Pressure sensing: a novel method for characterizing the processing zone in vapour phase soldering systems
A. Geczy, Zs. Peter, B. Illes, Zs. Illyefalvi-Vitez
Department of Electronics Technology, Budapest University of Economics and Technology, Budapest, Hungary Goldmann Gy. ter 3., H-1111
gattila@ett.bme.hu
Abstract:Vapour Phase Soldering (VPS) [1] has become a significant alternative method for reflow soldering in the last years [2-3]. Since the technology is based on heat transfer principles, which are totally different from the conventional reflow technologies (forced convection or IR), a new approach is needed for the characterization of the processing zone inside the vapour chamber. This paper presents the results of a novel method for characterization: the measurement of pressure differences in the processing zone of a VPS tank. For this purpose, high resolution pressure sensors were applied to a batch type, experimental VPS station. High resolution of the measurement range is a key factor of the method due to the small pressure differences of the process. The measurements correlate to the Galden vapour density and the temperature, which are also important aspects of the experiment.
1. Introduction
After the comeback [2,3] of vapour phase soldering, the method became a competitive alternative solution for reflow soldering in the field of surface mounted assembly technologies. The literature of the topic, however, is still not extensive enough to cover all aspects of the process.
During the VPS process, an inert, non-harmful liquid (Galden) is boiled at the bottom of a tank. With proper heating parameters it forms a layer of saturated vapour over the boiling liquid. The assembly (which is already prepared for soldering by paste deposition and component placement) is then immersed into the saturated vapour layer. The saturated vapour condenses on the colder surface of the assembled board, providing its latent heat of condensation to melt the solder alloy (to reflow the solder paste).
The saturated vapour layer is a key to the process.
The control of heat transfer depends on the control of the vapour. Our aim was to investigate the pressure relations inside the soldering process zone of an experimental batch type VPS station (Figure 1.) and to characterize the heating process by correlating the pressure differences and the change of Galden vapour density (the formation of the saturated vapour blanket) with the temperatures measured inside the tank.
. Fig. 1. Batch type experimental VPS station.
2. Experimental
The setup for the measurement consists of the batch type experimental station, which is a flexible system prepared for positioning different sensor types at different points inside the process zone.
Thermocouples and platinum resistor temperature sensors are used for temperature measurements, and different pressure sensors (with silicone hose probes positioned in the vapour) are used for pressure measurements. Data is logged with a universal data acquisition tool connected to a PC.
For the measurements the same initial parameters were set. The following parameters were applied:
Table 1. Parameters of the process
Parameter Value
Galden Type HT170
Galden Boiling Point, °C 170
Galden Volume, dm3 1,3
Heating Power, W ~550
Cooling Power, W ~50
Galden HT170 type is used for faster cycles and more efficient resetting of the system for new measurements.
2.1. Pressure measurement devices
For pressure measurement three different types of sensor devices were used. Each of them is coupled to the experimental process zone via silicone hoses.
These hoses can be considered as the probes for the sensor. The devices work by different measurement principles; it was also an aim of the investigations to highlight the applicability of these sensors in the field of VPS.
Fig. 2. Sensirion SDP1108 (left) and Sensortechnics LBAS500U (right)
The first type (Figure 2, left) is a Sensirion device, SDP1108, which works as a differential sensor by dynamic principle. The device offers square root output characteristics (to enhance resolution and accuracy at very low differential pressures) and provides a fully calibrated voltage output. The root- square output characteristics allows 0,05 Pa resolution at the lowest ranges. The full range is 0-100 Pa, which is suitable for the given vapour pressures.
The next sensor applied in the experiment is Sensortechnics LBA-type (Figure 2, right) differential dynamic pressure sensor, which is capable of measuring ultra-low gas pressures also offering analog signal conditioning, temperature compensation and
output amplification. LBAS500U type was used, which has 0-500 Pa range with a linear output, allowing sub-Pa resolution for measuring fine differences.
The SDP and LBA type sensors are based on thermal principle, where a heater generates a specific temperature profile on a MEMS membrane. The applied pressure difference enforces an air flow on the small channel showed on Figure 3. This air flow breaks the symmetry of the temperature profile which is measured by the two temperature sensors.
Temperature difference is a direct function of the applied air flow (pressure difference). Figure 3 presents the basics of the CMOSensTM Sensirion device; however, the Sensortechnics sensor element has a similar construction.
Fig. 3. MEMS based pressure sensing principle [4]
Due to the dynamic measurement principle, a little air flow is required for the sensor, which leads to dependence on the length of the hose. Tubes up to 1 m have an error of less than 1%. This error margin was compensated according to manufacturer application notes.
The third device has a different principle. The Fluke 922 differential manometer (Figure 4.) is able to work by dynamic principle; however, it can measure static pressure as well.
Fig. 4. Fluke 922 [5], with hose positioned in the VPS tank The Fluke sensor has ±4000 Pa range and 1 Pa resolution with ±1 Pa tolerance. It is not able to
highlight fine changes in pressure characteristics;
however, it is adequate for showing the actual pressure values of the saturated vapour. The static mode of the device was evaluated with basic water column pressure measurements.
For measurements, a single hose is connected to the input (+) ports of each device. The input hose is then positioned into the given height inside the VPS process zone vertically. The reference (-) ports are left unconnected, open to the ambient space. The outputs are giving the differential pressure of the input with respect to the reference ambient space. A positive reading means that there is higher pressure at the end of the input hose-probe. The diameters of the hoses are matched for the given sensors; the length of each hose is 1 m. The moving vapour blanket forces the air inside the hoses to flow; this way the vapour does not interact with the sensors directly, only air flows through the devices. Also, with a 1 m probe length, heat transfer via the hoses and temperature increase at the sensors can be neglected. Due to vertical positioning inside the VPS, condensing vapour pours back into the bottom of the tank. The Sensirion and Sensortechnics devices are also coupled to the VPS process zone according to Figure 3.
2.2. Temperature measurement device
For in-situ temperature measurements during pressure measurements, Chromel-Alumel (K-Type) thermocouples are attached to the end of the silicone hoses. The thermocouples have an error margin of +/- 1 °C, which is acceptable for the measurement purposes. The small welded hot junction has negligible thermal capacitance, allowing fast output response.
Fig. 5. K-Type thermocouple with small welded hot junction
The full process zone characterization was performed beforehand [6] with a batch of sensors,
comprising ten Pt500 platinum resistor thermometers positioned in points along the vertical Z axis.
3. Results
Figure 6. shows a pressure measurement of a full heating up process with fix parameters (Table 1.), where the end of the hoses are positioned 1,5 mm above the top of the heated Galden. The measurement shows that the Sensirion and Sensortechnics sensors indicate the relative change of the pressure, not the exact pressure values at the given height in the tank.
On the other hand, the interpolated Fluke characteristic show the actual pressure values. The slow relaxation after the maximum can be described with the extremely low pressure level in the range of the sensor. It is suggested that the absolute maximum of +18 Pa (0.5% of the measuring range) is insufficient for the sensor to hold its output signal.
According to the dynamic sensors, there is no change in the pressure values beyond the maximum points, so the measured maximum output of the Fluke can be considered as the absolute maximum pressure.
Fig. 6. In-situ pressure measurement with three sensors; the hoses are positioned 1,5 mm above Galden level The values measured by the Fluke manometer are in correlation with preliminary calculations, where a 17 cm tall column of Galden vapour at 170 °C should have ~20 Pa pressure at the bottom.
With this method (involving the Fluke manometer) the pressure of the saturated vapour can be measured at given vertical positions inside the process zone.
This method points to a more specific vapour processing zone monitoring, where not the temperature, but the exact vapour pressure (and thus the density) is measured. From the aspect of soldering,
the saturated vapour indication is more precise and immediate than measurements with thermocouples.
To correlate temperature values with pressure values, the temperature of the vapour (at the level where the hoses were positioned) was also measured by thermocouple during a full heating up with fix parameters (Table 1.). The following Figure (Fig. 7) reveals three stages of the heating process.
Fig.7. In-situ pressure and temperature measurement with two sensors; the hoses are positioned 1,5 mm above Galden level; Three stages of the process are highlighted (I.-II.-III.) Note: the dynamic pressure sensors are indicating only the relative change of the pressure. The first stage (I.) of the heating process shows a linear increase in temperature. This is followed by the first linear increase in vapour pressure (the increase in vapour density as well according to the general gas law). At the end of the first stage, temperature is nearing the saturation value, but it has still not reached the steady state. The ΔT/t rate for the second stage (II.) is considerably lower than before, and the vapour formation is slowing as well, according to the pressure measurements. Then, at the end of the second stage, there is a rapid pressure increase, which means that the saturated vapour has just reached and surpassed the level of the probes. The third stage (III.) is the steady-state at the given height inside the process zone, where the vapour density does not increase further, and the temperature is fixed at the saturation value (the boiling point of the liquid).
4. Conclusion
In this paper a novel method was presented, where pressure values were measured inside a process zone
of a vapour phase soldering system with three different sensor types.The pressure values were in correlation with the temperature of the vapour space at a specific point of the process zone. Three stages of the temperature profile and of vapour density can be distinguished. The soldering is recommended only at the third, final stage (III.), where vapour density (which affects the condensation and the heating up of the solderable assemblies) already reached its saturation. With sole temperature measurements, the optimal circumstances for soldering could be indicated in the second stage (II), but it is not optimal from the aspects of soldering and heat transfer. This approach may help to improve the overall process and current constructions of VPS stations working only with temperature sensors.
Future work will include investigations at different heights of the soldering station, to prepare full saturated vapour identification along the vertical coordinate of the processing zone.
5. Acknowledgement
The work reported in the paper has been developed in the framework of the project ”Talent care and cultivation in the scientific workshops of BME"
project. This project is supported by the grant TÁMOP - 4.2.2.B-10/1--2010-0009.
References
[1] R. C. Pfahl, H. H. Ammann, Method for Soldering, Fusing or Brazing, 1975, US Patent 3,866,307;
[2] H. Leicht, A. Thumm, Today’s Vapor Phase Soldering - An Optimized Reflow Technology for Lead Free Soldering, in Proceedings of Surface Mount Technology Association Int. Conference at Orlando, USA, 2008.
[3] A. Thumm, Going Lead Free with Vapor Phase Soldering, in: Proceedings of Surface Mount Technology Association Int. Conference at Orlando, USA, 2010.
[4] Measuring Differential Pressure and Air Volume with Sensirion’s CMOSens® Technology, Application Note, www.sensirion.com, (as is) 2012.04.07.
[5] Fluke 922, Users Manual, Rev. 01., www.fluke.com, (as is) 2012.04.07.
[6] A. Géczy, Zs. Illyefalvi-Vitéz, P. Szőke, Investigations on Vapor Phase Soldering Process in an Experimental Soldering Station, Micro and Nanosys. (2) (2010) pp.
170-177.
