Hood,
I know the feeling, I have an MPL-A430H which according to this table requires 12A whereas the 2098-DSD-020 I actually have driving it is capable
of only 10A. In use truly how often do you come even close to maxing your servo and drive?

I find that if I require maximum torque (output current) its usually at very much less than maximum speed.  Given that the drive can be likened to
a buck regulator where the total out output current exceeds the DC Link current in the manner of a DC transformer that really I very seldom come
anywhere close to maxing the drive.

I can only conclude that the drives are generously specified and can accommodate a mild overload without blowing up. It comports well with my
perception of the quality and performance of the drive.

I some while back bought a refurbished 2.65kW servo, that is 14A at 180Vrms. It was manufactured in '95 and is fitted with an 8 pole resolver.
Trying to find a drive for it proved impossible, or least within my budget. I decided to make one, electronics being my thing and control engineering
being my training.

To get 14A rms per phase is going to require 28A rms input from the supply. I'll not get away with using such a device in the village I live, the power supply
degradation is likely to cause a complaint and thereafter the power company will track me down. I will have to therefore make a boost type PFC (powerfactor correction)
circuit. It will require a 10 second overload of 70A. You can be assured I've given the matter of input current distortion and output current, continuous and
overload, considerable thought in the process of designing my drive.

The 2098-DSD series will still being made in 2012 and thus were sold as compatible with the MPL series servos which started manufacture in the early 2000's.

Craig

Yes the drives were still available and the firmware updated through time to encompass newer motors but originally they were designed for the motors available at the time, in fact they are basically just upgrades of the Ultra 200 drives which were upgrades of the ultra 100 drives which in turn were upgrades of DM drives. I have a few DM drives here although Osai branded rather than AB, then again the original makers were Electrocraft and I think AB eventually bought them out. That was probably why Giddings and Lewis stopped selling the Centurian drives as they were just rebranded ElectroCraft DDM and DSD drives. One of the 030 drives on my lathe is actually a G&L Centurion.

BTW the DSD drives are very adaptable beasts, I have a few motors running on them that were originally resolver motors, all I had to do was find a suitable encoder for them and then set up a custom motor profile.
My manual lathe uses such a beast, the motor is a 1326AB-A520E  and the encoder is from an N series motor (same pole count) so it was ideal. The 1326 motors were meant for 1391 drives. The drive on the manual lathe however is an older DDM-075.  They are similar drives but you can only have one custom motor in the database at a time, you have to overwrite it if you want a different one, the DSD drives allow as many custom motors as you like.

You do not even have to have commutation signals on the encoders as they will self sense on start but the drawback is when you power up the drives the motor will rotate slowly approx 1/4 turn to sense the commutation so it is not the ideal situation in a lot of setups. It would have been fine for my manual lathe but if I recall the DDM drives do not have self sense.

Hi Hood,
kool, fascinating bit of history. It shouldn't surprise me that they were developed over a period of years. I mean it would be a tall
order for a manufacturer to come out with a new design that good right from the get go.

The history goes that I was looking for a servo for a mill spindle motor where I wanted torque at low speed for steel and stainless.
I got this 2.65kW unit I've talked about and thereafter decided to build a drive for it. Of course its taken ages and then I saw the Allen
Bradley 1.8kW servo and drive turn up on the NZ auction site and I bought it.

This is the first modern AC servo and drive I've ever had any experience with and simply it blew my mind as to how good and flexilble it is.
As a result of the experience I reallise now that I could use a DSD-030 drive to power my original purchase as you have recommended.
Really the pressure is off trying to complete my home built drive, I already have the Allen Bradley servo converted into a spindle and doing
sterling work for me.

I wish to persevere with making my own drive. I have had to learn to program a Texas Instruments TMS320F2069M micro controller
in C and C++ and complete with all the hardware peripherals it contains, the latest and greatest PFC devices, isolated gate drives and the
list goes on. I have always thought that the success of a hobby should be measured by what you have to learn in its pursuit. If that is indeed
the measure then making my own drive is very worthwhile.

Just exactly what a 2098-DSD-030 can or cannot manage I am at least sure it will power a servo that if Mick chooses to use as a lathe spindle he
will be very happy with. If he wants more grunt either get three phase power or go for some sort of extreme single phase solution, likely to cost
more than getting a three phase connection to his shop.

Craig

If it would be big enough (power wise) then I think a DSD-030 with an F4075 motor would be the way to go for a lathe spindle as the high inertia is ideal.

It may be possible to use the Indexing features of the 030X ( or SE) drives to act as a C Axis as well, just depends on whether he just wants indexed positions or needs to have the axis rotating. If the former then you can set up 4 inputs to the Indexing drives to give you up to 64 predefined positions, it is what I do on the turret of my lathe.
Now for that to work it would have to be possible to disable the spindle in Mach3 and for the CSMIO to be happy with that, not sure if it can be done.

Hi Craig Hood

You pair are unreal. Wish I had 1% of your knowledge. Need to have a good read through all the posts and look at the motors and drives your quoting.

Cheers
Mick

 

Hi Hood,

If the former then you can set up 4 inputs to the Indexing drives to give you up to 64 predefined positions, it is what I do on the turret of my lathe.
Now for that to work it would have to be possible to disable the spindle in Mach3 and for the CSMIO to be happy with that, not sure if it can be done.

There are some interesting questions bound up in that. Micks CSMIO is an analogue controller, to use the indexing feature would require that the CSMIO
stand out of the way and allow the drive to close the loop and move to the required index position. Quite frankly I don't know.

I have an ESS which is a step/direction controller and can be forced to stand aside while the drive completes some autonomous function.

Additionally I now use Mach4 which allows me much easier means of programming indexing operations verses free running spindle operations.

I would have to recommend just getting the spindle going in free running mode, which is still the most commonly used mode, would be adequate
to start with. THEN try puzzle out some way to introduce index ops.

Craig