TGA gives the weight changes as a function of time or temperature, however DTA, is obtained from  differentiating of this changes (a kind of the rate of weight changes). If the weight changes are very slow, may be you can not distinguish in DTA, these three steps observed in TGA. 

Did you differentiate the DTA curve (second derivative of the TGA curve)?

@Atilla..thank you for the answer.

DTA as far as I know is the plot of temperature difference between the sample and the reference versus time. DTG is what you are talking about which is the differentiation of the weight with time.                                   

But I am talking about the DTA curve here. In DTA I am not getting peaks (As in endothermic or exothermic peaks) but I am getting many steps of weight losses in my TGA curve. I am wondering what the reason could be.

You are right. DTA is the plot of the temperature diffrence between sample and a reference versus time.  What is the reference did you use?  

I am agree with Atilla

TGA and DTA has different function, and complimentary each other.

TGA measures weight change when the sample is  heated, while DTA (or DSC) measures temperature different between sample and reference when both are heated, and gives information such as glass transition temperature, melting temperature, crystallisation temperature,etc. TGA and DTA are used to be in the same instrument called TGA/DTA.

Actually if you are not getting any peak in DTA ...As per my knowledge you can do one thing just check the melting point and/or glass transition point of your compound.Then just make some changes in the methololgy, change the ramp  and initial temperature of the method.start the method from low temperature with low ramp......you may get some positive result...

What is the nature of your sample? It might be that your TGA steps are related to adsorption or desorption processes that are ina a broad range of temperatures so are not visible in your DTA trace. We have observed diferences in DTA information when compared to DSC, for example. Atilas' question regarding your reference sample is very relevant.

If you find a constant  (either positive or negative) slope on the TG curve, his could be a baseline shift, which is caused by the air flow during heating, due to the  so-called chimney effect or stack effect. 

In combined DTA/TGA  instruments the DTA signal is correlated with TGA signal. With other words any changes in TGA signal should lead to changes in DTA signal  which is proportional of the weight change and magnitude of related thermal effect of that change. TGA signal itself should be independent of DTA signal e.g.. melting, crystallization, glass transition.   Using higher speed of measurement or weight of sample will lead to higher   .sensitivity, lower speed to higher resolution. So try to measure lower  speed.  It could be helpful to  try also to stop the measurement after every step in weight, cool dawn and heat again.