There are three main tools for hypersonic flight vehicle design, namely flight testing, CFD and ground-based testing. The flight testing cost and risk high. For reducing this risk, lots of work should been done with the CFD and ground-based testing. The impulse wind tunnel is one kind of ground-based facility. It can simulate hypervelocity flow but the test time can only last for several milliseconds. Force measurement in impulse wind tunnel is more difficult than that in continuous wind tunnel. When the impulse tunnel runs, the output of the balance includes the aerodynamic force and the inertia vibration. The vibration caused by the mass of model, balance and sting (MBS) system. Usually the frequency of the MBS is same order of magnitude with the test time, so if we want to get the pure aerodynamic force, we need to improve the frequency of the MBS. The higher the frequency is, the more test curve period we get within test time, the higher the accuracy is. For improving the frequency of the MBS, it is necessary to make the balance and the sting more rigidity, and reduce the weight of the model. For the kind of blunt body model, such as Mars lander configuration, the frontal area is several times larger than common model, so the axial force is very large during the test, even far more than the range we usually measured. Aiming at this issue, considering the Mars lander configuration is symmetrical, the mx component for improving the frequency of the MBS is removed. A five-component balance has been designed by the method of finite element simulation, considering both natural frequency and sensitivity requirements. It shows that the calibrated precision meets the military standard indicators, the accuracy of each component is within 5%. Using the balance to measure the force of the Mars lander configuration model in FD-20a shock tunnel of China Academy of Aerospace Aerodynamics, the Mach number is 6. The model chosen to test is a 70-deg-semiangle cone of 261mm axial length, and the model material is made of LC4. The test gas are air and carbon dioxide respectively. The test results show that the repetition is well. The axial force coefficient with carbon dioxide is higher than with air, but the increment decreases when the angle of attack a changes from 0 degrees to - 20 degrees.The degree of trim angle decreases with the carbon dioxide test gas. Compares to the CFD, the axial force coefficient is recovered to within 3% and 1% in the air and carbon dioxide test gas respectively.