ABB Direct torque control

Hi, does anyone here know anything about ABBs direct torque control drives. ie what is it. Regards Alan Case

Sounds that VFD drives solid torque and the speed dependent on the load ?

The reason I ask the above question is that a competitor has told a client that their brand of drives use DTC which makes thenm 10% more efficient. There was no mention of what the efficiency was compared to and the ABB site is very short on info on what DTC does.
I am beginning to think it is just another acronym for a system that is very hard to prove works or doesnt work. Regards Alan Case

I hope here is more information:
http://www.sescowi.com/epd/files/ABB/dtc.pdf

There are 2 kinds of VFDs, goodones and cheapones..

The direct torque control of ABB refers to the method which the processor of the drive calculates the commutations of the igbt. The DTC controller receives two data: the stator flux and the torque of the motor. With these two scalar data and the position of the flux of the rotor the command pulses will be calculated to the IGBT. If the motor is a brushless the position of the rotor flux is provided by the encoder. If the motor is an asynchronous the position will be calculated. The other field orientation methods control the torque in an indirect way

ABB affirms that this control improves dynamic performances at low speeds with respect to the other field orientation methods. I don't think that the DTC control changes the power efficiency.

After using many, I've never really seen any difference in actual performance between ABB's DTC, and any other similar class vector drive out there.

DTC does have a marked advantage over a straight V/Hz Frequency drive, but then again, they are nowhere near in the same class.

Acronyms are great for marketing hype

Alan, if I may venture into this thread a bit late--I've been gone on a three day family vacation.

First, you must know that I am an ABB rep, a certified startup technician, and a long time user of at least a dozen different brands of inverters over 15+ years. I may have some prejudices so be warned.

Direct Torque Control is ABB's name for an advanced form of sensorless vector motor control. The difference between DTC and other sensorless vector systems is primarily the precision that DTC acheives in building a motor model in its memory. This motor model is built during an initial motor ID run and is unique enough to be protected by several patents, I believe all Swiss in origin.

Once this motor model is built properly, the drive produces a torque feedback signal and speed feedback signal of remarkable accuracy from frequency and current data it collects from the motor leads using the motor model for its basis for all calculations. This method has a 6 microsecond update time and is so fast that the boards in the drives have to be linked by fibre optic strands to avoid timing errors.

I'd be the first to say that most applications do not need the maximum precision DTC is capable of so its advantages may be insignificant for the average job. On the other hand, I have faced really tough jobs where everyone else quoting is using closed loop vector (with motor shaft encoder) and I've quoted DTC without feedback and met the spec. I've done this repeatedly and several app engineers for other brands have grudgingly admitted that they simply cannot match DTC's performance with anything resembling sensorless vector.

Let me cite a couple of examples. First, a aircraft hydraulic pump test lab job where I had to meet the following performance spec:
Speed 0-10000rpm on special watercooled two-pole motor
Torque 0-1200in-lbs over the above speed range (including zero!)
Speed regulation not more than +/- 25rpm over whole load and
speed range
Do it motoring and braking in both directions

I quoted this job with an encoder but tried to meet the spec using just the DTC system. My best efforts resulted in +/- 8rpm error and the next twenty or so systems have been DTC only.

I have done several dozen stamping presses up to 1600 tons and have not required a regen package on any of them. The DTC requlator is so fast that the top of the stroke torque management leaves essentially zero excess speed coming down making regen much less demanding. DTC permits some flux braking internally without a braking resistor which helps too.

I've seen but never done hoisting systems that use DTC to stop and hold a load in midair using nothing but zero-speed motor torque to hold it. The system continues to hold the load stationary until motor heating renders the motor model invalid. At that point, a safety brake is necessary to keep the load from dropping. I've never seen another sensorless vector system do that successfully.

Now, all of that being said, if the person programming the drive doesn't know how to take advantage of this capability, lesser performance results. Actually, ABB makes three different levels of motor model precision available in their ACS600 and the newer ACS800 drives plus open loop scalar. So, you get to pick what level you want to go to.

Since the motor is a key component in acheiving tight speed regulation, the most responsible way of expressing drive speed regulation performance is in "%of motor slip". DTC is routinely able to regulate speed within 10% of motor slip from no load to full load.

Again, consider the source and hope it helps answer some questions.

Thanks Dick. I wondered what had happened to you lately. The motor is on a vertical lift water pump so speed accuracy required is minimal. The advertising for ABB quote "up to" 10% savings at 20% load falling to 2% savings at 50% load, but do not mention as against what style of VFD. Are the savings are quoted against a V/Hz drive, sensorless vector, encoded vector ?? or ???. Thanks for your time to answer. Alan Case

Alan, as usual, the problem with % anything is that it is meaningless unless the base number for the percent is given.

With the numbers you've given me, it isn't clear. I can say that I have seen responsibly demonstrated that properly set up DTC can produce slightly more motor shaft torque per amp than a pure sine wave can. I don't think this translates necessarily into higher efficiencies, tho.

On a vertical lift pump, the energy savings comes almost entirely from reducing the speed of the pump under light demand conditions. Pump horsepower is generally a function of speed cubed so you sure don't have to slow down much to save serious money. But, this savings can be acheived with the lowest tech scalar open loop drive out there. DTC is fine on these applications but doesn't translate into any efficiency or control benefits as far as I know.

In ABB's product line however, above 50hp, the only offering is the ACS600/800 which has DTC built in. It is, however, software selectable so, if you want to operate open loop, you can. That's Parameter 9904. I suppose that, above 50hp, the power circuit components dominate the cost of the drive so adding DTC in the software doesn't affect the system cost significantly.

So, Alan, the bottom line is that it sounds like the ABB salesman has said more than he likely should have said. No excuses, but, in today's market, with claims and counterclaims flying around wildly with little or no basis for proof, the temptation to survive is very strong. I've actually had to go off on my own parttime as a consultant to survive, too. But I try to stick with the facts as I know them and rely on testing or personal experience for the remaining questions.

Hope this finally answers your question.

Hi Dick, The savings/efficiencies I mention come straight from an ABB brochure page 21
http://www.sescowi.com/epd/files/ABB/dtc.pdf
link courtesy of Seppo
Regards Alan Case

I'd like to take a look at that document, Alan, but the link doesn't work for me. If you can give me a document number (on the back cover or page), I'll check it out. The document number might also be on the bottom of each page.

You might be looking at a document written in Australian. Naturally, I'd need one in American English to be able to read it. (snicker!)

Hi. I couldnt find a reference number and I cant copy part of the PDF but it is called the Technical Guide No 1 and it is on the ABB web site. It is quite well written and informative guide to VFDs. Regards Alan Case

Found them

Ref number 3AFE58056685

and link http://138.227.174.121/GLOBAL/SCOT/scot201.nsf/VerityDisplay/299C5D307CB78A34C1256D28003FED0C/

Thanks, Alan. I checked the reference and it is a true statement but shouldn't refer to DTC. The reference mentions flux optimization and that's where the energy savings come from. I've seen flux optimization on sensorless and scalar drives and it all works basically the same.

Flux optimization works by detecting a light load on a motor and, after a delay period, reduces the voltage below what would normally be expected at that frequency. This causes the magnetizing current to drop. That's where the energy savings comes from---the lower magnetizing current.

The problem with this scheme is that it totally wrecks any attempt at good speed regulation. That's why the booklet refers to fan and pump applications and more energy savings at light load than at heavy load. The other problem, if its a problem, is that the higher percent savings occur at very light loads so the actual savings is most often miniscule. It may be 10% at 20% load but its still 10% of almost nothing which calculates to, well, almost nothing. Now, if there was a way to save 10% at maximum load we'd all have something worth crowing about.

You can tell I'm no big fan of flux optimization. On the other hand, the statement is not false as I suggested it might be earlier. It is software selectable and I don't ever recall a job where I used it.

DickDV said:

I believe all Swiss in origin.

Click to expand...

I believe all Finnish in origin. I think they have designed and manuf. in Finland and some models in Milwaukee (New Berlin) for US market.

Originally posted by DickDV
I've seen but never done hoisting systems that use DTC to stop and hold a load in midair using nothing but zero-speed motor torque to hold it.

Click to expand...

I have used ACS600 for Roll Lift. Earlier Lifters ware hydraulic, but they do not like 'oil' in the paper mill floor, so El.-Motor is better for that. Starting from zero speed needs magnetism in motor before Brake can opened, otherwise lifter lowering a little in the start. I understand many more thinks now, thanks Dick ! I hope you understand my english.