Page Yield/Ink Cartridge Reliability
Comparison Study
HP Inkjet Print Cartridges
vs. Refilled Brands
April 2007
For distribution in North America
5401 Tech Circle 鈥? Moorpark, CA 93021 鈥? 805 531 9030
6148 N. Discovery Way, Ste. 175 鈥? Boise, ID 83713 鈥? 208 424 1905
漏 2007 QualityLogic Inc.
Reliability Comparison Study
Executive Summary
In April 2007, QualityLogic completed a study for HP designed to test the page yield and
reliability of Hewlett-Packard (HP) inkjet print cartridges compared to 13 brands of refilled
inkjet print cartridges for the HP DeskJet 5650 (C6490A) and HP DeskJet 6940 (C8970A)
printers. All brands tested were sold in North America. The refilled inkjet print cartridges
were comprised of refilled cartridges carrying the store brand, refilled cartridges sold under
a third-party brand, and empty cartridges that QualityLogic had refilled at refill stations.
The results of the study show that HP inkjet print cartridges outperformed the refilled inkjet
print cartridges.
Page Yield 鈥? When looking at the total pages printed from those cartridges tested, on
average Original HP inkjet print cartridges produced 57.5% more pages than the refill
brands tested.
Cartridge Reliability 鈥? None of the Original HP inkjet print cartridges were dead-on-arrival
or premature failure (as defined in Appendix 2 鈥? Definitions), whereas, on average 20.9% of
the refill cartridges tested were either dead-on-arrival or failed prematurely.
Additional Observations 鈥? For refill stations, cartridge refill times varied from between
15-20 minutes to more than 24 hours. Not all cartridges could be successfully refilled. Of all
cartridge refill attempts at refill stations, 40 units failed or became unusable during the refill
process.
Refilled Brands Tested
Store/Refill brands Refill station brands
鈥? OfficeMax store brand 鈥? OfficeMax refill station
鈥? Office Depot store brand 鈥? Office Depot refill station
鈥? PrintPal refilled cartridges 鈥? Cartridge World refill station
鈥? Caboodle Cartridge refilled cartridges 鈥? Island Ink-Jet refill station
鈥? Rapid Refill Ink refilled cartridges 鈥? Walgreens refill station
鈥? Staples store brand
鈥? Nu-kote refilled cartridges
鈥? Corporate Express store brand
Page 2 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Detailed Results
Page Yield
When looking at the total pages printed from those cartridges tested, on average Original
HP inkjet print cartridges produced 57.5% more pages than the refill brands tested based
on comparing the normalized 90% lower confidence bound values (as defined in Appendix 2
鈥? Definitions).
More specifically, when comparing HP to the store/refill brands tested only, based on the
total pages printed for this subset of cartridges tested, on average Original HP inkjet print
cartridges produced 36.2% more pages than the store/refill brands (brands A-H) tested
based on comparing the normalized 90% lower confidence bound values (as defined in
Appendix 2 鈥? Definitions).
When comparing HP to the refill station brands tested only, based on the total pages printed
for this subset of cartridges tested, on average Original HP inkjet print cartridges produced
110.1% more pages than the refill station brands (brands I-M) tested based on comparing
the normalized 90% lower confidence bound values (as defined in Appendix 2 鈥? Definitions).
In the chart below, each brand is referenced in comparison to the performance of the HP
brand. The normalized 90% lower confidence bound page yield values (as defined in
Appendix 2 鈥? Definitions) were used for this comparison. These values from each cartridge
model tested were combined to create a single index value for each brand tested. The score
for the HP brand is represented as 100.0%, with the scores for each refilled brand
represented as compared to the HP brand (higher or lower based on performance.) For the
HP/Brand Relative Page Yield comparison, the percentages shown indicate that HP
cartridges printed that percentage more pages than the given brand. For example, Brand A
had a score of 76.0% compared to the HP brand, indicating that HP printed 31.6% more
pages than Brand A.
Page 3 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Number of Normalized 90% HP/Brand
Cartridges Lower Confidence Relative Page
Tested Bound Page Yield Yield Difference
Brand Type Brand
HP HP 68 100.0% n/a
Store/Refill Brand Brand A 68 76.0% 31.6%
Store/Refill Brand Brand B 68 92.8% 7.8%
Store/Refill Brand Brand C 68 57.2% 74.8%
Store/Refill Brand Brand D 68 71.9% 39.1%
Store/Refill Brand Brand E 68 61.0% 63.9%
Store/Refill Brand Brand F 34 91.2% 9.6%
Store/Refill Brand Brand G 35 54.7% 82.8%
Store/Refill Brand Brand H 34 82.2% 21.7%
Refill Station Brand I 68 63.0% 58.7%
Refill Station Brand J 72 38.3% 161.1%
Refill Station Brand K 68 45.1% 121.7%
Refill Station Brand L 68 51.5% 94.2%
Refill Station Brand M 45 40.1% 149.4%
Sum/Average of All Store/Refill Brands 443 73.4% 36.2%
Sum/Average of All Refill Station Brands 321 47.6% 110.1%
Sum/Average of All Refilled Brands 764 63.5% 57.5%
Table 1:
Cartridge Page Yield
Cartridge Page Yield
Normalized 90% Lower Bound
120% 100.0%
100%
73.4%
80% 63.5%
47.6%
60%
40%
20%
0%
s
E
P
s
F
H
s
D
C
M
L
K
B
I
G
J
A
nd
nd
nd
d
H
d
d
d
d
d
d
d
d
d
d
d
d
an
an
an
an
an
an
an
an
an
an
ra
an
ra
an
an
ra
Br
B
Br
Br
Br
Br
B
B
Br
Br
Br
Br
Br
Br
Br
Br
n
i ll
d
io
il le
ef
at
R
ef
St
e/
R
or
ill
ll
St
ef
A
R
of
ll
A
ll
e
A
of
ag
of
e
er
ag
e
Av
ag
er
er
Av
Av
Graph 1:
Cartridge Page Yield
Page 4 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Cartridge Reliability
None of the Original HP inkjet print cartridges tested were dead-on-arrival or premature
failure (as defined in Appendix 2 鈥? Definitions), whereas, on average 20.9% of the refill
cartridges tested were either dead-on-arrival or failed prematurely.
More specifically, none of the Original HP inkjet print cartridges tested were dead-on-arrival
or premature failure (as defined in Appendix 2 鈥? Definitions), whereas 15.3% of the
store/refill brand cartridges (brands A-H) tested were either dead-on-arrival or failed
prematurely, and 28.7% of cartridges refilled at refill stations (brands I-M) were either
dead-on-arrival or failed prematurely.
Dead-On- Premature Problem
Arrival Failures Cartridges
Number of
Cartridges
Brand Type Brand Tested No. % No. % No. %
HP HP 68 0 0.0% 0 0.0% 0 0.0%
Store/Refill Brand Brand A 68 4 5.9% 2 2.9% 6 8.8%
Store/Refill Brand Brand B 68 1 1.5% 3 4.4% 4 5.9%
Store/Refill Brand Brand C 68 4 5.9% 11 16.2% 15 22.1%
Store/Refill Brand Brand D 68 2 2.9% 4 5.9% 6 8.8%
Store/Refill Brand Brand E 68 3 4.4% 16 23.5% 19 27.9%
Store/Refill Brand Brand F 34 1 2.9% 2 5.9% 3 8.8%
Store/Refill Brand Brand G 35 6 17.1% 6 17.1% 12 34.3%
Store/Refill Brand Brand H 34 0 0.0% 3 8.8% 3 8.8%
Refill Station Brand I 68 8 11.8% 6 8.8% 14 20.6%
Refill Station Brand J 72 12 16.7% 12 16.7% 24 33.3%
Refill Station Brand K 68 9 13.2% 9 13.2% 18 26.5%
Refill Station Brand L 68 13 19.1% 2 2.9% 15 22.1%
Refill Station Brand M 45 13 28.9% 8 17.8% 21 46.7%
Sum/Average of Store/Refill
Brands 443 21 4.7% 47 10.6% 68 15.3%
Sum/Average of Refill Stations 321 55 17.1% 37 11.5% 92 28.7%
Sum/Average of All Refilled
Cartridges Tested 764 76 9.9% 84 11.0% 160 20.9%
Table 2:
Cartridge Reliability
The dead-on-arrival and premature failure cartridges are listed separately for each brand
and also combined to create a total problem cartridge value for each brand. The
percentages shown are referenced to the number of cartridges tested per brand. These
values were then combined to create a sum/average value for the store/refill brands, refill
stations, and all refilled cartridges tested. For example, Brand A had a total of 68 cartridges
in the test. Four cartridges (5.9%) were found to be dead-on-arrival, and two cartridges
(2.9%) were found to fail prematurely for a total of six (8.8%) problem cartridges in the
test for Brand A.
Page 5 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Cartridge Reliability Problems
(DOA's and Premature Failures)
35.0%
28.7%
30.0%
Problem Cartridges
25.0% 20.9%
20.0%
15.3%
15.0%
10.0%
5.0%
0.0%
0.0%
HP Average of Average of Refill Average of All Refilled
Store/Refill Brands Stations Cartridges Tested
Graph 2:
Cartridge Reliability Problems
Page 6 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Appendix 1: Test Methodology
The following is a summary of the methodology used for this study:
The printers and print cartridges selected for this study were as follows:
Printer Black Cartridge Color Cartridge
HP DeskJet 5650 (C6490A) HP 56 (C6656AN) HP 57 (C6657AN)
HP DeskJet 6940 (C8970A) HP 96 (C8767W) HP 97 (C9363W)
A total of 762 refilled print cartridges and 68 HP print cartridges were tested using a total of
42 HP DeskJet 5650 printers and 30 HP DeskJet 6940 printers.
Printing was performed in a continuous mode in a controlled environment using the five-
page test suite from ISO/IEC 24712, and the environmental conditions specified in ISO/IEC
24711. To account for reliability-driven cartridge issues, defective cartridges were included
in the page yield calculation. Consequently, the reported page yield numbers are not based
on the ISO/IEC 24711 standard since ISO/IEC 24711 requires that defective cartridges are
excluded from the page yield calculation.
Test pages were as follows:
Page 1 鈥? Business Page 2 - Page 3 - Page 4 - Slide Page 5 -
Letter Spreadsheet Newsletter Diagnostic
QualityLogic procured all printers, paper, and HP print cartridges through standard retail
channels. Refilled print cartridges were obtained through multiple retail channels in multiple
cities, where possible, or directly from the manufacturer. For the refill station brands tested,
approximately 50% of the test data is based on cartridges that were refilled once,
approximately 30% is based on cartridges that were refilled twice and approximately 20% is
based on cartridges that were refilled three times.
QualityLogic selected Xerox Commercial 4200 plain paper (8陆 x 11, 20 lb., 92 Brightness)
for all printing for this study.
Cartridge transportation policies were designed to replicate customer behavior. Cartridges
were shipped via ground shipment or air shipment utilizing air-tight, pressurized containers
to protect against transportation-induced reliability problems.
Printer settings were left at the factory default. Driver quality settings were set to Normal
and plain media.
Each cartridge was inspected for leaks or other damage upon entering the test. A cartridge
with substantial visible ink spilled in the bag or on the cartridge was declared DOA due to
the leak. All other cartridges were printed to End-of-Life (EOL) as defined in Appendix 2:
Definitions.
Page 7 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Printing continued until at least 17 black and 17 color cartridges successfully reached EOL
for each printer model tested. Color and black cartridges were tested in parallel. As 17 color
or black cartridges were completed for a brand, HP original cartridges were used to
complete the rest of the brand. All results and effects of these HP original cartridges were
ignored in the test.
Testing was conducted to align with the sample size requirements in ISO/IEC 24711, which
requires a minimum of three physical cartridges to be tested in each of three printers. To
ensure a statistically significant result, testing continued until a minimum of nine of each
cartridge arrived at EOL due to Fade.
The test methodology for this reliability comparison study was developed by Hewlett-
Packard and implemented by QualityLogic.
Page 8 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Appendix 2: Definitions
Test Project Terminology Definition
Substantial ink leakage - Ink visibly spilled in the plastic bag containing the cartridge.
- Ink visibly spilled in the interior of the cartridge packaging.
- Ink visibly spilled over the print head nozzles.
End-of-Life (EOL) A condition determined by one of three mechanisms:
1. Fade has occurred on the diagnostic page.
2. Streak removal procedure steps have been exhausted.
3. Cartridge is Dead-on-Arrival (DOA).
Fade A significant decrease in density on the bands or blocks of the last page in
the test page suite, which is a diagnostic page. This decrease in density
does not have to occur completely across the page to be considered fade.
For a comparison to determine if fade is occurring, reference the 10th page
printed by that printer.
Page 9 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Streaks Very thin lines of color or the lack of color where it should be, in the blocks
surrounding the edge of the diagnostic page. Streaks differ from fade in the
width and severity of the reduction in density. Streaks can appear due to a
number of reasons, including thermal issues and clogged nozzles.
Black Streaks Color Streaks
Streak Removal Procedures This is the cartridge cleaning procedure (servicing) used to restore print
performance. If streaks are observed on three consecutive diagnostic
pages, a streak removal procedure should be implemented. Streak removal
operations should be conducted according to the HP printer manual
documentation. If there were additional cleaning steps advised for the non-
HP cartridges, they were included within the cleaning process. Use of a
鈥渓ight鈥? and 鈥渟trong鈥? cleaning procedure counts as one cartridge cleaning
operation. Cleaning is verified by reprinting the diagnostic page. If streaks
are still present, the clean procedure is repeated. Any pages printed during
the cleaning operation are not counted in the yield calculation. Due to the
significant amount of ink that is used for cleaning, the maximum
permissible number of times that the streak removal operation can be used
on a given cartridge is three times. Cartridges which require a fourth
service are considered to be at EOL.
Individual Cartridge Yield Individual cartridge yield is calculated by counting the number of diagnostic
pages printed between cartridge installation and EOL, then multiplying by
five. The diagnostic page is the last plot printed in the test suite.
Page 10 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Average Cartridge Page Yield Average cartridge yield for a given cartridge type
Where
xi is the individual cartridge yield.
鈥?
n is the sample size i.e.( 17 for example), excluding cartridges
鈥?
identified as 鈥淒ead-on-Arrival鈥? as per definition above.
This average cartridge page yield number is solely used for the purposes of
calculating premature failures.
90% Lower Confidence Bound Page Calculated as:
Yield
Where
t伪,n-1 Can be found on a Students鈥? t-Distribution Table with n -1 degrees of
freedom (df or 鈥樜解??) and an 伪 of 0.1. This provides a 2-tailed confidence
interval with 90% confidence. A different sample size and/or different
confidence interval will yield a different t伪,n-1.
The 90% lower confidence bound value means that one can be 90%
confident that the true mean page yield is equal to or greater than the
value of the lower bound of the confidence interval.
The page yield values used for this calculation include cartridges identified
as dead-on-arrival and premature failure.
Normalized 90% Lower Confidence Converted 90% Lower Confidence Bound Page Yield into index values by
Bound Page Yield setting the 90% Lower Confidence Bound page yield value for the HP brand
to 100% and the 90% Lower Confidence Bound page yield value for the
refill brands tested as a percentage value relative to HP.
Example:
90% Lower Confidence Bound page yield HP = 800
90% Lower Confidence Bound page yield refill brand A = 600
Normalized 90% Lower Confidence Bound Page Yield HP = 100.0%
Normalized 90% Lower Confidence Bound Page Yield refill brand A = 75.0%
Dead On Arrival A condition determined by one of three mechanisms:
1. Cartridge found to have substantial leakage (as defined above) at start
or during testing.
2. 50 or less pages printed by a cartridge.
3. Color mix (as defined below) has occurred and incorrect colors printed
on 10 consecutive pages 鈥? color cartridge failure mechanism only.
Page 11 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
Reliability Comparison Study
Color Mix Defined as a color cartridge that cannot print the correct Cyan, Magenta
and Yellow colors as shown on the diagnostic page 5 of the page yield test
suite. Ink has mixed in an unintended manner inside the cartridge and has
caused a discoloring of the ink.
An example of Color Mix is provided below. Compare the colored blocks in
the correct example to those of the color mix page.
Correct Diagnostic Page Color Mix page
Test Page Suite A series of five pages that are printed consecutively in order as a single job,
ending with a diagnostic page. ISO/IEC 24712.
Premature Failure A cartridge that has a page yield of less than 75% of the average page
yield (as defined above) for that cartridge brand.
Problem Cartridge Cartridges identified as dead-on-arrival and cartridges identified as
premature failures (as defined above).
Page 12 of 12
Test results provided by QualityLogic.
Tests were performed under laboratory conditions and your results may vary.
|