Fellow GSers,
Having experienced the dreaded fuel starvation problem on my '99 (stock except for a drop-in K&L) like so many of us, I'd like to propose a fix which I believe I haven't seen in other topics on this forum or anywhere.
My bike had the classical symptoms of sputter/cutoff after a few minutes at sustained 7000 rpm or more. They went away when I switched the petcock to PRI, so I know that the problem is there. As we know from the wiki or the service manual of your choice, the petcock is vacuum-operated and the vacuum comes from one of the carbs via a hose on the back of the petcock as shown here:
(http://www.flickr.com/photos/85606104@N06/7842001262/)
All the fixes I've found online involved either fitting in another petcock (04-09 GS, Honda CRF 250 x, etc.), drilling the fuel passages wider, or converting the stock petcock to non-vacuum by plugging the vacuum line with JB Weld. How about taking the problem from the other end and increasing the suction force on the diaphragm to pull it out further and increase the fuel flow?
It turns out that the vacuum chamber has a built-in restrictor plate:
(http://www.flickr.com/photos/85606104@N06/7842001182/)
Eureka! Make the holes wider, increase the suction force (= vacuum pressure diff * cross-section). I re-drilled the center hole with a 1/16" bit (1.5mm would work in the metric world), cleaned the outer holes with a wire, and blew everything clean with 100-psi air - including the vacuum line to the carb for good measure. The restrictor plate is thin and soft, so I did not need any power tools:
(http://www.flickr.com/photos/85606104@N06/7842001054/)
The total cross-section (all holes combined) is now ~25% wider and so is the suction force. It works spectacularly well. Fuel starvation is *gone* up to 9000 rpm (I did not explore higher), the choke response is much more progressive on a cold San Francisco summer start, and part-throttle response is much better without the flat spot I used to have. Definite proof that the fuel flow has improved.
Total cost: zero, and the fuel circuit is still stock and 100% compliant with existing repair manuals. Let me know what you think!
I dont see any pics
Nice, I'll remember this if I ever have the problem. But I have to wonder what year this problem started? I have an 06 and I can push it to 9 or 10k without a problem. In fact it hits a power band at 7500rpm and jumps up to 9 real quick......bone stock GS. If it was just a restrictor plate shouldn't all of them have the problem?
I had a bike that had a similar problem but it was in the tank cap. The vent in the cap was stuck and 2 or 3 times at high rpm's it would just die....and would not re-start. I would open the cap and and it would start right up. From the on I would just knock on the cap with my knuckle when it felt like it was starting to bog down.
Pictures:
(http://farm9.staticflickr.com/8285/7842001262_564833d43d.jpg)
(http://farm9.staticflickr.com/8431/7842001182_2d1b455f4a.jpg)
(http://farm9.staticflickr.com/8294/7842001054_7d4b50aa59.jpg)
@RossLH, thanks for posting the pics. They do make things clearer, don't they?
@pliskin, the petcock on post-2004 models is a completely new design which solves the fuel starvation problem once and for all.
:bowdown: :bowdown: :bowdown: :bowdown: :bowdown:
Great man!
Absolutely great! People will love you for that!
Would it be called the "ggg petcock mod"?
ggg practices world domination by stealth :) How about something more anonymous such as "Petcock Penicillin?"
Will this modification result in a decrease in miles per gallon? If not, I'm gonna give it go. Fuel starvation is a sucky feeling and I don't want to have to worry about being eaten by an 18 wheeler while I fiddle around for the petcock lever at 70 mph.
As long as your float needles are functioning at 100%, it shouldn't touch your mpg.
It looks to me like it's basically allowing more vacuum pressure to the diaphragm, and in turn allowing more gas to flow when vacuum is present. Makes you wonder why they have the restrictor plate in there in the first place. Maybe if too much vacuum goes to the diaphragm, it will damage it?
Enginerd guesstimate:
The holes are small so that when the diaphragm fails, you don't dump too much gas into the vacuum line?
If $ are not too tight, a new petcock is not that bad to buy every 10-15 years...
@bombsquad: I suspect it's a design compromise. The vacuum line to the carb has to be reasonably large, otherwise it would create its own restriction (the pressure diff would be noticeable over such a long, narrow tube) and it would also foul quickly over time. However, the large cross-section probably creates too much suction on the diaphragm and would shred it if unrestricted.
@DoD: A new petcock on the old design will have the same restriction issue as the old one if left unmodified!
Quote from: ggg on August 23, 2012, 09:57:26 AM
ggg practices world domination by stealth :) How about something more anonymous such as "Petcock Penicillin?"
That too hard to right or spell...but you are the creator, you name it!
This should be wikied and stickied, I think...
This mod would only work if you have a vacuum leak leading in between the 'restrictor' and diaphragm... Pressure times area squared is true for flow rate, but when the pressure between the petcock and intake is in equilibrium there will be zero flow unless you have a vacuum leak inside of the petcock.
As a side note, there's a good chance that the restrictor is there in order to provide a buffer to the constant vacuum-no vacuum cycle of the intake in order to reduce wear/cycling on the diaphragm/spring.
@Badot: There isn't any fluid flow indeed in the vacuum line to the petcock in the absence of a vacuum leak. However, widening the restrictor plate changes the static equilibrium of the diaphragm and therefore the fuel flow.
The diaphragm is subject to two forces:
- the fuel pressure, which pushes it away from its seat. This force is the diaphragm area times the pressure differential between the fuel (at atmospheric pressure since it comes from the tank vented to the outside) and the petcock vacuum chamber (more on this below)
- the elastic force from the spring, which pushes it back in. This force is proportional to the displacement of the spring, i.e., the distance between the diaphragm and its seat
The pressure in the vacuum chamber is different from the vacuum pressure in the carb (which is roughly the same as in the vacuum line to the petcock). The difference is caused by the restrictor plate. If the restrictor hole is small enough, it creates enough resistance to fluid motion that not all the air can be sucked out of the petcock when the engine starts and a vacuum is created in the carb. That's an exact analog to electrical resistance.
(Try this experiment with a party balloon: Inflate it, tie it shut, then pierce it with a needle near the knot where the rubber is still thick and soft. It won't burst, let it deflate naturally. You'll find that a sizable volume of air remains trapped inside and you need to squeeze the balloon to get it out. Now inflate another balloon, tie it shut, and cut a larger hole with scissors near the knot. That balloon will deflate faster and there'll be less air inside at the end. The only difference between the two balloons is the size of the hole and the one with the smaller hole has more residual air, which means that the pressure inside is higher due to the greater resistance of the hole. The vacuum petcock works on the same principle.)
If I make the restrictor hole larger, then the pressure difference between the carb and the vacuum chamber will be lower. The carb always produces the same vacuum pressure, so the pressure in the petcock vacuum chamber will be lower. This implies that the fuel will want to push the diaphragm further out and the spring will need to compress a little more to balance that larger force, i.e., the diaphragm will indeed move out and the fuel passage cross-section will be larger.
This drawn-out explanation would call for a picture... volunteers? :)
Quote from: ggg on August 24, 2012, 10:28:46 PMThe pressure in the vacuum chamber is different from the vacuum pressure in the carb (which is roughly the same as in the vacuum line to the petcock). The difference is caused by the restrictor plate. If the restrictor hole is small enough, it creates enough resistance to fluid motion that not all the air can be sucked out of the petcock when the engine starts and a vacuum is created in the carb. That's an exact analog to electrical resistance.
(Try this experiment with a party balloon: Inflate it, tie it shut, then pierce it with a needle near the knot where the rubber is still thick and soft. It won't burst, let it deflate naturally. You'll find that a sizable volume of air remains trapped inside and you need to squeeze the balloon to get it out. Now inflate another balloon, tie it shut, and cut a larger hole with scissors near the knot. That balloon will deflate faster and there'll be less air inside at the end. The only difference between the two balloons is the size of the hole and the one with the smaller hole has more residual air, which means that the pressure inside is higher due to the greater resistance of the hole. The vacuum petcock works on the same principle.)
That experiment would not be accurate. If the hole were small enough, the contraction of the rubber on partial deflation could easily be enough for the balloon to seal itself, which is what I believe you are referencing. By the looks of that picture, the restrictor is made of a hard, resilient plastic, correct? Just a nonmoving part that the air flows through? Not a regulator - just a small hole?
A more accurate experiment would be filling a party balloon and letting it deflate by itself or through a tube. Repeat the experiment, only use a much smaller tube (restricting the flow). In both situations, so long as there is a path for the air to flow, it will continue flowing until the pressures equalize, at which point there will be no air flow. No matter how much you restrict the output, so long as air can still flow, it will never hold residual pressure. This is the exact thing happening between the petcock and vacuum port.
Like you said, it creates resistance to fluid
motion, it slows the motion and only the motion.
Again.....if it was as easy as opening out a hole why didnt Suzuki do this instead of redesigning the whole component :dunno_black:
A good question... All I know is that it did the trick on my bike.
Quote from: ggg on August 25, 2012, 11:28:02 PM
A good question... All I know is that it did the trick on my bike.
Anyone else willing to try?! (I don't have a GS500....)
Cruising down the freeway at 65+mph on my wifes 94 ( trey ).Full tank,80 degrees or so,5500rpm's bike starts acting like it's running out of gas and then dies.Pulled over,removed tank bag and opened fuel cap.After closing the cap it started right up and ran fine!!.
Could this be a petcock issue ???.
Sent from my iPhone using Tapatalk
Sounds like it could be vapor lock in the tank.
Ya thats I was thinking,just exploring all posible causes.
Sent from my iPhone using Tapatalk
pingle FTW
Quote from: NortwestRider on August 27, 2012, 06:38:51 PM
Cruising down the freeway at 65+mph on my wifes 94 ( trey ).Full tank,80 degrees or so,5500rpm's bike starts acting like it's running out of gas and then dies.Pulled over,removed tank bag and opened fuel cap.After closing the cap it started right up and ran fine!!.
Could this be a petcock issue ???.
No. Tank vent, in the gas cap.
How to clean the gas cap tank vent (from the FAQ section)
http://gstwins.com/gsboard/index.php?topic=41061.0 (http://gstwins.com/gsboard/index.php?topic=41061.0)
I'd say the same thing.
Full disclosure on the petcock fix: I've had to disassemble and reassemble the petcock twice since I rejetted the restrictor plate because the fuel starvation symptoms were coming back. Both times, I found fuel on the vacuum side. The diaphragm is not punctured, so it's possible that the extra suction force is pulling a little too hard on the diaphragm and creating a fuel leak at the edge. When I rejetted, I rebuilt without sealant, the fix lasted 50 miles. I tried a small amount of multi-purpose grease in the diaphragm seat groove, that lasted 230 miles. Last night, I put gasket sealant in the groove (on the fuel side) and around the rim of the diaphragm (on the vacuum side). We'll see how this works.
If fuel is getting into the vacuum side and the diaphragm isn't punctured, air is getting in too. The leak is more likely than not the original cause of your problem.
Again, there is no extra suction since you widened the hole, unless doing so offset a leak -- that would be like saying when a tire gauge reads the pressure through a valve stem that the pressure in the tire is higher than the pressure in the gauge, or that in a compression test the pressure in the cylinder is higher than the gauge reads since it has to go through the hose (neglecting the volume of the hose).
That may very well be... Today, the bike ran great for 70 miles, then the problem came back. As before, switching to PRI got rid of the issue. More interestingly, I switched to RES after a few minutes on PRI and the problem did not come back after a few minutes like it does when I go to ON. Maybe the diaphragm is porous after all, maybe the fuel on the vacuum side comes up the vacuum line--at any rate, I have reached the limit of my patience on this issue. I just ordered a new petcock on eBay and we'll call the fix unproven until someone else replicates it and reports results.
One more data point: the bike ran flawlessly today with the petcock on ON, and I didn't change anything since last night. Also, I got 160 miles on a tank (same commute route, same traffic, same everything as usual) without needing to switch to RES. The other time I knew the petcock was leak-free after rejetting it, I had to switch to RES after 157 miles. Before the fix, it was 135-140 reliably. More questions than answers at this point:
- Could the problem be linked to ambient temperature? Since I did the fix, starvation has always happened on my way back from work in the afternoon with temps in the 70s, not in the morning with temps in the 50s
- Does the fix help get more gas out of the tank? It can't improve fuel economy--that's controlled by jet size and fuel pressure, and neither has changed. My theory is that with a larger fuel passage under the diaphragm, there is less resistance to fuel flow, so the fuel can keep flowing at a lower fuel pressure, i.e., with less fuel in the tank (fuel pressure = atmospheric + weight of the fuel column in the tank).
Thoughts?
Just a theory, but it may be air bubbles working their way out of the system. I had odd symptoms like you described with the bike becoming fuel starved and switching to reserve and back fixed it, this was after converting my petcock to manual flowing. Happened 2 or 3 times on my 320 mile ride home, but didn't ever happen again.
I was just thinking along similar lines before reading your post in view of today's ride home. 15 miles into it, the bike bogged down again. I switched to PRI for one mile, then back to ON, and everything was normal for the rest of my 50-mile commute. Switching to PRI bypasses the diaphragm but not the petcock and would indeed work around any air left in the diaphragm chamber.
Your theory makes even more sense considering that I did not re-prime the fuel circuit after resealing the petcock two nights ago, because I wanted the sealant to cure overnight and I wasn't sure that exposing the uncured sealant to fuel was a good idea. I'll keep an eye out for 1-2 tankfuls--and maybe I'll return that new petcock I bought on eBay after all. (I can reassign that $50 to something more useful such as Progressive springs.)
Another day, another data point. No starvation issues whatsoever, still on ON after 158 miles since filling up. Looks more and more like Badot was right and my recent issues were caused by air bubbles--and also like the fix is working.
Just for fun, I took the GS out for a highway run to test for fuel starvation. As expected, after a minute or two at 70-75 mph, the bike began to sputter. I suddenly remembered to flip the petcock switch to PRIME and the bike roared back to life. I think I'll stick with this solution rather than start widening holes in my petcock's baffles. Although I do applaud ggg's ingenuity.
Same symptoms indeed - mine took 4 miles at steady 7000 rpm in 6th. The verdict for today's tank: 163 miles before switching to RES. Definitely on to something.
Quote from: mimikeni on August 31, 2012, 06:00:37 AM
Just for fun, I took the GS out for a highway run to test for fuel starvation. As expected, after a minute or two at 70-75 mph, the bike began to sputter. I suddenly remembered to flip the petcock switch to PRIME and the bike roared back to life. I think I'll stick with this solution rather than start widening holes in my petcock's baffles. Although I do applaud ggg's ingenuity.
Could be the petcock is wearing out. Any part on the bike with rubber can wear out. So replace it. Might be a leaky, pinched or plugged up vacuum line from the carb. The space between airbox and frame isn't much. I find it easy to get the hoses pinched a little unless I'm careful and shove them where they fit best.
600 miles on the petcock fix and no signs of fuel starvation, even after a washer-under-needle rejet which should trigger it a little earlier. I'd say we're moving from possible to probable validation.