download

Report 

Evaluating the CAF (conductive anodic filament) 

resistance of multi-layered PWBs 

Understanding the Technology 

New JTM standards for temperature test chambers 

– Methods of testing and indicating performance 

Topic 

New model: Super Clean Oven SCO-2 

1 

18 

29

Evaluating the CAF (conductive anodic filament) resistance of multi-layered PWBs 

Report 


Akiko Kobayashi, Yuichi Aoki, Keiko Toi Technical Development Headquarters 

A 

dvances in the high-density wiring of electronic parts have been accompanied 

by insulation degradation stemming from ion migration, creating serious 

problems. While a great deal of research has been carried out investigating the 

formation of dendrites, which are one form of ion migration, there has been a 

noticeable lack of clarification of the role of CAF (conductive anodic filament) 

formation. As a result, we decided to evaluate the CAF resistance of multi-layered 

PWBs (printed wiring boards). We investigated the relationship between the 

shape of the CAF that formed and the behavior of insulation resistance levels 

during CAF formation. We found that during CAF formation the insulation 

resistance levels repeatedly dropped temporarily and then recovered. In addition, 

we observed a substance presumed to be CAF at the sites of probable insulation 

degradation, and we were able to confirm that this substance occurred along the 

glass fibers. 

1 

Introduction 

The electronic components of equipment such as cell phones comprise a great number of 

electronic parts embedded onto PWBs. These PWBs normally contain features such as copper 

patterns stamped onto the boards and throughholes that penetrate the boards. These lines and 

throughholes must have continuous insulation protection from adjacent lines and throughholes. 

(Fig.1) 

Ion migration, which causes insulation degradation, is a phenomenon in which metallic ions 

from the electrodes migrate either on or through the insulation, causing short-circuiting. 

Representative types of this phenomenon are dendrites and CAF. (Fig.2) 

Fig.1 Enlarged PWB photo 

- 1 - 

Fig.2 Sites of ion migration 

Espec Technology Report No.24


Until recently, measures such as refinements in PWB materials had managed to suppress the 

problem of ion migration, but with recent advances producing higher density wiring, ion migration 

has once again become a serious problem. In addition, the use of multi-layered circuit boards and 

built-in circuit boards has caused the formation of CAF to become an increasingly serious 

problem. 

CAF is a failure mode in PWBs that occurs under conditions of high humidity and high voltage 

gradient. The filament, a copper salt, grows from the anode toward the cathode along the 

epoxy/glass interface. 

There is a growing trend toward standardization in evaluating CAF resistance, and in 2003, the 

IPC Association Connecting Electronics Industries issued standards for the CAF Resistance Test. 

Testing is being performed in Japan as well, with such examples as Company X and Company 

Y cited in Table 1, and calls are being made for proposals leading to international standardization. 

The information cited above indicates the growing importance placed on resistance to CAF. 

Current research indicates that CAF is a conductive copper-containing salt created 

electrochemically that grows from the anode toward the cathode subsurface along the 

epoxy/glass interface. 1) However, because CAF grows inside the PWB, the mechanism of its 

occurrence and growth has not been fully resolved. This need for further clarification led us to 

carry out temperature and humidity testing and HAST (highly-accelerated stress testing), and to 

investigate the relationship between the shape of CAF formation and the insulation resistance 

between throughhole walls. 

Table 1 Examples of testing CAF-resistance evaluation (including evaluating 

the reliability of the insulation between throughholes) 

Test 

IPC-TM-650 

2. 6. 25 

Temperature 

(℃) 

65±2 

or 

85±2 

Humidity 

(%rh) 

Bias voltage 

- 2 - 

(V) 

Test voltage 

(V) 

Test time 

(h) 

87 +3 

-2 100 100 500, 1000 

Company X 85 85 100 * 500, 1500 2)、3) 

Company Y 

85 85 50 * 1000 4) 

110 85 50 * 300 5) 

*Not given 



2 

Test method 

For this research, we manufactured PWBs in which it was thought that the sites of insulation 

degradation would be easily identifiable. To carry out the reliability testing, we modified 

temperature and humidity, PWB materials, distance between throughhole walls, and throughhole 

diameter. Using this approach, we investigated the relationship between CAF formation and 

insulation degradation. We continuously monitored insulation resistance during the tests, and 

investigated its behavior. Following the tests, we used a grinder on the sites of presumed 

insulation degradation to visually observe each type with a microscope. 

Photo 1 shows the test equipment. Test equipment included a Bench-top Type Temperature 

and Humidity Chamber in parallel with a Highly Accelerated Stress Test System (HAST Chamber) 

and an Ion Migration Evaluation (Electrochemical Migration Evaluation) System. We used four 

sets of temperature and humidity conditions: 60℃ at 85%rh, 85℃ at 85%rh, 110℃ at 85%rh, 

and 120℃ at 85%rh. The applied test voltage and the measured test voltage were both 50V DC. 

During the tests, we continuously monitored leak current. We used a leak detection cycle that 

stopped the application of voltage upon detecting leak current above a pre-set resistance, and 

also a leak behavior mode (which triggers a behavior check cycle after confirming leak behavior) 

that stopped the application of voltage after being entered above a pre-set number of times. The 

standard leak current setting is 1μA (resistance: 5 x 107Ω), and failure time was determined as 

either leak current reset, or entering the behavior check cycle in five successive checks done at 

one-minute intervals, whichever came first. 

Highly Accelerated Stress Test 

System (HAST Chamber) 

Bench-top Type Temperature and Humidity 

Chamber and Ion Migration Evaluation 

(Electrochemical Migration Evaluation) System 

Photo 1 Test equipment 

- 3 - 



3 

Test specimens 

Fig.3 shows a photo and a diagram of the surface pattern of the PWB used in the tests. The 

PWB type was an eight-layered FR-4, and we used types A, B (halogen-free), C, and D model 

PWBs. Ten throughhole pairs were used on each PWB, with the following relationship of distance 

between throughhole walls to size of throughhole diameter (refer to Fig.4): 0.3/0.2, 0.4/0.2, 

0.5/0.2, 0.3/0.6, and 0.3/1.0 (units: mm, same below). The throughhole pairs were linked using 

a copper pattern, and during the tests each of the ten pairs received voltage application 

simultaneously. The specimens were constructed so that after the tests the copper patterns could 

be cut apart to permit each throughhole pair to have its resistance measured individually. Using 

this method, the resistance of each throughhole pair was measured, and the sites of insulation 

degradation were specified. 

Fig.3 Photo and diagram of surface pattern 

- 4 - 

Fig.4 Diagram of PWB with 

throughholes cross-sectioned 

vertically 



Table 2 lists the test conditions and specimens. 

Conditions 

Specimens 

4 

Test results 

Temp. /Humid. 

Table 2 Test conditions and specimens 

60℃, 85%rh; 85℃, 85%rh; 110℃, 85%rh; 

120℃, 85%rh 

Applied voltage Bias voltage: 50 V, Test voltage: 50 V 

Material type FR-4(8-layer) 

Models PWB A, PWB B (Halogen-free), PWB C, PWB D 

PTH wall-to-wall spaces/ 

Drilled hole sizes 

4-1 Behavior of insulation resistance 

0.3/0.2, 0.4/0.2, 0.5/0.2, 0.3/0.6, 0.3/1.0 

Fig.5 shows the correspondence between time and the behavior of insulation resistance of PWB 

C tested at 110℃ and 85%rh. The relationship of distance between throughhole walls to size of 

throughhole diameter used was 0.3/0.2. To confirm the behavior of insulation resistance leak 

current, standard settings were 50μA (resistance: 1 x 10 6 Ω), continuously testing leading to 

failure, with the test ending at approximately 45 hours. Fig.5 indicates that insulation resistance 

temporarily drops at the time of CAF formation, with resistance then recovering and the pattern 

being repeated. 

Fig.5 Behavior of insulation resistance 

(50V DC at 110℃ and 85%rh, PWB C, 0.3/0.2) 

- 5 - 



4-2 Test results for temperature acceleration 

Fig.6 shows a Weibull plot for temperature acceleration of specimens at 60℃ and 85%rh, 85℃ 

and 85%rh, 110℃ and 85%rh, and 120℃ and 85%rh tested at 50V DC for 2400 hours. The 

PWB was model C with a relationship of 0.3/0.2 for distance between throughhole walls to size of 

throughhole diameter. The “m” in the graph is the shape parameter. 

The Weibull plot trends show a difference only for the conditions of 120℃ and 85%rh. The 

failure mode in the test at 120℃ and 85%rh differed from the failure modes at the other 

temperature and humidity conditions. The conditions of 120℃ and 85%rh in this test may have 

been too harsh for the specimens used. 

Fig.6 Weibull plot of temperature acceleration 

(50V DC for 3000 hours, PWB C, 0.3/0.2) 

- 6 - 



4-3 Test results for different PWBs 

Fig.7 shows a Weibull plot for each type of PWB tested for 500 hours at 110℃ and 85%rh. The 

relationship used for the distance between throughhole walls to throughhole diameter was 

0.3/0.2. 

The test results indicated that PWB C had a shorter time to failure than PWBs A and B. 

Only three of the nine specimens of PWB B (halogen-free) failed within the 500 hours of test 

time, yielding the longest time leading to failure of any specimen. 

Fig.8 shows the humidity absorption characteristics of these PWBs at 110℃ and 

85%rh. No correlation was seen between humidity absorption rate and failure time, indicating 

that it may be difficult to determine CAF resistance solely on the basis of humidity absorption 

characteristics. Factors considered to affect CAF resistance include the flux residue in the inner 

layers, the damage to throughhole walls from drilling, and the interface bond between the epoxy 

and glass fibers. 

Fig.7 Weibull plots for the different PWBs 

(50V DC for 500 hours at 110℃, 

85%rh, 0.3/0.2) 

- 7 - 

Fig.8 Humidity absorption 

characteristics 

(110℃, 85%rh) 



4-4 Test results for different distances between throughhole walls and diameter sizes 

Fig.9 and 10 show Weibull plots for each specimen based on differing distances between 

throughhole walls. The specimens were tested for 3000 hours at 85℃ and 85%rh. PWB C was 

used. 

With the range of distances between throughhole walls of 0.3 to 0.5 mm, the smaller the 

distance between the throughhole walls, the shorter the time leading to failure. 

No clear difference was seen among the different throughhole diameters, from 0.2 to 1.0 mm. 

The data was not consistent, and so we plan to experiment further. 

Fig.9 Weibull plot for each distance 

between throughhole walls 


85%rh, PWB C) 

- 8 - 

Fig.10 Weibull plot for each throughhole 

diameter 


85%rh, PWB C ) 



5 

Results of observation 

5-1 Results of observation (1) 

Fig.11 shows the changes in insulation resistance during the test at 110℃ and 85%rh for PWB 

C, Sample 1. The relationship of distance between throughhole walls to size of throughhole 

diameter was 0.4/0.2. Approximately 90 hours after test start-up, insulation resistance fell below 

5 x 10 5 Ω (marked with a red x in the diagram), and the test was completed. Following the test, 

we observed cross sections of the sites of insulation degradation on the PWB. 

Fig.11 Changes in insulation resistance (Sample 1) 


- 9 - 



Photos 2 and 3 are microscope photographs of pre- and post-test specimens. The unevenness 

of the throughholes in the pre-test photo is thought to be caused by damage from drilling where 

traces of Cu plating have become embedded. In microscopic observation following the test, 

substances thought to be CAF were observed along the glass fibers in the epoxy. To obtain a 

more detailed analysis, we performed a Cu mapping analysis using a metallurgical microscope 

and EPMA (Electron Probe Micro-Analyzer). 

Photo 2 Pre-test specimen microscope 

image 

- 10 - 

Photo 3 Post-test specimen microscope 

image 

(Sample 1) 

Photos 4 and 5 are images obtained from the metallurgical microscope and Cu mapping. From 

the metallurgical microscope image, this substance is believed to be metallic. From the Cu 

mapping image, the main component of the substance is found to be Cu. 

Photo 4 Metallurgical microscope image 

(Sample 1) 

Photo 5 EPMA Cu mapping image 

(Sample 1) 




Next, we shall consider changes in insulation resistance during the test, and post-test 

observation of insulation degradation sites for Samples 2, 3, and 4 of PWB C at the same test 

conditions of 110℃ and 85%rh. 

Fig.12, 13, and 14 show changes in insulation resistance for Samples 2, 3, and 4. The green X 

in the diagrams indicates the formation of leak touch, with a resistance below 5 x 10 7 Ω. A red X 

indicates a resistance of 5×10 5 Ω or below. (However, since the leak current standard setting for 

Fig.14 was 50μA, resistance indicated is below 1 x 10 6 Ω.) 

The repeated fall and recovery of resistance seen in Fig.13 is conjectured to be repeated 

breaking and connecting of a fine section of CAF. 

Fig.12 Changes in insulation resistance 

(Sample 2) 

(50V DC at 110℃ and 85%rh, 

PWB C, 0.5/0.2) 

Leak current setting, 1μA 

(resistance, 5 x 10 7 Ω) 


(Sample 4) 

(50V DC at 110℃ and 85%rh, 

PWB C, 0.3/0.2) 



- 11 - 


(Sample 3) 

(50V DC at 110℃ and 85%rh, 

PWB C, 0.4/0.2) 





Photos 6 though 9 are post-test microscope images of sites of insulation degradation. For 

Samples 1 and 2, the test was concluded when resistance fell below 5 x 10 5 Ω. For Sample 3, the 

test was concluded when resistance fell to 7.21 x 10 8 Ω. For Sample 4, the test was concluded 

when resistance fell to 5.11 x 10 9 Ω. 

Photos 6 through 9 indicate that in the specimens with insulation degradation set at below 5 x 

10 5 Ω the CAF diameter is thicker than that in the 7.21 x 10 8 Ω and 5.11 x 10 9 Ω. This experiment 

indicates that the density may have increased after the CAF connected the anode and cathode, 

leading to insulation degradation. 

Photo 6 Site of insulation degradation 

(below 5 x 10 5 Ω) 

(Sample 1) 


(7.21 x 10 8 Ω) 

(Sample 3) 

- 12 - 


(below 5 x 10 5 Ω) 

(Sample 2) 


(5.11 x 10 9 Ω) 

(Sample 4) 




Fig.15 shows changes in insulation resistance during the test at 110℃ and 85%rh for PWB D, 

Sample 5. The relationship of distance between throughhole walls to size of throughhole diameter 

was 0.3/0.2. Following the test, we observed cross sections of sites of probable insulation 

degradation on this PWB. Photos 10 and 11 are microscope images of the pre- and post-test 

specimen. 

Observation indicated that the insulation degradation on this specimen was not caused by CAF 

along glass fibers, but rather by ion migration forming between the core material* and the 

prepreg sheets**. 


(50V DC at 110℃ and 85%rh, PWB D, 0.3/0.2) 

Photo 10 Pre-test microscope image 

*Core material: inner layer laminate of multi-layer PWB 

**Prepreg sheets: Adhesive sheet used to bond core material layers 

- 13 - 

Photo 11 Post-test microscope image 

(Sample 5) 




Next, we shall consider PWB C Samples 6 and 7 in tests at 110℃ and 85%rh with regard to 

examples of CAF observed even in specimens analyzed as non-defective when the test was 

concluded prior to insulation degradation. For Sample 6, Fig.16 shows changes in insulation 

resistance, Photo 12 is a post-test microscope image, and Photo 13 is a post-test metallurgical 

microscope image. For Sample 7, Fig.17 shows changes in insulation resistance, Photo 14 is a 

post-test microscope image, and Photo 15 is a post-test metallurgical microscope image. 

From Fig.16 and 17 and Photos 12 through 15 we can see that neither Sample 6 nor Sample 7 

have any major changes in insulation resistance. However, post-test observation revealed CAF 

along the glass fibers. Observation of Sample 7 revealed CAF growing from both the anode and 

the cathode, but since three-dimensional observation and observation during the test are 

problematic, this did not lead to clarification of the process of CAF growth. 




(Sample 6) 

- 14 - 

Photo 13 Post-test metallurgical 

microscope image 

(Sample 6) 






(Sample 7) 

- 15 - 



(Sample 7) 




Up to this point, we have considered observation of PWBs with the throughholes 

cross-sectioned vertically. Next we shall consider observation of PWBs with the throughholes 

cross-sectioned horizontally. Fig.18 shows changes in insulation resistance during the test at 

120℃ and 85%rh for PWB D, Sample 8. The relationship of distance between throughhole walls 

to size of throughhole diameter was 0.3/0.2. Photo 16 is a pre-test microscope image, while 

post-test images of sites of insulation degradation are presented in Photo 17, a microscope 

image, and Photo 18, a metallurgical microscope image. 

The metallurgical microscope image shows CAF growing from the anode toward the cathode, 

and it was confirmed to be extending obliquely toward the back. This is presumed to be caused by 

the existence of CAF along the glass fibers. 


(Sample 8) 

(50V DC at 120℃ and 85%rh, 

PWB D, 0.3/0.2) 


(Sample 8) 

- 16 - 

Photo 16 Pre-test microscope 

image 



(Sample 8) 



6 

We carried out High Temperature and Humidity Tests and Highly-Accelerated Stress Tests 

(HAST) to evaluate the CAF resistance of multi-layered PWBs. These experiments produced the 

following results. 

- At the time of CAF occurrence, insulation resistance exhibits a cyclical behavior of repeatedly 

falling temporarily and then recovering. 

- PWB C exhibited a shorter time leading to failure than either PWB A or B, but since PWB C did 

not exhibit a noticeably high level of humidity absorption in the Humidity Absorption test, 

determining CAF resistance characteristics solely on the basis of humidity absorption 

characteristics seems to be problematic. 

- The tests at 120℃ and 85%rh may have been too harsh for the PWB C specimen used in these 

tests, so that the failure mode may have been different from the failure mode at the other 

conditions of temperature and humidity. 

- CAF along the glass fibers and a substance believed to be ion migration occurring between the 

core material and the prepreg sheets were observed at presumed sites of insulation 

degradation. In addition, CAF occurrence was observed in specimens that did not exhibit 

insulation degradation. 

7 

Conclusion 

Challenges for further research 

In this research, we observed CAF growing from the anode as well as from the cathode, and we 

plan to carry out further research regarding the growth process and mechanism. In addition, 

investigation is required on what effect the manufacturing conditions and pre-conditioning of 

multi-layered PWBs have on the formation and growth of CAF. 

It must be stressed that the evaluations produced by this research are relative, and their 

correlation to actual failure in the field are a challenge for further research. We plan to carry out 

further reliability testing with other temperature and humidity conditions to investigate 

accelerated temperature and humidity characteristics. 

[Bibliography] 

1) L. J. Turbini, Ph.D.: “Conductive Anodic Filament (CAF) Formation: An Historic Perspective”, 

ECWC10 Conference, 2005. 

2) F. Ishigami, H Sakai, and Y. Nakamura, Hitachi Chemical Technical Report, No.39, pp.25-28, 

2002. 

3) M. Miyatake, H. Murai, T. Fukuda, and S. Shimaoka, Hitachi Chemical Technical Report, No.45, 

pp.31-35, 2005. 

4) Y. Nakamura, T. Asano, and N. Ito, Matsushita Electric Technical Report, No.75, pp.14-18, 

2001. 

5) K. Komori, T. Watanabe, and Y. Matsushita, “Low-Dielectric-Constant Multi-layer PCB Material 

‘MEGRON5,’” Matsushita Electric Technical Report, Vol.52, No.1, pp.23-27, 2004. 

- 17 - 


New JTM standards for temperature test chambers - Methods of testing and indicating performance 

Understanding the Technology 

New JTM standards for temperature test chambers 

- Methods of testing and indicating performance 

Toshimi Ishida Environmental Test Business Headquarters, Business Control Department 

J 

apan Testing Machinery Association (JTM) has established standards for 

environmental test equipment. 

JTM has issued a new standard for environmental test chambers called JTM K 07, 

"Temperature chambers-Test and indication method for performance". This 

standard conforms to standards established by IEC (International 

Electrotechnical Commission) and JIS (the Japanese Standards Association), and 

has major changes from previous methods of testing and indicating performance 

for environmental test equipment. As a result, we would like to present the 

background leading to the establishment of this standard as well as its principal 

contents. In addition, we would like to discuss the coming trends in methods of 

testing and indicating performance for temperature and humidity test chambers. 

1 

Trends in standards for environmental test equipment 

Efforts to standardize environmental test equipment have been under way continuously since 

the "Engineering industry standardization project" of 1975. In 1976, MITI (the Ministry of 

International Trade and Industry) organized special committees to promote standardization, 

including the Japan Machinery Federation (JMF), Japan Testing Machinery Association (JTM), and 

related organizations, and investigations began on means of standardization. Following that, 

investigative research from those committees began accumulating, and JTM established 

standards related to performance and safety. Below is a list of the establishment and revision of 

some of the major performance standards. 

1985: JTM K 01, Temperature and humidity chambers - Test and indication method for 

performance (Established) 

1988: JTM K 03, Environmental temperature and humidity rooms - Test and indication 

method for performance (Established) 

1991: JTM K 05, High temperature chambers - Test and indication method for performance 

(Established) 

1998: JTM K 01 (Revised) 



- 18 - 



The driving force for promoting environmental testing software has included the United States 

military standards (MIL) and the IEC and JIS standardization of electrical, automotive, and 

maritime standards. JTM has been at the forefront for promoting the standardization of 

hardware. These industrial standards have been established for the purpose of contributing to the 

improvement and maintenance of product quality and to international standardization. 

The IEC standards and JIS standards for environmental test equipment have been established 

as follows. 

2001: IEC 60068-3-5, Environmental testing - Part 3-5: Supporting documentation and 

guidance - Confirmation of the performance of temperature chambers 

2006: JIS C 60068-3-5, Environmental testing - Part 3-5: Supporting documentation and 


The contents of this JIS standard matches the contents of the above IEC standard. The IEC and 

JIS standards for temperature test chambers are utilized for all equipment used in environmental 

testing and are provided for the general use of the user to confirm the performance of test 

chambers. Since these standards constitute guidance, there is a danger that when put into 

practice they may be subject to a number of interpretations. Because of that, JTM has taken a 

proper understanding of the contents of these standards, and including portions that are unclear 

or not stipulated in the standards as well as contents that made up the previous JTM standards 

has developed new JTM standards. 

The new JTM standards target the “temperature test chambers” and have changed the 

contents of JTM K 01 (Temperature and humidity chambers - Test and indication method for 

performance), JTM K 03 (Environmental temperature and humidity rooms - Test and indication 

method for performance), and JTM K 05 (High temperature chambers - Test and indication 

method for performance) to create the new JTM standards. However, the new standards will be 

phased in over the course of 5 years following their publication. 

In addition, JTM is planning to publish new standards targeting “temperature and humidity test 

chambers” as well. 

- 19 - 



2 

Table 1 presents an overview of the major revisions to the standards. 

Method of indicating 

performance 

Method of confirming performance 

New test and indication methods for performance 

Distance of 

working 

space 

Range of temp. 

fluctuation 

Temp. 

uniformity 

Heating/cooling 

time 

Table 1 Principal changes in the new JTM standards 

Current standards New standards Comments 

Range of temp. fluctuation Temp. fluctuation Method of confirming 

performance also 

changed 

Temp. uniformity Temp. gradient, Temp. 

variation in space 

- 20 - 

Method of confirming 


changed 

Heating/cooling time Temp. rate of change Method of confirming 


changed 

Space excluding 1/6 of 

distance from inner walls 

Measurement point: single 

point at center of chamber 

Calculation method: take the 

difference between average 

high and average low temps, 

divide by two, and express as 

± value 

Concept: Difference between 

center of chamber and 8 corner 

points 

Definition: Time required to 

reach max/min temp. from 

ambient temp. 

Expression: n1℃ to n2℃, 

x minutes 

Let’s consider the major points of the new standards. 

Space excluding 1/10 of 

distance from inner walls 

Measurement points: 9 points 

(chamber center and 8 

corners) 

Calculation method: take the 

highest value of the standard 

deviation among all of the 

measurement points, double it, 

and express as ± value. 

Concept: Difference between 

center of chamber and 8 corner 

points, and difference among 

corners 

Definition: Rate of temp. 

change during one minute for 

specified times 

Expression:K/minute or ℃ 

/minute 

Minimum values 

specified for different 

test chamber volumes 

Time to reach temp. 

extreme conforms to 

prior 

heating/cooling 

time 



2-1 Working space 

The working space is a test chamber area that can maintain specified temperature conditions 

within an allowable range. The test chamber generally takes the shape of a rectangular box, and 

the working space is composed of the area surrounded by the shaded surface shown in Fig.1. The 

distance from each of the wall surfaces surrounding the working space to the next corresponding 

chamber surface shall be termed as L1, L2, and L3, respectively. Current standards define the 

working space as the space between the surrounding walls excluding the area adjacent to the 

walls composed of a distance divided by 6 from the total surface distance, expressed as L1/6, L2/6, and L3/6. The new standards define the working space as the space between the 

surrounding walls excluding the area adjacent to the walls composed of a distance divided by 10 

from the total surface distance, expressed as L1/10, L2/10, and L3/10. However, since it is 

necessary to consider the effect of thermal emissions in the space between the inner walls of the 

test chamber and the test specimens, the distance from the inner walls must be specified as a 

minimum value according to the volume of each test chamber. 

Fig.1 Working space 

- 21 - 



2-2 Range of temperature fluctuation 

The range of temperature fluctuation in the current standards is defined as the difference 

between the average high temperature and the average low temperature at the center of the test 

chamber. The new standards refer to temperature fluctuation, which is defined as the 

difference between the maximum and minimum temperatures at discretionary points in the 

working space measured at specified time intervals after the temperature has stabilized. The new 

standards have set the discretionary points as being at the center and at the corners of the 

working space. As can be seen in Fig.2, to find the range of fluctuation according to current 

standards, average temperatures are first calculated for a measured point in the center of the 

chamber. Then, the average temperatures are separated into high and low temperatures and the 

average high and low temperatures are calculated, and the fluctuation is defined as the difference 

between these high and low averages. This difference is divided by two, and then that result is 

displayed as a ± value. 

Fig.2 Temperature fluctuation in current standards 

The calculation method used by the current standards is as follows. 

Average temp. = 

Average high temp. = 

Average low temp. = 

Range of temperature fluctuation = ±(average high temp. - average low temp.)/2 

- 22 - 



No specific measurement procedures are presented in JIS C 60068-3-5 (IEC 60068-3-5). 

Actual measurements are performed at user-selected sampling intervals, and so one cannot say 

that the data has necessarily captured the peak temperatures. 

The new standards call for statistical estimation of temperature fluctuation. After the 

temperature has stabilized, the temperature is measured at regular intervals at the specific 

measurement points, a minimum of 10 times for 30 minutes, and is represented as X1,X2, …Xi… X n (n ≧10). XAve represents the average value of X, and in X i, the i represents a number from 1 

to n. Using these, the σ n-1 specimen standard deviation is defined as follows. 

Temperature fluctuation is found by taking ±2σn-1 for each of the 9 measurement points inside 

the working space and notating the greatest of these values as the temperature fluctuation. Units 

are represented in K (Kelvin, absolute temperature) or ℃ (Celsius). 

Fig.3 Temperature fluctuation in new standards 

- 23 - 



2-3 Temperature uniformity 

In the current standards, temperature uniformity is defined as the temperature differential 

between the temperature at the center of the chamber and discretionary points in the working 

space. For the high temperature chamber or the temperature and humidity chamber, 

measurements are taken at a minimum of five points, which include the center of the chamber 

and at selected symmetrical points around the center, four of which are as far from the center as 

possible. For the temperature and humidity chamber, measurements are taken at nine points, 

which include at the center of the chamber and at each of the eight apices. First the average 

temperatures are calculated for each point, and then the average temperature at the center point 

is divided by the high and low temperatures. Next, the average temperatures are found for the 

high-temperature and low-temperature points. The temperature uniformity is the difference 

between the high and low average temperatures. This value is displayed as a ± value divided by 

2. 

The new standards, however, do not use the expression temperature uniformity, opting instead 

for the two terms temperature variation in space and temperature gradient. 

Temperature variation in space refers to the difference in average temperature between the 

center of the working space and separate discretionary points at some point in time after the 

temperature has stabilized. Measurements are taken at nine points: at the center of the working 

space and at the eight apices. 

Temperature variation in space is defined as the maximum average difference between the 

average temperature at the center of the working space and the average temperatures at each of 

the other measurement points. This difference is expressed as an absolute value. 

Average temperature at the center of the working space: XCenterAve 

Average temperature at each of the working space apices: XCornerAve(j); j = 1 to 8 

Temperature variation in space = |Max(XCornerAve(j) - XCenterAve)| 

The temperature gradient is the greatest average difference between the average 

temperatures at two separate points within the working space at discretionary points in time after 

the temperature has stabilized. The nine measurement points are specified as at the center of the 

working space and at the eight apices. The temperature gradient is defined as the maximum 

difference in average temperature among the measurement points. 

Maximum average temperature at each measurement point: XAveMax 

Minimum average temperature at each measurement point: XAveMin 

Temperature gradient = XAveMax - XAveMin 

To sum up, temperature variation in space indicates how great a difference exists between the 

center of the working space and the rest of the working space. Temperature gradient indicates 

how great a difference exists among each of the various points in the working space. The two 

terms together express the concept of temperature uniformity. 

- 24 - 



2-4 Heating and cooling time 

In the current standards, heating and cooling time refers to the time required to reach the 

maximum or minimum temperature of the temperature control range from the ambient 

temperature under standard conditions at the center of the chamber. In the new standards, this 

term has been replaced with time to reach temperature extremes. The new standards have 

also added the term temperature rate of change. The temperature rate of change indicates 

the temperature for one minute, and refers to measuring the rate of temperature change 

between two specified points in time at the center of the working space. Fig.4 shows rate of 

change for an interval ten percent prior to both the maximum and minimum temperatures. 

Fig.4 Temperature rate of change for heating and cooling a test chamber 

The temperature rate of change is found using the following formula. 

Temperature heating rate = △ t/T1 Temperature cooling rate = △ t/T2 The closer the temperature comes to the maximum or minimum temperature, the slower the 

rate of change, and so the temperature rate of change excludes that portion just prior to reaching 

maximum or minimum temperature. Because of this, it is not possible to know the actual time at 

which the maximum or minimum temperature is reached. Therefore, the heating and cooling 

time of the previous standard has changed only in its expression as time to reach temperature 

extremes, but the content adheres to the current concept with no change. The time to reach 

temperature extremes refers to the time to reach or pass through the temperature extreme from 

the ambient temperature under standard conditions at the center of the working space. 

- 25 - 



3 

Current standards and the activation of new standards 

The current standards will expire five years after the publication of the new standards. The new 

standards are for the temperature test chamber. At the current stage JIS has not yet 

published new standards for the temperature and humidity test chamber, and so the new 

standards for the temperature test chamber will precede those for the temperature and humidity 

test chamber. As a result, current standards will continue to apply to the temperature and 

humidity test chamber, and only those contents concerning the temperature test chamber will 

proceed to application of the new standards. 

Table 2 Test chambers and standards subject to transition to new standards 

Current 

standard 

number 

JTM K01 

JTM K03 

JTM K05 

Name of test chamber 

subject to current 

standard 

Temperature chamber Transition to K07 as 

temperature test 

chamber 

Temperature and Current standard K01 

humidity chamber stays in effect 

Temperature room Transition to K07 as 


chamber 

Temperature and Current standard K01 

humidity room 

stays in effect 

High temperature 

chamber 

Details of transition Comments 

Full transition to K07 as 


chamber 

- 26 - 

Measure expires 5 

years after transition 

Enactment of newer 

standard already 

planned 



Enactment of newer 

standard already 

planned 





4 

Coming standards 

The current standards have been established over a long period of time utilizing both domestic 

and foreign investigations as JTM has worked to standardize environmental testing equipment. 

The current standards are used not only in Japan but also around the world. 

IEC 60068-3-5, enacted in 2001, has finally become a global standard. Five years have elapsed 

since its enactment as an international standard, and yet this standard is still not widely accepted 

throughout the world. In Europe there are wide variations in how this standard is applied. In 

America and Asia, even the contents are not well known. 

JIS C 60068-3-5 was enacted in 2006, and so there has been a lot of discussion in Japan as to 

whether new JTM standards are necessary. The new JTM standards will not bring about changes 

in the hardware for environmental testing equipment, but rather will bring about changes in 

testing methods and in ways of indicating or expressing performance. Therefore, although the 

equipment may be the same, there is some concern that these changes have produced the 

misunderstanding that a loss of performance has occurred. 

However, if we should disregard this international standard and remain at the current stage of 

domestic industrial standards, the situation would not be conducive to preparing for the coming 

global development. Japan Testing Machinery Association has borne the mission of supporting 

the development of the Japanese testing technology and the environmental equipment testing 

technology that have laid the foundation for the high quality of Japanese products, and JTM will 

bear the mission for taking a leading role globally in the future as well. Because of this, the new 

JTM standard is coordinated with the contents of the IEC 60068-3-5 standard, supplementing 

obscure points and producing a more well-developed standard. 

Regarding temperature and humidity test chambers, the IEC 60068-3-6 standard was enacted 

in 2001. For this international standard as well, JTM plans to create a new JTM standard that 

supplements the obscure points and further develops the standard. 

- 27 - 



5 

Conclusion 

The IEC and JIS standards for temperature test chambers are utilized for all equipment used in 

environmental testing and are provided for the general use of the user to confirm the 

performance of test chambers. However, the contents are guidance, and there is a danger that 

they can be subject to all kinds of interpretations when applied. To eliminate the variations caused 

by the interpretations among different users, we highly recommend that the new standards 

published by JTM be confirmed. 

[Bibliography] 

1) The Japan Machinery Federation : 1981, Japan Testing Machinery Association: Investigative 

report for the standardization of temperature and humidity equipment 

2) The Japan Machinery Federation : 1986, Japan Testing Machinery Association: Investigative 

research report for the standardization of temperature and humidity rooms 

3) JIS C 60068-3-5: 2006, Environmental testing - Part 3-5: Supporting documentation and 


4) IEC 60068-3-5: 2001, Environmental testing - Part 3-5: Supporting documentation and 


5) Japan Testing Machinery Association: 1994, Investigative research report for the 

standardization of measurement and control technology for environmental equipment 

6) Japan Testing Machinery Association: 1995, Investigative research report for the 

standardization of measurement and control technology for environmental equipment 

7) JTM K 01: 1998, Temperature and humidity chambers - Test and indication method for 

performance 

8) JTM K 03: 2001, Environmental temperature and humidity rooms - Test and indication method 

for performance 

9) JTM K 05: 2000, High temperature chambers - Test and indication method for performance 

- 28 - 



Topic 


Kazuhiro Sakonaka, Manabu Shirahase Flat Panel Display Products Division 

1 

Introduction 

We at Espec are proud to present our new 500℃ clean oven, perfectly suited to flat panel 

display manufacturing processes such as field emission displays, low-temperature polysilicon, 

and organic electroluminescence. This model maintains a high Class 4 (0.3μm) classification for 

air cleanliness at all temperature ranges, including temperature ramp-up and cool-down, as well 

as during heat treatment at oxygen concentrations of 10 ppm and below. 

This new oven not only offers a space-saving slim design, it reduces the environmental load by 

more than half from previous models by virtue of features such as low energy and low nitrogen 

gas consumption. 

Photo 1 Super Clean Oven 

- 29 - 



2 

Product features 

2-1 Moving up to Class 4 (0.3μm) classification for air cleanliness 

Previous ovens have had difficulty maintaining the level of cleanliness during heat ramp up and 

cool down, but the new Espec proprietary oven construction utilizes a special heat-resistant filter 

that attains the Class 4 (0.3μm) classification for air cleanliness in accordance with 

ISO14644-1/JIS B9920 (previously Class 10 under the canceled FED-STD-209E standard) not 

only at stable temperatures but during temperature change as well, reducing the amount of 

particulate matter falling onto the glass. 

How we moved up to Class 4 (0.3μm) classification for air cleanliness 

- Using the special high-heat-resistant filter developed by Espec for this model as well enables 

continuous operation up to a maximum 500℃. In addition, this filter makes it possible to 

maintain a high level of cleanliness even during temperature ramp up and cool down. Using this 

special heat-resistant filter at 500℃ delivers a clean oven in a hot air circulation environment. 

- We simplified the air conditioning method, rearranging the water-cooled heat exchanger 

placement and eliminating the damper inside the chamber. In this way, we were able to 

eliminate the sliding section that could generate dust particles. 

- By undertaking a fundamental reappraisal of the sheet metal construction and parts inside the 

chamber, we were able to minimize the contact and friction of the sheet metal and parts caused 

by thermal stretching. 

Fig.1 Level of cleanliness at 500℃ operation 

- 30 - 





- 31 - 



2-2 Improved temperature uniformity 

Rearranging the water-cooled heat exchanger placement allowed us to eliminate the damper. 

This new design simplifies the air conditioning method, making it easier to maintain temperature 

uniformity. 

How we attained improved temperature uniformity 

- When we simplified the air conditioning method by moving the water-cooled heat exchanger 

and eliminating the damper from inside the chambers, pressure loss was reduced, making it 

possible to increase the quantity of circulation air that is required to maintain performance. 

- Eliminating excess space required for air conditioning also achieves a space-saving design while 

maintaining sufficient double inner walls and sufficient air insulation space on all sides of the 

inner chamber racks. 

- In addition, we revised the insulation materials, using new materials with a smaller coefficient 

of thermal conductivity. 

Fig.4 Temperature uniformity at 500℃ operation 

- 32 - 





- 33 - 



2-3 Heat treatment capability in a low-oxygen atmosphere 

Heat treatment is possible not only during temperature ramp up and at stable temperatures, 

but is now possible during cool down as well, even in low-oxygen atmospheres of 10 ppm oxygen 

or less. 

Fig.7 Oxygen concentration at 500℃ operation 

How we achieved heat treatment in a low-oxygen atmosphere 

- The cooling method consists of circulating out the hot air, and so in general cooling is achieved 

by means of introducing outside air into the chamber. However, with low oxygen specifications, 

a water-cooled heat exchanger* replaces the induction of outside air to achieve a hermetically 

sealed construction. The introduction of a minimal amount of nitrogen gas is used to attain an 

oxygen concentration of 10 ppm inside the chamber. 

- Sensors on every port leading to the inside of the chamber are used to maintain the 

hermetically sealed environment with the use of metal seals. 

*The water-cooled heat exchanger is not placed inside the chamber, and so safety is maintained 

with no danger of the water boiling when the operation is stopped at 500℃. 

- 34 - 



2-4 Energy-efficient and space-saving 

By revising air conditioning and cooling methods, materials used inside the chamber, heaters, 

circulating fans, and the construction of heat-resistant filters, we have been able to reduce by half 

the “equipment footprint,” “power consumption,” and “nitrogen usage (at low oxygen 

concentrations only)” when compared to our previous HTC-70A model. 

2-4-1 How we reduced the equipment footprint by half 

We undertook a fundamental reappraisal of the water-cooled heat exchanger placement with 

special regard to the standards for low oxygen concentration specifications. The results of this 

reappraisal led to the elimination of the damper that was inside the chamber. By eliminating 

excess air conditioning space, we were able to reduce the overall size of the oven from 11 m 2 to 

6 m 2 . The elimination of the damper reduced the maintenance area as well. (Compared to the 

Espec HTC-70A) 

Fig.8 Comparison of equipment footprint 

2-4-2 How we reduced power consumption by half 

Eliminating the excess air conditioning space reduced the amount of heat required inside the 

chamber, making it possible to cut the number of circulating fans in half, from two to one, and 

reducing the heater capacity from 54.6 kW to 24 kW. This cut our power consumption in half. 

(Compared to the Espec HTC-70A) 

2-4-3 How we reduced nitrogen usage by half 

The reduced volume of the chamber and the reduced number of circulating fans, from two to 

one, cut the nitrogen requirement in half. We were able to lower the replacement nitrogen to 

500 L/min. (Compared to the Espec HTC-70A) 

- 35 - 



3 

Specifications 

Table 1 shows the main specifications of the new oven. Fig.9 shows door operation, and Fig.10 

shows a detailed diagram of the oven construction. 

Table 1 Main specifications 

Model No. SCO-1 (Glass accommodation: max. 400 x 500 mm) 

Temperature range +100℃ to +500℃ 

Temperature 

uniformity 

SCO-2 (Glass accommodation: max. 650 x 750 mm) 

SCO-4 (Glass accommodation: max. 800 x 1300 mm) 

±3.0℃ (at 300℃) 

±4.0℃ (at 400℃) 

±5.0℃ (at 500℃) 

Level of cleanliness Class 4, 0.3μm 

Nitrogen consumption 500 L/min. (SCO-2) 

(equiv. to Class 10 of former FED-STD-209E) 

Power consumption 200 V AC, 3φ, 29 kVA (SCO-2) 

Please contact us for information about other models and voltages. 

Fig.9 Oven door operation 

- 36 - 



4 

Conclusion 

Fig.10 Oven construction 

4-1 Effect of manufacturing conditions of MEA specimens 

The Super Clean Oven SCO-2 is not only perfectly suited to flat panel display manufacturing 

processes such as field emission displays, low-temperature polysilicon, and organic 

electroluminescence, this oven is also well-suited as research equipment for applications such as 

improving other manufacturing processes and developing materials. 

Contact information 

For further information on the Super Clean Oven SCO-2, or to request a catalog, please contact 

the Espec Corporation System Sales Engineering Department of Flat Panel Display Products 

Division. 

- 37 -

download

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?