I would choose between NETZSCH Jupiter series, TA Instruments Q600 and Setaram LABSYS. The difference in performance is not very large. I used mainly TA Instruments devices and I like their software, but I was always afraid that the sample holder of Q600 looks too easy to break.

I have continuously used TA instrument for a year with no problem. I suggest ordering the accessories like the cooling system for running the instrument at low temperatures, as well.

I use STA 8000 Perkin Elmer unstrument. I can use both TGA and DSC.

STA8000 connect with water chiller so the temperature can go from 20 Cto 1500 C

I think it may be suitable for you research

I would like to thank everyone for their time and answers, they proved very useful.

I have worked extensively with both TA Instruments devices and with SETARAM instruments devices, so that informs my response: 

TA Instruments machines are cheaper in up-front cost, but much more expensive to operate in terms of consumables and repairs. Many TA maintenance and repair operations are not easily performed by the user as they require specialized tools or an extremely steady hand which only comes from practice - so you will need a service contract with regular service engineer visits. The hang-down TGAs are extremely delicate and not something I would allow an undergrad or even an early graduate student unfettered access to unless you want your repair costs to balloon - nor would I want multiple users in the equation. They are one-user instruments for best performance - because that way the instrument technician has ownership over taking good care of the instrument. The DSCs are a bit prone to calibration drift and thus relatively high-maintenance but otherwise fairly easy to use. The instruments do produce good data and are generally quite reliable - but when they do break down, expect at least five figures in the repair bill and a service visit. TA instruments machines tend to come with a lot of bells-and-whistles options which affect the look and image of the instrument without significantly affecting the performance like touch-screen interfaces etc. The consumables are in general between 2-5x more expensive than Setaram consumables (being the other brand I'm familiar with). I would say TA instruments devices are the choice if you care more about it looking good and being user-friendly on a daily use basis than about high performance, extreme environments, ease-of-maintenance and operating costs. They make good workhorse instruments, if a bit on the expensive side to run.

I have not worked with NETZSCH instruments before - the general reputation is that they are in between TA and SETARAM in terms of up-front cost, and on par with TA in terms of recurring cost, more versatile than TA but less than SETARAM, and more bells and whistles than SETARAM but less than TA. I have heard that some NETZCH systems can be delicate and prone to breaking easily, but I've also heard that they're otherwise well-engineered. I don't have personal experience with NETZSCH instruments so I can't give an informed opinion about what the deciding factor is there or what the user experience is.

SETARAM instruments in general are a bit higher in up-front investment required, are extremely versatile in terms of atmospheres and samples that can be accommodated, and are both inexpensive and easy to maintain. SETARAM also offers high-sensitivity Tian-Calvet calorimetry in some of its instruments, and is able to accommodate more extreme environments than other manufacturers including offering the only commercial instrument to my knowledge which can measure under pure hydrogen atmosphere up to 1750°C (unsure if your rare earth work is using hydrogen storage alloys - but it might be relevant), as well as having the capability to accommodate other corrosive gasses such as HF and Cl2, and high pressures up to 10K bar (depending on the instrument - their Setsys DSC is the one with the corrosive gas and H2 capability. Labsys is more of a typical workhorse instrument and is a solid choice for routine lab work - it does have a high sensitivity option, but the high sensitivity is not as good as a Sensys or Alexsys would be. Sensys or Alexsys would be the instruments of choice for high pressure DSC at high temperatures). Most routine maintenance and repair is user-doable on a SETARAM instrument, and consumables are the cheapest of the big manufacturers, making a low year-on-year operating cost which to my mind balances out the higher up-front cost. I would not recommend opting for a service contract for a SETARAM instrument as they tend to be low-maintenance so you genuinely don't need it.  As a company, they focus more on the high performance and the ability to handle extreme environments (pure hydrogen, corrosive gasses, radioactive environments, extremely high temperatures or pressures, etc) than on flashy looks or "bells and whistles" type of options. They require a bit more in terms of up-front training only because SETARAM trains users in routine maintenance and replacement of consumable parts - and once you're done training, they're very reliable and straightforward to operate in my experience. SETARAM is the instrument of choice if you want top-end performance, robust construction, and great ease of maintenance, or if you want to work with extreme environments.

Oh, an afterthought on this - if you're looking to change atmospheric conditions fairly regularly, you may want to consider a Tian-Calvet (3D) calorimeter. Tian-Calvet calorimeters use a 3D DSC sensor which completely surrounds the crucible. Because of this, they are able to capture the majority of the convection and radiation heat transfer as well as capture heat transfer by conduction (somewhere between 90%-95% of the total heat flow signal, depending on the design of the calorimeter, compared with a 2D DSC which can only detect heat transfer by conduction and thus sees somewhere between 20-30% of the heat flow in the system). An added benefit of the 3D sensor is that the calibration is then independent of atmosphere or gas pressure, as the degree of heat "missed" by the sensor is fairly independent of the convection in the system, unlike in a 2D system.

This would have the effect of eliminating the need to recalibrate the instrument every time you want to change the gas flow rate, pressure, or composition of the gas in the system. As well - Tian-Calvet systems can be calibrated in an absolute method using a Joule effect device to generate a known heat flow, as opposed to calibrating relative to a number of standards. This has the effect of greatly reducing calibration error in the system (because as I'm sure we all know, the quality of your DSC data is directly proportional to the quality of your DSC calibration).

It may not be worth the added investment unless you're looking to switch up the atmosphere regularly (I know some researchers into catalysts which find this aspect of the Calvet method very useful as they don't have to recalibrate for each pressure and gas composition profile), but for those who need to switch gas composition regularly, it is hugely helpful. 

DSC Vesta from rigaku is good choice

I suggest TA instrument.