Although I had the rear bumper debadged except for the main emblem, I added back the RX-7 logo in an updated, “Efini” form. These are stick-on and go on the right, passenger side of the bumper. As if this wasn’t cool enough, I wanted the more obscure, red lettering version.
The Mazda part number for this is F100-51-711B. Even down to an emblem, everything needs to be as authentic and quality as possible. While expensive, I was at least able to find one available.
There is another portion to this emblem that has a rectangular Efini badge underneath, but I decided to omit it.
Deciding to step up to the Revolution/AutoStaff “Radiator Outlet Tank” cost almost 3 weeks of downtime. I had to order this part from RHDJapan in the midst of the water pump service. Unfortunately, this was the last piece of the puzzle and I wasn’t able to finish everything up until the part was in hand.
After waiting 2 weeks with no shipping update from RHDJapan (they first have to receive the part from the manufacturer), I opted to pull some strings just to ensure I could get the car operational again. I have little patience when it comes to these types of things, which probably alludes to how I powered through the single turbo conversion in a few months. It really bugs me to leave a job unfinished.
So I settled for second best, as it was better than nothing at all. A member on the RX7Club forums actually made a small run of a similar product, inspired by the original. These were sold out before I could grab one, but I chased another avenue and it turned out a buddy of mine bought one and never installed it. I explained my situation and he was generously willing to pass it on to me. As I said before, the community is an imperative resource.
Of course, the very next day after all this happens, I get an email from RHDJapan noting that my order was shipped. Miraculously, according to the tracking number, the Revolution part was already in San Francisco and waiting for customs. It actually shipped 2 days prior, but I got the tracking number too late… talk about complicating things.
Revolution made the outlet tank with a zero pressure, screw on cap. Meaning, it was simply a fancier replacement of the stock thermostat housing and, like stock, a separate Air Separator Tank (AST) was required. I wanted to integrate the AST directly, so I had the screw on filler chopped off and a new billet, radiator filler welded in its place (thanks FFTEC, again). This would allow me to run a pressurized 13psi cap, effectively obviating the need for a separate AST.
For comparison’s sake, here’s a side by side of the Revolution tank (left) versus the locally made copy (right).
The scalloped section on the bottom allowed for perfect clearance with the idler pulley.
The difference in quality is easily recognizable. Yes, the Revolution part is prohibitively expensive and a hassle, but you truly get a well-made product. The entire flange is billet and even the outlet for the radiator hose is billet.
I also took the opportunity to refresh the thermostat with a new OEM replacement. It seems as though Mazda changed the design and build of the thermostats. The later versions are simpler and the jiggle pin is moved inward near the center opening.
Left: Old; Right: New
Finally, here is the Revolution tank installed with a 90* 1/8″ NPT to 1/4″ barb for the overflow nipple.
I think it adds a nice touch to the engine bay! A vast improvement over the ungainly stock counterpart.
Complimenting my OCD tendencies, I’ve typically held a strict policy of keeping my cars & bikes clean and sanitary. The reason why installs take me so long to do is because I have to scrub down all parts and fasteners before reassembly. I take pride in upkeep.
However, due to the amount of garage work that took place during the past few months, there was simply no point in preserving the cleanliness of the RX-7. By the end of it, a thick layer of dust and grime had accumulated. And to think a brand new paint job essentially laid underneath all of this…
Finally, with all the major wrenching complete, I could restore the car’s finish with a much needed detail. I’ve been looking forward to this part. Regardless of the repaint, swirl marks and micro scratches are inevitable if a car sits for any length of time without consistent maintenance. Properly cleaning a car is almost a science, and it’s shocking to realize just how finicky and temperamental paint is. All it takes is one bad wipe of a towel to introduce new marks in the clear coat.
To perform the job thoroughly, I picked up a Griot’s Garage Random Orbital with a set of Lake Country pads. My brand of choice for the detailing products is Wolfgang. This stuff is the real deal, forget the subpar consumer-grade offerings like Meguiar’s, Turtle Wax, Zymol, etc. I also grabbed a few Cobra Gold Plush microfiber towels.
I’ve been most looking forward to applying this wax… the Pinnacle Souveran. I’ve had this sitting on my shelf for half a year, and never got the chance to use it.
While detailing a car may be therapeutic in many ways, you can’t escape the fact that it requires a lot of effort.
I dedicated an entire day just to polish the car. I ended up doing 2 passes and it was extremely time consuming and slow work. Once the polishing was done, I applied sealant over the car (except the hood because it still had fresh paint, and I wanted to give it more time to outgas).
The sealant needed 12 hours to cure and I ran out of time last weekend to finish the detailing. I’m going to pick it back up tomorrow and top everything off with the Souveran wax. The results so far have been appealing… I’ll have to try and arrange a photoshoot once it’s all ready.
After investigating further, I was able to determine the true source of the odd coolant leak. I initially thought it was the lower radiator hose, so I replaced every single clamp with stainless t-bolt clamps. The leak still persisted. I eventually traced the leak to a region higher up – the bottom weep hole of the water pump. Thankfully, the increased visibility and accessibility from going single turbo enabled me to finally find the source. This is the same mystery leak I saw when the car was still in the stock, twin turbo configuration. With the undertray in place and all the clutter blocking any sort of reasonable access, my attempts at diagnosing the problem was greatly thwarted back then. (Albeit, the leak was less prominent at that time.)
I noticed that the leak would only occur during a cold-start and disappear once the car was fully warmed up. This is likely due to the seal inside the water pump expanding from the heat. Regardless, when the weep hole is leaking, it is a sure sign of a failing water pump.
As much as I wanted to lay down the wrench for a while and just drive the car, I knew the only way that could happen is if I postpone my “vacation”. There was no skirting around the issue, I took immediate action and obtained all the necessary parts for the water pump service by the week’s end.
Allow me to demonstrate the Grand Mighty approach to maintenance. If I’m forced to go through the effort of removing and replacing a part, I may as well upgrade it. So in addition to brand new gaskets and hardware (every fastener I needed to touch for this job was replaced), I elected to swap in a new RE-Medy water pump from MazMart. I got lucky when I called MazMart, because I needed the water pump quickly and they happened to only have 2 left in stock.
The first step was to remove the idler and water pump pulleys, remove the thermostat housing, and then remove the water pump. While a water pump service on a rotary may not be as involved as piston-engined counter parts, where the timing belt is interconnected with the pump, there is a catch… The water pump housing is sandwiched between the water pump and the front cover on studs. I don’t think there is a reliable way to remove JUST the water pump without compromising the water pump housing’s gasket. Maybe if the gasket remains in one piece and any movement is severely minimized, but there’s still no guarantee against disturbing the seal.
To be comprehensive, I moved forward with removing the water pump housing as well. And as anticipated, the paper gasket broke apart instantly and needed to be scraped off.
The RE-Medy water pump uses a brand new OEM core, but features a CNC’d impeller blade. This benefits from being more effective at moving water and is said to be void of cavitation up to 10,000 RPM.
Here is a comparison between the stock water pump (top) and the RE-Medy water pump (bottom).
And a closer look at the new impeller.
Replacing the water pump is overall a relatively straightforward job, and especially so with my setup because much of the clutter up-front was removed after going single turbo. The V-Mount also provided adequate workspace, but I did need to remove the studs holding on the water pump housing, because there wasn’t enough room to slide it fully off. However, I spread the work over 3 days. The most time-consuming part was cleaning everything.
I gradually sanded and basically polished all of the mating surfaces. There was slight pitting on the water pump housing to water pump’s surface, so I took the time to fill them in (darker colored areas in the picture below).
Once the parts were ready to go, I dressed all the mating surfaces with a light film of RTV for added security (in addition to new gaskets, of course), and buttoned it all back up.
I’m not quite finished, as you can see, because I’m waiting on a thermostat housing from Japan. I’m used to having parts shipped Next or 2nd Day when in the States, but the whole “overnight parts from Japan” thing isn’t quite as easy. Until then, I’ll have to wait and pick this back up later.
A couple weeks ago, I got out of classes early one day and made my way over to AutoRnD. I used to always get my alignments done at Auto Innovations in Milpitas, but Rishie at AutoRnD has a nice, new Hunter rack and I wanted to take advantage of it.
This was the first time in my experience when over an hour of labor time was spent, and when the alignment results came out exact across the board. Rishie was methodical and adjusted everything into spec. Since alignments are typically flat-rate jobs, most places simply do what they can and get within ballpark to the desired measurements, and then send you on your way.
I opted to run -2.5* of camber all-around with 1/16″ of toe up front.
After swapping in the manual steering rack and guesstimating the toe, the front alignment was initially way off and required broad adjustments. Unfortunately, once it was all complete, the front-end of the car actually dropped by almost 1/4″. This meant I had to go back home and bust out the coilover spanners to redo the ride height once more. Obviously, doing so also negated the perfect alignment I just received. I’ll eventually go back and get a second alignment.
With the single turbo conversion complete, I figured it’d be good to shed my insights and add closure. This was easily the most rigorous and difficult project that I have personally pursued. I know you can cut corners and convert a car to single turbo fast and cheap (~sub-$6k), but this is not acceptable to me. It became extremely time-consuming and involving to wade through a job of such magnitude and scope. There were literally hundreds of different items that required individualized attention.
It was an absolute frenzy making countless trips to Vic Hubbard‘s, Dale’s Hardware, and various other stores to seek out specific parts and fasteners. In the end, I’d estimate that the cumulative total time I’ve invested in achieving the single turbo conversion equates to around 700 hours. This includes all the researching, shopping, parts-getting, and labor. Albeit I was busy accumulating the necessary inventory of parts many months in advance, I did not start the physical disassembly of the stock setup until mid-December. I finished the conversion mid-March.
Depending on your perspective, my timeframe is either slow or relatively fast. During the past 3 months, the majority of my days were very long. I had to maintain a constant rhythm of work in order to complete the project in the time that I did.
As you can imagine, I am very relieved that I can finally take a breather. The deed is done. This project has provided serious relativity to me. “Normal” life now seems oddly slow-paced and easy-going. I can also appreciate anyone who has undertaken a similar task, because this wasn’t a small feat. I had to thoroughly learn the process from scratch by pouring over diagrams, seeking assistance from others, and conducting a bountiful amount of searching. You need to have steadfast perseverance and concentrated diligence. Simply said, an unmotivated or inept individual is not the right candidate.
Now, let’s summarize the objective-oriented achievements that underpin the effort exerted. It wasn’t a small decision to tear down a 25,000 miles stock engine and convert it to single turbo. But after careful consideration and deliberation, I knew it was a necessity.
As nice as the idea of sequential turbos may be, its benefits are heavily outweighed by its complexities. The sheer number of vacuum lines and solenoids required to make the system function is insane. Debugging an issue here would have been a nightmare.
My fundamental goal with this car is to simplify and modernize. Allow me to outline the factual points that substantiate the superiority of a single turbo setup:
- Immensely reduced complexity and fail points.
- There are now only 6-7 vacuum lines, which are ALL easily visible and accessible.
- The multitude of solenoids has been reduced to just one.
- Mechanical simplicity: Boost is controlled off of the wastegates’ springs, which is about as robust and reliable as you can get.
- It’s important to remember that the 3rd Generation RX-7 is now 20+ years old. This was an opportunity to bring core components out of antiquity.
- The turbocharger is now cutting-edge technology and features all of the latest features (double ball bearing, built-in heat sink so it’s only oil cooled, aerodynamic CEA, etc.)
- The fuel system is also modernized with high-grade Injector Dynamics injectors. Additionally, the Fuel Pulsation Dampener feature is now integrated into the Weldon Fuel Pressure Regulator. Before, the stock FPD was a major culprit of dangerously leaking fuel when old.
- The new turbo system is overbuilt and worked far below capacity.
- 12 psi is very minimal for the turbo. It is realistically efficient up to 30 psi. For reference, the stock turbos are only suitable to about 14 psi before they turn into nuclear hair dryers.
- The max injector duty I’ve seen so far has been around 67-69%.
- Overall increased accessibility. Everything is now so much more reachable and open. This greatly helps with maintenance, because you no longer have to tear down a million things that are in the way. I can put a wrench on every clamp on every radiator hose and intercooler coupler. Removing the UIM now takes under 10-minutes, whereas it took over 30 minutes before. The entire turbo system is now sitting in plain view – and not hidden under a layer of various intake and intercooler piping.
- Better flow, better cooling.
- The stock system was subject to finicky boost creep. With just the addition of a midpipe, you can experience overboost issues. Now, with the dual 44mm wastegates, boost creep is a thing of the past (I’m actually experiencing the opposite of creep).
- Greater flow and less restriction in the exhaust also aids in reducing under-hood heat.
- A V-Mount is a superior design and package. Although the weather is still cool and I have only done preliminary testing, water temperatures average around 85*. IAT’s are usually below 40*. This is a phenomenal leap in performance over stock.
- The stock system was subject to finicky boost creep. With just the addition of a midpipe, you can experience overboost issues. Now, with the dual 44mm wastegates, boost creep is a thing of the past (I’m actually experiencing the opposite of creep).
- Lighter weight. A single turbo setup is significantly reduced in weight and offers the best power to weight potential in a rotary application. I’d estimate about 40 lbs was saved from the conversion. A single turbo 13B is about the same weight as a N/A 20B.
- More power! And a far higher ceiling for power (if ever desired).
And there you have it! A stock sequential setup is great on paper, but does not translate as well in reality. Proponents of leaving a car like the RX-7 99-100% stock are living in the past, they are advocating an untapped platform. I like to approach the matter in a more pragmatic fashion.
As stock as my car’s exterior is, I am very content with it. ’99 front bumper, mud flaps, FEED spoiler… that’s all I wanted. Well, except for one more thing – the FEED hood. I have a soft spot for this hood’s design as I find it striking yet subtle. From certain angles, you can barely tell that it’s any different from a stock hood. The dual vents are highly reminiscent of the Jaguar XJ220. The fact that this kind of hood is available for the RX-7 made it a must-have, with skepticisms notwithstanding. I’m sure the alternative hoods that are more aggressively vented perform better, but this is one of the rare instances where I will favor form over function. The FEED’s vents will provide additional heat relief, which is sufficient for the cause.
Shine Auto Project used to make a FRP & Carbon version of the FEED hood, but no longer do so. Regardless, I’m not a fan of any composite material hoods. Even though the FEED hood is the only aftermarket variant I’d purposely install on my car, I still didn’t want to go aftermarket, per-say. I was keen on retaining exact OEM fitment and also obviating the need for hood latches. This meant I was left with one, more interesting option of “procuring” this hood – make my own hybrid. You gotta do what you gotta do…
I hit up my man Ken @ Shine and was able to convince him to make me a copy of just the dual vents. I then brought this over to William and had him work some magic into my stock hood. Needless to say, the molding and bodywork process was quite difficult and intricate ($$$). I asked William to take pictures along the way, check it out.
Here is what the vents look like placed over the stock hood, for reference.
In order to transpose the vent accurately into the hood, very precise and measured cuts needed to be made.
Inserted into position:
Next came the task of bonding everything together.
And then a lot of filling and smoothing.
The holes for the vents were then cut out. Only the center sections of the vents were cut because this was a balance of compromising structural rigidity. Furthermore, the framework of the stock hood actually impedes the vents on the side areas, so it wasn’t worth it.
The job took about 3 weeks to complete. Here is the hood painted and installed.
Looking good! And I’m glad to finally have a hood back on the car.
I will be taking the car and hood back to William again soon for touchups. The backsides of the vents will be painted flat black to add contrast. William is also going to add in vents and mesh material on the undersides to help increase the rigidity.
It was quite the momentous occasion when I mounted on the RegaMasters. I was confident in their fitment offset and width wise, because I have that down to a science with this car. The whole stepping down to 17″ thing, however, garnered some anticipation.
I’d reckon that the majority of RX-7s with wheels go for 18″. For the cars that did have 17″ wheels, I could hardly find any proper or good looking example. It’s important to note that wheels must work holistically with the car. All of the other variables must be thoroughly accounted for to pull off a successful setup: ride height, offset, width, tire choice (all-seasons or budget brands need not apply), etc.
There was no choice but to take the initiative on this one. Once I put the wheels on the car, any trace of apprehension or doubt went out the door – I took a step back and knew I was looking at my grail setup. I invested heavily to procure, piece together, and refinish this specific set of RegaMasters, so I’m delighted that it paid off. This has affirmed and solidified my thinking that 17″ wheels are more ideal for a stock body RX-7.
I did have to spend a weekend correcting the ride height, I ended up dropping all four corners slightly to a height of 24-3/4″ from floor to fender. This closed up the wheel gap and is what I think is optimal, it strikes a balance between form and function. The front end of the car was suspiciously higher than before after the single turbo conversion… especially the passenger (turbo) side.
Here’s one of the rear wheels. I love the period-correct and understated look of RegaMasters. I’m using a set of Project Kics R40 REVO lugnuts in Neo-Chro. I think this is the fourth time I’ve bought these particular lugnuts, they’ve become a staple of mine, more or less. I even tried looking for an alternative set, but wasn’t able to find anything as intriguing.
On one of the front wheels, the paint started to crack around the lugnut holes when tightened down. I had to bring this wheel back to Wheel Techniques to fix, for fear of the paint cracking further. They were able to sand and bevel the lugnut seat areas.
Finally, here’s a full picture of the RX-7 with the wheels mounted. I realize that I don’t have many shots of the car as a whole, most of what I have are up-close of parts. This will have to change once I can get the car detailed and finalized. If any competent photographer is reading this and wants to assist, let me know! I’ve become extremely lazy with pictures. As you can see in the below, I simply pulled the car out in front of my house and snapped away. Regarding the glare, I wasn’t patient enough to wait for the sun to go down more.
With the tuning over, the next step was to wrap up and polish out all the loose ends. I started by redoing my fan’s wiring.
I wanted the wiring to be routed off to the side of the front duct area, so as to not impede air flow or look out of place hanging in the middle. This meant it needed to be extended, which I carefully did prior. Once the splicing was complete, I left the connector dangling on the floor because I didn’t want to connect it until the car was ready to start. I was greeted with a pleasant surprise when I eventually did go to plug it in… the connector was chewed up.
Do not be fooled by Blue’s innocent looks or charming demeanor, an anarchist hides underneath that furry coat.
Apparently, one morning he was left in the garage alone and unattended for a while. I guess he got bored and crawled under the car to have a snack on my hard work. The fact that I couldn’t even repair the damage is the worst part. One of the wires was gnawed through right next to the base of the connector, which made re-splicing it nearly impossible.
I spent a week searching around and calling Rotary specific shops across the country to see if anyone had any broken stock fans, so I could pilfer the connector off. I ended up finding the connector I needed by randomly asking an RX-7 buddy of mine, who happened to have a spare fan setup in his closet (thanks again!).
With a replacement connector in hand, I used the sealed solder and crimp connectors to extend the wiring back. I had temporarily crimped what I could together just so the fan would still operate during the tune session. I finished it off with heatshrink and Painless mesh wrap to give it a professional look. (Don’t worry, all is forgiven with my dog, I can’t stay mad at him…)
Next, I flushed the coolant system again and picked up a Lisle funnel to refill.
I noticed that one of the lower radiator hoses developed a leak around the clamp area. I was using worm gear clamps on all the hoses, but decided to step them up to stainless T-bolt clamps from HPS. The larger the hose, the more relaxed the tolerances… thus needing beefier clamps.
In addition to greater clamping potential, the T-bolt clamps have a wider 3/4″ band which should leverage more surface area. Hopefully this will solve the leak issue.
Another item that was left unfinished was the cold side intercooler piping to compression elbow connection. The silicone coupler there ended up being slightly too short, so I had to pull it forward to reach. This left a gap on the elbow which was aesthetically incomplete. I called Flex Technologies (SiliconeHose.com) and requested a custom, manufactured part. I needed the same coupler to be 0.5″ longer at one end, and 0.25″ longer at the other. Flex Technologies was nice enough to fulfill my order within a week. Major kudos to them for coming through because these guys have important contracts with large OEM’s, and my single coupler was not a simple off-the-shelf affair.
New, lengthened coupler on the left and Old on the right.
Back in business.
Lastly, I picked up some pricey DEI fire sleeve for the rear wastegate’s vacuum line. A section of this line runs close to the turbine housing, so I wanted to double up the heat protection there for added security.
I want to do a quick comparison between the silicone couplers that SiliconeIntakes.com and SiliconeHose.com sells. When I was hunting down the couplers I needed, many of the other brands did not have all the sizes in stock. This lead me to SiliconeIntakes.com, which benefits from a readily available inventory.
They were well priced too, so I decided to try them out. The couplers are made from a 4-ply construction as advertised, but after inspecting them in person, something about the quality was off to me.
I was then turned to SiliconeHose.com and decided to re-buy my set of couplers. These ones are made by Flex Technologies in-house at their Southern California location. I was far more impressed with the build of these couplers, they felt dense, rigid, and less noodle-like in comparison to SiliconeIntakes’ product.
SiliconeIntakes (2.5″ to 3″ reducer) is on the left and Flex Technologies (2.75″ to 3″ reducer) is on the right. You can see how the latter has the nice orange inner layers. As always, buy the good stuff.
Rotaries are a dying breed and there are not many people knowledgable in their tuning. I didn’t want to take chances and let someone inexperienced with rotaries handle such a crucial task. So I decided to get Steve Kan to do it. Although he is quite a popular and well-known rotary tuner, the catch is he lives in Texas.
In order to make this work, I had to set up a group tuning session of at least 5 people and he would fly out. Back in December around Christmas time, I made this thread on RX7Club to start gathering others for the tune. To put things into perspective, I had only just taken off the stock turbos around this time. How’s that for project management?
Obviously the actual prospective tune date was set for around late February, but nonetheless all of this needed to be done in advance. Right from the onset, I knew I didn’t want to sit around for months to years for this car to come together. In order to achieve my lofty ambitions and goals, I needed a lot of determination, copious amounts of hard work, and a bit of luck.
Eventually the tuning session filled up and we had enough people for a 2-day event. I collected the deposits and sent them over to Steve; the date was set for March 2nd-3rd. It was time to get busy.
It came quite close because I chose to redo the fuel lines and radiator fan’s mounting at the last minute. I literally finished up the fuel lines last Friday, a day before the tunes started. I was originally set to go in on Saturday, March 1st – but I moved myself to March 2nd so I could actually get the car started.
On that same Friday night, I primed the turbos and the scene was set for start-up the next morning. Come Saturday, I placed faith in my work and quality of the parts I chose… and turned the key. It fired up on the second try and was uneventful. I then set the FPR’s idle pressure to 40 psi and spent the rest of the day double checking things. Later that night, I put the stock wheels on and dropped the car back onto the ground for the first time in over 3 months. Just like that, the car was ready for its dyno tune the very next day.
My hood was still at the painter’s and I didn’t want to drive the car without a proper tune, so I coordinated with a flatbed to pick me up on Sunday morning, and then drop me back off once the tuning was over.
Steve was using Tech3 Performance’s DynoJet in San Jose.
A couple hours later, the car was tuned and ready to go! Check out the video below.
My car was easily the loudest one there that day, thanks to the twin wastegate dumps.
And here is my dyno graph. The car made around 310 rwhp @ 6500 RPM @ 12 psi. There was pressure drop in the high RPMS and that is reflected in the graph by the downward curve. The boost actually tapered off 1.5 psi – 2 psi, so I was essentially boosting around 10 psi after 6500 RPM.
I consulted with many experts on the subject afterwards. The pressure drop is actually inherent of running off of the wastegate springs. I routed my wastegates’ vacuum lines to the compressor housing and intercooler piping; this area gives the FASTEST response and thus opens the wastegates sooner. I’m basically experiencing reverse creep. One option is to route the vacuum lines post-intercooler to inhibit a delay in the response, and gain back the pressure drop. Another is to install a boost controller. This is also where a street port can also come into play, because then the power would continue to climb for longer.
I’m going to do neither because I’m happy with the setup as is, and I like the reduced complexity and mechanical simplicity of running only wastegate springs. Many racers configure their turbo systems in the same manner to purposely achieve pressure drop at high RPMs, because in a race or track setting, lower boost at higher RPMs equals better reliability. Peak horsepower is only a number. Sure, I would have liked 350 rwhp, but that’s just getting greedy. The 5866 is still a relatively small turbo, don’t be fooled by its large frame size, so low to mid 300′s is within range @ 12 psi.
At the end of the day, I was tired from all the anxiety but mostly felt relief. I didn’t even feel that much excitement, to be honest. I mean, when you’re done with a build of such magnitude, you have to set your expectations high, it would be counter intuitive otherwise. Anything less than a fully working car would be failure.
Alas, over 3 months of work is funneled into one weekend. I admit, I am quite awed now at the fact that the car was started for the first time just a day before the tune. This just goes to show that the best way to accomplish a task is to have full confidence and commitment.
“If you want to take the island, then burn your boats. With absolute commitment come the insights that create real victory.” – Tony Robbins