Fuel Pump Upgrade Guide
for the Mitsubishi 3000GT VR4 and
Dodge Stealth R/T Twin Turbo

by Jeff Lucius

How to Select
Bosch Fuel Pumps
Walbro Fuel Pumps
Nissan Fuel Pumps
Denso Fuel Pumps



Denso fuel pump diagram The solution for more power out of the 3000GT/Stealth (3S) turbo models, in its simplest form, is just to add more air (increase boost) and add more fuel. After a point, in order to add more fuel, a larger-capacity-than-stock fuel pump is required, along with larger injectors and a fuel-mixture controller. The fuel pump must be able to supply the amount of fuel required by all the injectors in wide open throttle (WOT) conditions at the maximum boost level anticipated. A 20% safety margin (that is, the pump can supply 20% more fuel than is actually required) is often cited to avoid overheating the pump or a momentary lean mixture (and possible damage to pistons, valves, or rings).

The stock fuel pump used in 3S cars is a constant- or fixed-displacement electric pump that is submerged in the fuel tank. It is called a "wet type" pump because all of its components, including the internal DC motor and impeller-type pump, are in contact with the fuel. The very-high electrical resistance of gasoline (more than 1 megohm) prevents electrical shorts inside the pump. A fixed volume of fuel is delivered for every revolution of the pump. The amount of fuel discharged into the fuel line is determined by how fast the pump rotates. The pump will rotate faster if the supplied voltage is increased or if the fuel line pressure at the discharge port is reduced. If the pressure at the discharge point exceeds about 70 psi, a relief valve built into the factory 3000GT VR4 pump opens to reduce pressure inside the pump. There is also a spring-loaded check valve in the pump to preserve high pressure in the fuel line when the pump is stopped.

Fuel pumps supply fuel volume; they do not create pressure in the fuel lines. In a return-line system (like in the 3000GT and Stealth), the fuel pressure regulator restricts the return fuel flow in order to create pressure in the supply line. As the fuel supply line pressure increases, such as during boost conditions for forced-induction engines, the pump has to work harder (it actually rotates slower and the current draw increases) and so the volume flow decreases.

Flow rate is reported in gallons per hour (gph), liters per hour (lph; 1 gallon = 3.7854 L), or pounds per hour (lb/hr). The average density of gasoline is 690 to 760 g/L (or 5.76 to 6.34 lb/gal).

How to Select

Know Your Needs
In order to determine the size of the fuel pump you need, you must first decide on (or know) the injector size and the maximum boost you plan to run. For example, many people upgrade the stock 360 cc/min injectors to 550-cc/min injectors. Six 550 injectors theoretically will flow 198 lph at 100% injector duty cycle (IDC). A boost pressure of 25 psi would probably be considered by many to be the upper limit for most street engines. The fuel pump you need then must be able to supply at least 200 lph at 68 psi fuel line pressure (43 psi base line pressure for 3S turbo models plus 25 psi to compensate for boost pressure in the plenum) at the voltage the pump is receiving. The voltage the pump receives is critical because flow can vary by 5 to 80 percent per volt depending on the particular pump, the voltage level, and the line pressure.

Get the Ratings
To determine if a fuel pump can meet your needs, you will need either a flow chart (see the examples below) or a flow rating for the pump. A flow chart is the best choice because it shows the measured flow rate over a range of line pressures, usually in increments of either 5 or 10 psi, for a particular supplied voltage. A flow rating cites the flow rate at one line pressure. For a flow rating to be useful it also must state the supplied voltage. If the voltage is not stated, it is probably fair to assume it is 13.5 volts. The flow rating pressure is usually 0 psi (also called the free-flow rating) or some integer multiple of bars. Three bars (43.5 psi) is a popular rated pressure for electric pumps, as is 5 bars (72.5 psi).

I need to mention here that there is an inconsistency in the flow rate information RC Engineering, Inc. (Torrance, CA; 310-320-2277) sends to the client in the form of an Excel spreadsheet. RCE notes that they use a test fluid with a 0.76 specific gravity, which is equivalent to a density of 760 grams/liter (g/L). During a flow test, RCE measures the volume flow with a flow meter. However, on the spreadsheet that RCE delivers to the client, the independent value is pounds per hour (Lbs/Hr). From this value, gallons per hour (Gal/Hr) are calculated. And from gallons per hour liters per hour (Ltr/Hr) are calculated. In every spreadsheet I have seen, RCE uses a density for gasoline of 6 pounds/gal for the conversion from Lbs/Hr to Gal/Hr. However, the test fluid with a density of 760 g/L actually has a density of 6.34 pounds/gal. It is unclear how RCE arrives at the Lbs/Hr figure. Their response to my email questions on this matter did not address this topic. In my opinion, the result is all of RCE's flow charts and spreadsheets overestimate the Gal/Hr and Ltr/Hr flow of the tested pumps by about 5.5 percent. I have adjusted all the spreadsheets available at my web site, and the charts shown on this web page, to reflect the correct conversion factor from Lbs/Hr to Gal/Hr and Ltr/Hr. Note that charts and spreadsheets available from my web site before March 20, 2004 do not include this correction. These corrected results compare favorably with other independent fuel pump flow tests. You can re-adjust the values in the Excel spreadsheets if you like by substituting 6.0 for 6.34 as the conversion factor.

Check the Current
Pumps are usually tested and rated at 12 or 13.5 volts. Unfortunately, substantially less than 12 volts may be reaching the fuel pump. I have measured current at the 10.2 to 10.5 volt range at the fuel pump at 15 psi boost as shown on my web page 2-fuelpumpvoltage.htm. The cause of this is the relatively small gauge wire used in the fuel pump circuit, the inclusion of the engine control relay in the circuit, and the use of an additional relay and resistor in the turbo models to reduce voltage to the pump at idle and low-load cruising. Sometimes re-wiring the connector at the resistor will restore voltage to the 11.5 to 12 volt range (for example, see John Moore's web page http://moojohn.com/fuelpump/ or my instructions at 2-fuelpumprelaybypass.htm). Road///Race Engineering has some tips for increasing the voltage to the fuel pump at their web page http://www.roadraceengineering.com/fuelpumptechtip.htm.

If the resistor by-pass mod does not restore sufficient voltage then the engine control relay can be by-passed and a new thicker positive cable can be run from the battery to the pump (including a new fuse and relay). Examples of this modification include my web page 2-fuelpump-rewire.htm and Bryan Roger's web page 2-fuelpump-hotwire.htm.

Erik Gross offers a different and much easier method to increase the current to the fuel pump on his web page http://www.team3s.com/~egross/3S/Mods/TT/FPRewire/index.html. Erik shows how to provide current directly to the fuel pump relay for the resistor, rather than running a new wire to the fuel pump. Erik's method provides increased voltage to the pump when the resistor is by-passed by the ECU, yet maintains low voltage to the pump during idle and low load cruising.

Another method to increase voltage to the fuel pump is to use one of the following "step up" voltage regulators, which would include re-wiring the fuel pump circuit similar to the methods described above.

If You Still Need More
If you need more fuel than one pump can supply, then I have been told that two 50-mm Denso pumps (like the VR4 and Supra pumps) will fit in the tank if the opening is notched 1/8" or so. Because of the heavy current draw (above 16 amps for each pump), you must install another circuit to supply power for the second pump. Two Denso pumps on the same circuit will lower voltage far below 12 volts and have unacceptably lower pump flow. The Walbro fuel pumps are narrower than the Denso pumps listed here and two will easily fit side-by-side through the tank opening and on the assembly. Walbro fuel pumps also draw about half as much current as Denso pumps. The Bosch and Nissan fuel pumps have a larger diameter than the Denso pumps.

The diameter of the fuel supply lines may restrict the amount and pressure of fuel that can be delivered regardless of pump capabilities. There is always some pressure loss in the lines, filter, and rails. For those of you interested in increasing the supply line size as well as the fuel pump capacity, remember that volume flow in smooth round pipes (with turbulent flow like in fuel lines) is proportional to the 2.5-power of the fuel-line radius increase. So a 33 percent increase in the fuel pipe diameter will double the volume flow (1.3332.5 = 2.05). The stock high-pressure supply line is a -5 AN pipe (5/16" outside diameter). A -8 AN supply line is sometimes recommended to support 550 bhp. Bear in mind though, that many recommendations come from the non-turbo hot-rodders that use low fuel-line pressures and low-pressure fuel pumps. The loss of few psi due to small fuel lines affects low-pressure systems much more than it affects high-pressure systems like ours. Our high-pressure pumps just need to work a little harder to maintain correct line pressure at the injectors.

Pick a Pump
There are basically four fuel pump manufacturers to consider when upgrading the in-tank fuel pump: Bosch, Denso, Nissan, and Walbro. If you are considering eliminating the in-tank pump in favor of an external pump, there are other choices for manufacturers, as well as much larger-capacity pumps available. Weldon Racing Pumps and SX Performance are examples.

This web page focuses on in-tank replacements. Only Denso makes a direct "drop-in" replacement for the stock pump (a 50-mm diameter pump made by Denso). The Bosch, Nissan, and Walbro pumps require modifications to the fuel pump assembly.

For fuel pump replacement instructions, check out my web page 2-fuelpump.htm. Please note that the o-ring that seals the fuel pump discharge port into the fuel pickup tube can be damaged during installation and cause starting and drivability problems. Mitsubishi does not sell this o-ring but the following one is a replacement: Borg-Warner fuel injection seal part # 274571 which has NAPA cross referenced part #2-12093, seal kit. To monitor fuel line pressure, I have fuel pressure sensor installation instructions at 2-fp_install.htm.

Bosch Fuel Pumps

Bosch makes excellent fuel pumps. Porsche uses them in the 928 models. Bosch pumps don't flow much at 43 psi line pressure, but at very high line pressures they tend to flow better than other pumps. However, for our cars they are still inadequate for injectors larger than 450 cc/min. Considering that Bosch pump model 10208 is 2.45" in diameter and 7.5" long, which about 0.5" wider and 2" longer than our stock pump, and requires extensive modifications to the pump assembly to make it work, there are better choices that cost less, flow more, and drop right in. The only advantage the Bosch 10208 has over the other selections here is that it flows 158 lph at 100 psi line pressure, far better than any of the others. For more information about Bosch pumps see the links below.
The picture below is adapted from http://www.theoldone.com/.

Bosch pump 10208

Walbro Fuel Pumps

Some of the Walbro fuel pumps work in the 3000GT. There are basically three models of Walbro pumps. For each of these three models there are three "sub-models" that differ only in the fuel line connection. The Walbro GSS242, GSS250, and GSS278 are 190 lph models. The Walbro GSS307, GSS315, and GSS317 are 255 lph models. The Walbro GSS340, GSS341, and GSS342 are 255 lph high-pressure models. The Walbro GSS341 model (also called the 255 lph HP) would be the best Walbro model for the 3000GT/Stealth.

The Walbro pumps are inexpensive, only about $100 to $150. Both the 315 and 341 models will work if 450 cc/min injectors are used and about 12 volts are supplied to the pump. The 341 model is suitable for use with 550 cc/min injectors in the VR4 if voltage to the pump is maintained at 13.5 volts or more. Walbro models 315 and 341 flow similar amounts of fuel up to about 55 psi line pressure (12 psi boost). A change in the relief valve design of the 341 permits better flow than the 315 at higher pressures. The reason for this is that relief valves often allow some amount of fuel to pass below the cracking pressure. The 341 model uses a relief valve that seeps less fuel at pressures above 60 psi than the 315 model's relief valve.

The Walbro GSS341 and GSS315 fuel pump flow data in this chart are from the links below. The data for the fuel pump labeled as Walbro GSS341 T&HTP are from the April 2004 issue of Turbo & High-Tech Performance magazine.

EP Racing Walbro GSS341 kit
Walbro fuel pump flow vs. line pressure

On the flow chart above, and the others on this web page, I note the theoretical maximum-flow capabilities of popular fuel injectors for the VR4. Please note that most injector manufacturers recommend a maximum, continuous IDC (injector duty cycle) of 80 percent, and that peak injector output often occurs at only 90 to 95 percent IDC. If you click on the chart above, it will appear in a new window. Click on this or other charts in succession to facilitate easy comparison.

Nissan Fuel Pumps

The Nissan Skyline fuel pump has held a legendary position as the ultimate upgrade fuel pump for the VR4, reportedly flowing 310 lph at 43 psi. Unfortunately, there are no hard data to support this. Nevertheless, some of the Skyline fuel pumps (R33 stock, and HKS and A'PEXi aftermarket) would make excellent upgrade choices for the 3000GT VR4 and Stealth TT.

From what I have read at various web sites, the R32 Skyline came with a 190 lph pump, and the R33 Skyline had a 195 lph pump. I am not sure about the current R34 model. In the April 2004 issue of Turbo & High-Tech Performance magazine, a stock R33 Skyline GT-R fuel pump was tested to flow 252 lph at 43.5 psi and 13.5 volts.

Both Mines and HKS sell upgrade pumps for the Skyline models that flow "only" about 270 lph. The HKS fuel pump (part number 1407-RN019) was also tested in the April 2004 issue of Turbo & High-Tech Performance magazine and it flowed 268 lph at 43 psi and 13.5 volts. The HKS 1407-RN019 and the R33 Skyline fuel pumps tested in T&HTP look just like the 300ZX Turbo fuel pump pictured below.

A'PEXi sells an upgrade fuel pump for the BNR32 and BNR33 Nissan Skyline GT-R. Max Cooper, http://www.maxcooper.com/, published results of flow tests by RC Engineering for the BNR32 upgrade fuel pump (A'PEXi part number 404-A011) at his web site. This pump outflows any other in-tank fuel pump for 3S cars that has been tested (that I am aware of). With a tested flow of about 300 lph (adjusted for the 6.34 density of the test fluid) at 43 psi and 13.5 volts, six injectors up to 720 cc/min can be supported by this pump. In their 2003 online catalog, A'PEXi shows a list price of $499 for this pump and the one for the BNR33 (part number 404-A010). I have seen prices for this pump at internet stores for between $359.20 (http://www.autocarparts.com/) and $410.16 (http://www.overboost.com/). The picture below is from Max Cooper's web page devoted to this fuel pump.

A'PEXi BNR32 fuel pump

The Nissan 300ZX Turbo fuel pump (part number 17042-40P05) has been tested by RC Engineering for George Balejian (aka ViperPilot99 on 3SI) to flow about 241 lph (adjusted for the 6.34 density of the test fluid) at 43 psi with 13.5 volts. The results (see the table in the Summary and the chart below) show this pump flows like the Cosmo and Supra Turbo fuel pumps, depending on fuel line pressure. George contributed the pictures below of his installation of the Nissan 300ZX pump. Some modification to the fuel pump assembly is required. Dealer list price is about $300 for this fuel pump.

Nissan 300ZX/Skyline pump 1

Nissan 300ZX/Skyline pump 2

Nissan 300ZX/Skyline pump 3

Nissan fuel pump flow vs. line pressure

Denso Fuel Pumps

Denso fuel pumps that have a 50-mm diameter and have part numbers that start with 195130 are direct "drop-in" replacements for the stock pump in all 3S turbo cars and 1st-generation DSM cars (1989-1994). Second-generation DSM cars (1995-1998) and all non-turbo 3S models use a 38-mm diameter Denso pump. My web page 2-fuelpumps.htm compares the turbo and non-turbo 3S fuel pump assemblies.

The chart below shows the flow test results for the stock Mitsubishi 3000GT VR4 (Denso 195130-0810), the Mazda Cosmo 20B (Denso 195130-0771; special thanks to Paul B. Prentis, Jr. for this data), the Toyota Supra Turbo MKIV (Denso 195130-1020), and the Mazda RX-7 FDS3 (probably Denso 195130-0782) and HKS 1407-001US (a modified Denso 195130-0771) fuel pumps. The 3000GT VR4 was used in my '92 Stealth TT for about 56,000 miles. There were about 2,000 miles on the Supra pump. I'm not sure how many miles were on the Cosmo pump. RC Engineering, Inc. (Torrance, CA; 310-320-2277) performed these tests using a pump dynamometer (this service costs about $50 per test). Testing can also be performed by Kinsler Fuel Injection, Inc. (Troy, MI; 248-362-1145). The complete RC Engineering flow test data (with the correct conversion factor of 6.34 lb/gal for the test fluid) are reproduced in Excel spreadsheets: misc/fuelpumpcomparo.xls and misc/fuelpumptests_t&htp_4-04.xls.

Denso fuel pump flow vs. line pressure

The chart above points out the limits of the stock VR4 fuel pump (at least a used one with 56000 miles on it) with the stock VR4 360 cc/min injectors. Above 15 psi boost (58 psi line pressure) the pump operating at 13.5 V cannot supply enough fuel for the injectors. At 12 volts the pump runs out of capacity at lower boost. Boost pressures of 12 to 14 psi are widely considered the safe limit (if detonation is controlled) for a 3000GT VR4 or Stealth TT with a stock fuel system in good shape and the factory turbos. I have instructions for measuring voltage at the fuel pump on my web page 2-fuelpumpvoltage.htm.

My old stock VR4 pump tested 17 lph less than what the stock pump is rated at, which is 180 lph at 43 psi at 13.5 volts. However, I have never seen Mitsubishi or Denso documentation for this. This may be due to the fact that electric fuel pumps run constantly. After 56000 miles and 8 years of service the armature bushings, brushes, commutator, pump vanes, and the rollers or gears have worn causing a loss of flow. Reduced flow from older pumps is something you should consider when increasing boost or upgrading injectors.

As I have mentioned, voltage to the pump is critical in determining how much fuel is discharged. John Cribb, a TT Supra owner, had his new Toyota Supra MKIV fuel pump (Denso 195130-1020) flow tested at Kinsler Fuel Injection, Inc. at five different voltages. Kinsler measured the flow and current draw (amperage) at each of these voltages from 40 to 100 psi line pressure. The test fluid specific gravity was 0.79 (or 6.6 lb/gal). John was kind enough to share the results with us. They are available in an Excel spreadsheet, misc/jcribb-supra-fp-test.xls. I added the LPH column and made minor changes to the structure of the original spreadsheet. The data have not been changed. Many thanks to John Cribb for sharing this data.

In the chart below, I plotted the volume flow versus fuel line pressure for the five voltages John's Supra fuel pump was tested at. For reference, I included the flow-test results for my used stock VR4 pump at 13.5 volts. The flow rates for John's Supra pump are slightly lower than those for my Supra pump. This may be attributed to one or more of the following factors: a difference in testing methods (RCE vs. KFI), a difference in test fluid specific gravity (less dense fluids may pump easier), and Denso production variations.

Nevertheless, one of the most interesting features of this data is the tremendous change in output at lower voltage levels. The change in output from 9 volts to 12 volts at 40, 50, and 60 psi line pressure is 70%, 200%, and 236%, respectively. This would be about 23%, 67%, and 78%, respectively, per volt change in output! Obviously, this fuel pump responds well to small increases in voltage. Output increases are less dramatic at levels above 14 volts, but still significant. This type of variation with supplied voltage may be typical for other high-flow Denso pumps. We won't know until further testing is performed.

Because the Supra pump flow varies so much with supplied voltage, the actual voltage at the pump determines the injector sizes that are appropriate to use with the pump. If six injectors are used and if the injectors can actually flow at their rated output at 100% injector duty cycle (always open), then the general recommendations below for injectors sizes are suggested for different supplied voltages. These suggestions would apply also to the Walbro GSS341 fuel pump.

Supra Turbo MK IV fuel pump flow vs. line pressure

Mach V sells (or use to) the Buschur Racing Upgrade Pump that is a direct drop-in replacement with a claimed capacity of 369 lb/hr at 70 psi line pressure, or about 231 liters (61 gallons) per hour at 70 psi (there is no mention of supply voltage). This pump must be the Supra Turbo fuel pump or a modifed Denso fuel pump.

The Cosmo pump would be good for 550 cc/min injectors only if 13.5 volts or more are supplied to the pump. Otherwise, it is safe to use with 450 cc/min injectors. The Cosmo pump, or Denso 195130-0771, is modified and sold by HKS with their part number 1407-001US. For many years this pump was advertised by retailers with a rating of "340 lph, 165 psi" or "90 gph, 165 psi". I am not sure what that means. HKS now advertises this pump on their web page http://www.hksusa.com/products/?id=717 with a rating of 71 gph at 45 psi (no voltage listed) and a suggest retail price of $630 (wow!). From the RCE pump tests, it looks like the relief valve for the factory Cosmo pump is set at about 90 psi, so I'm not sure what the "165 psi" in the old advertisements meant. In the Summary section below you can see new (April 2004) RCE flow tests of the HKS 1407-001US fuel pump, which show this HKS pump does flow what it is advertised, 269 lph at 45 psi line pressure. This Denso pump still has the "Cosmo" yellow tag but it flows almost exactly what my "black tag" Supra Turbo Denso fuel pump flow. I would guess that HKS has modified a standard Denso 195130-0771 (Cosmo) pump to flow similar to the Denso 195130-1020 (Supra) pump.

The pictures below are of the pump that was sold to me as "same as the HKS 90-gph pump". The part numbers were ground off, but I installed it anyway in the Spring of 2000. On March 13, 2001 I removed the fuel pump from my tank and inspected the ground-off numbers with a 10-power hand lens. The remnants of "1020" can still be seen. This pump is the Toyota Supra Turbo fuel pump. I sent this pump and the old stock one to RC Engineering for flow testing and the results are shown above. The Supra Turbo pump can be purchased for about $200 (sometimes less) from places like Conicelli Parts Center. In retrospect, I never should have accepted a part that had the part numbers ground off, and will not in the future. On a humorous note, this pump actually performs slightly better than the "HKS" pump.

Denso 255 lph pump 1

Denso 255 lph pump 2

Denso 255 lph pump 3

Below are some pictures of the HKS fuel pump that I purchased. I returned it after seeing it was the Cosmo pump, which, if unmodified, flows much less than the Supra pump I already had. I cannot caution people enough to avoid the "HKS" pump when it is sold at the exorbitant price of $400 to $700. Recent tests (April 2004 issue of Turbo & High-Tech Performance magazine) show this "Cosmo" pump has been modifed to flow very similar to the standard Denso 195130-1020 Supra Turbo fuel pump, which can be purchased for less than $200.

HKS 340 lph pump 1

fuel pumps


Please note that flow values in the tables below are for 13.5 volts supplied at the pump. Voltage at the pump in your car may be only 10.5 to 12 volts, and so actual flow may be much less than the values indicated below.

Fuel Pumps for the 3000GT/Stealth
Name Part
Rated flow
lph @ 43psi
Rated flow
gph @ 43psi
Mitsubishi VR4 Denso
180 48 This pump also may be used in the Celica GT-4.
Measured at 163 lph @ 43 psi @ 13.5 V for a used pump.
Light blue or green tag on top.
Mitsubishi Evo8 Denso
191 51 Factory pump in the Lancer Evo8. Flow data from T&HTP.3
Design is like the 3000GT NA pump so an adapter would be
needed for the pickup tube and electrical connection on
TT models. An upgrade for NA cars.
Bosch Bosch
210 55 Flow data is from The Old One
Requires modifications to the pump assembly.
Mazda RX-7 Turbo Denso
223 59 Flow data from T&HTP.3
Road///Race Engineering says 210 lph @ 12 v.
Nissan 300ZX Turbo
 (Skyline the same?)
241 64 This flow data is from RC Engineering tests for
George Balejian.
Requires modification to pump assembly pickup tube.
Dealer list price is ~$300.
ABC Nissan (1-800-500-8722) may sell it at a
25% discount (~$227).
Mazda Cosmo 20B Denso
248 66 This flow data is from RC Engineering tests for Paul
Prentis Jr.
Road///Race Engineering reports 250 lph flow.
Yellow tag on top.
Nissan Skyline R33 Nissan
252 66 Stock pump.3
Requires modification to pump assembly pickup tube.
Walbro Walbro
GSS 341
262 69 This flow data from T&HTP.3 Flow data from
Auto Performance Engineering says only 255 lph.
The Walbro 341 can be purchased from Extreme PSI
for about $100. An adapter is required for the pickup tube
and electrical connection.
HKS Skyline R33 HKS
268 71 Relief valve opens near 90 psi.3
Requires modification to pump assembly pickup tube.
Mines Skyline Mines
270 71 Mines and their retailers do not mention voltage or pressure.

272 72 I bought one of these pumps and the Denso part
number was 195130-0771, the "Cosmo" pump! I sent it back.
HKS used to claim "340 lph". Now they claim "71 gph
@ 45 psi"
(269 lph), just what was measured3.
HKS MSRP = $630!!!!!!
Recent tests3 show this yellow-tag Denso pump flows
very similar to the black-tag Denso below. I would guess
the pump is modified to perform like the 195130-1020.
Note that the picture on their web page is of an in-line pump
and not of the actual in-tank pump that they sell.
Toyota Supra Turbo Denso
277 73 260 lph @ 43 psi; 220 lph @ 58 psi; @ 12 V 1.
290 lph @ 43 psi; 260 lph @ 58 psi; @ 14 V 1.
Conicelli sells this pump for $180 ($244.71 list price).
Black tag on top.
300 79 An upgrade for the BNR32 Nissan Skyline GT-R
297 lph @ 40 psi @ 13.2 V 2.
290 lph @ 60 psi @ 13.2 V 2.
258 lph @ 70 psi @ 13.2 V 2.
Buy this pump for about $360 at http://www.autoparts.com/.
Requires modification to pump assembly pickup tube.
1 Measured by Road///Race Engineering.
2 Measured by RC Engineering.
3 Measured by RC Engineering. Data from Turbo & High-Tech Performance April 2004 issue

The table below uses data from the RC Engineering flow tests for the VR4, RX-7, Cosmo, 300ZX, Walbro 341, HKS, Supra, stock BNR33, and A'PEXi BNR32 fuel pumps. RCE tested my VR4/TT fuel pump (used for 56,000 miles) and my Supra MKIV pump (used for 2,000 miles). I thank Paul Prentis, Jr. for sharing the test results for his Cosmo pump, George Balejian for sharing the test results for his Nissan 300ZX pump, Turbo & High-Tech Peformance magazine (April 2004 issue) for the Mazda RX-7, Walbro 341, HKS 1407-001US, and stock BNR33 results, and Max Cooper for publishing the flow test results for his A'PEXi BNR32 fuel pump. Data for the Bosch pump were collected from the internet. The complete data are available in Excel spreadsheets: misc/fuelpumpcomparo.xls and misc/fuelpumptests_t&htp_4-04.xls. The 300ZX, Walbro 341, HKS 1407-001US, and Supra Turbo pumps can all supply sufficient fuel to support six 610 cc/min injectors up to 70 psi line pressure (27 psi boost) when 13.5 volts are supplied. The stock R33 Skyline GT-R fuel pump can supply sufficient fuel for six 640 cc/min injectors up to 70 psi line pressure with 13.5 volts supplied. Only the A'PEXi BNR32 pump is capable of providing adequate fuel to support six 720 cc/min injectors up to 70 psi line pressure at only 13.5 volts.

Fuel Pump Flow Tests Comparisons
Pump VR4 used
Mazda RX-7
Mazda Cosmo
Supra MKIV
R33 Skyline
GT-R stock
Volts 13.5 13.5? 13.5 13.8 13.5 13.5 13.5 13.5 13.5 13.2
PSIG Tested Flow in liters per hour (lph)
35 175.54 - 240.62 266.90 249.58 274.06 286.60 286.60 253.16 302.12
40 167.18 ~213 229.28 254.36 243.01 265.10 277.64 281.23 259.73 297.35
45 159.42 - 219.13 244.21 240.03 259.13 268.69 273.46 246.59 294.96
50 149.87 201.90 209.58 239.43 236.44 251.37 256.75 265.10 244.80 294.36
55 140.91 - 199.43 225.70 230.47 244.21 250.18 256.75 240.03 294.36
60 128.37 ~195 191.66 217.34 228.09 238.24 237.04 248.98 240.03 289.58
65 119.42 - 177.93 205.4 223.91 231.07 228.09 234.06 235.85 276.45
70 96.73 189.2 164.79 193.45 219.13 223.31 213.76 219.13 232.86 258.54
75 - - 155.24 171.96 202.41 216.74 200.62 197.63 228.09 240.03
80 - ~179 138.52 154.64 186.29 191.66 185.10 184.50 202.41 -
85 - - 117.63 131.36 167.18 179.72 167.78 164.20 180.92 -
90 - ~168 100.91 111.06 135.54 162.41 157.63 152.85 11.34 -
95 - - 77.62 - 120.61 149.27 134.34 127.18 - -
100 - 157.7 58.51 - - 128.37 115.83 - - -

Comparison of VR4, Supra, and Walbro fuel pumps

In their April 2004 issue, Turbo and High-Tech Performance magazine published flow test results for the nine fuel pumps noted below. RC (Russ Collins) Engineering tested these fuel pumps at 13.5 and 16.5 supplied volts with a fluid that had 0.76 specific gravity, which is at the heavy end of the range of typical gasoline densities, which is 690 to 760 grams per liter (g/L) or 5.76 to 6.34 pounds per gallon (lb/gal) depending on grade and additives. The chart below shows some of the results for these fuel pumps with the correction applied for a 6.34 lb/gal test fluid. Complete information for six of these fuel pumps (the ones that would be upgrades for a 3S car), including current draw values, is available in misc/fuelpumptests_t&htp_4-04.xls.

Comparison of a variety of fuel pumps


The A'PEXi BNR32 fuel pump 404-A011 (an upgrade pump for the Nissan Skyline GT-R) is the highest flowing in-tank fuel pump at line pressures up to 75 psi and at 13.5 volts. At present (February 2004) I am not aware that this pump has been installed in any 3S car. However, it is installed in some Mazda cars that use a factory Denso fuel pump nearly identical in size and design to the 3S factory Denso fuel pump. The relief valve opens at about 75 psi, limiting use to boost pressures up to 30 psi if base fuel pressure is set at 45 psi. With available flow measurements at only 13.2 volts, we do not know yet how this pump performs at lower or higher voltage. Nevertheless, if this pump is installed with a re-wired electrical circuit (providing at least 13.2 volts during WOT operation), six injectors up to 720 cc/min can be supported. List price in the 2003 A'PEXi online catalog is $499.00. I have seen it advertised at online stores for $360 to $410.

The factory R33 Skyline GT-R and HKS Skyline upgrade 1407-RN019 fuel pumps are the second highest flowing in-tank fuel pumps at line pressures up to 75 psi and at 13.5 volts. These fuel pumps may flow a little less than the other pumps below 65 psi, but this is not important when using large injectors. It is the flow at the high boost ranges when large injectors would be required that is important. These pumps can support six injectors up to 640 cc/min with 13.5 supplied volts.

The Denso fuel pump 195130-1020 (the Supra Turbo MKIV pump) is one of the best in-tank upgrade choices if you are using injectors up to 550 cc/min and are providing at least 13 volts to the pump. When 18 volts are supplied, the Supra pump is good for injectors up to 880 cc/min. At 13.5 supplied volts no other in-tank pump, except for the R33 Skyline GT-R and A'PEXi BNR32 fuel pumps, flows significantly more fuel up to 70 psi line pressure or 27 psi boost. It is a direct drop-in replacement; all other non-Denso choices require some modification to the fuel pump assembly. It is as quiet as the stock pump. Best of all, it can cost only $180 at Conicelli Toyota. One disadvantage of the Supra pump is its relatively heavy current draw (at 13.5 V, 16A @ 43 psi, 19A @ 70 psi) and large decrease in flow as supplied voltage lowers. Good, heavy-gauge wiring (meaning re-wiring the fuel pump electrical circuit) is a requirement to get the most out of this pump.

The Walbro GSS341 model (also called the 255 lph HP) flows a little less than the Supra Turbo pump up to about 67 psi line pressure (24 psi boost), when 13.5 volts are supplied to both pumps. However, at 12 supplied volts, the Walbro GSS341 outflows the Denso 195130-1020 above 55 psi line pressure (compare John Cribb's Supra pump data to the Walbro charts). Like the Supra fuel pump, the Walbro GSS341 is an excellent choice for 450 cc/min injectors, especially if the fuel pump has not been re-wired (12 supplied volts). Also like the Supra pump, it can be used with up to 610 cc/min injectors if 13.5 volts are supplied, and up to 720 cc/min injectors if at least 16 volts are supplied. The Walbro GSS341 draws about half as much current (at 13.5 V, 8A @ 43 psi, 10A @ 70 psi) as the Supra fuel pump. Some slight modification to the fuel pump assembly is required and it may not be as quiet as the Denso pumps. The price is even less than the Supra pump at $100 to $150.

Whatever pump you decide to go with, the Cosmo and 300ZX pumps are also good choices, the voltage at the pump is critical for good flow. Check the voltage at the pump. If it is below 11 volts try the resistor/relay by-pass modification first to see if voltage increases to 11.5 to 12 volts. For more voltage you will have to re-wire the circuit and maybe add a step-up voltage regulator.

If you have pump dynamometer measurements for any pump that fits the 3S cars please let me know the results and I will share them on this web page; send an email to jlucius at stealth316 dot com.

Back Home Forward

Except for the small gif and jpg images, the content, images, photographs, text, and multimedia displayed are Copyright © 2000-2004 by Jeff Lucius and K2 Software. All rights reserved. No part, section, image, photo, article, or whole of this site may be reposted or redisplayed without permission of the author.
Page last updated November 2, 2004.