My first regulator.... that really works

[rsterne]

No problem.... Now, while we're on regulators, I have a BASIC question that has been bugging me.... I've seen lots of regulators with stacks something like this....

)()(())(())(||

or whatever, but a mix of singles and doubles, all the same thickness (except the flat shims).... I find this totally confusing, for the following reason.... Let's say it take 100 lbs. to flatten a single 0.020" thick disc.... Let's further assume that the stack is subjected to a load of 100 lbs.... Wouldn't the single discs flatten completely, so the stack now looks like this?....

|||(())(())|||

In the original stack, the travel would be 8 times the "cup" of one disc.... If the cup was 0.020", the travel would be 0.160".... The "spring rate" (according to my math) would be K/6, where K is the rate of a single disc....

Once a load of 100 lbs. is applied, and 4 of the discs have collapsed (and the other 4 are collapsed half way), there is only 0.040" travel remaining.... The spring rate for that remaining portion of the travel would be much stiffer than it was initially.... My math says it would be K/2.... ie 3 times as stiff as it started out.... The stack started out 0.400" long with no load, and at 100 lbs. it was 0.280" long....

If the stack had been arranged like this.... (same number of bellevilles = 12)....

))(())(())((||

it would have a rate of K/3 and a travel of 0.120", but the spring rate would be a constant instead of losing half the travel and having a sudden change in rate.... At the same 100 lb. load, it would have compressed 0.060", and still have that much travel remaining.... This stack started out 0.360" long, and at 100 lbs. it is 0.300" long.... It has a linear spring rate, more travel when half loaded, and the only difference is that it may take a slightly different shim pack to set the pressure you want....

My questions are.... Firstly, do I understand this properly.... In a mixed stack of singles and doubles, do the singles go flat first?.... Secondly, if that is true, WHY would somebody assemble a regulator with a mixed stack of singles and doubles?.... I've found that once bellevilles are driven completely flat, they tend not to rebound conpletely but take a "set".... Why would somebody complicate their life by mixing the stacks?....

Bob

[Motorhead]

You think too much !!

[lloyd-ss]

Bob,
You have the stacking figured out correctly.  The top example will give a stepped rate. Soft then stiffer.  I have to go to work now.  But there is LOT to your question.  I'll be back!
Lloyd

[Bent]

The reason for single washers to be mounted is to have some extra movement/opening when pressure is low on the secondary side.

And I also made my own regulator for +300bar to 120-250bar. These are my "workdrawings".

I only made this one for when refilling my HW100, but the parts are made rather small so they can be retrofitted to an unregulated airguntank.
The belleville washers can be otherwise stacked or exchanged for other pressures.
https://www.dropbox.com/s/kluxmlitdv5fevh/2012-11-07%2002.09.51%20%28Medium%29.jpg

The overall lenght of the regulator is only  58mm.

[Cal]

Regarding the Belleville stack orientation,    Think progressive!

Regulators are not supposed to be on off switches.    Small changes need sensitive responses ;-)

[rsterne]

For those that think a regulator can respond during the 1-2 mSec it takes for a shot to occur, I would like to see some proof.... I have used regulators with a gauge on the output side, and you can see the needle drop during the shot and take up to a second to reach full pressure again.... I simply don't believe it happens....

Regarding the idea of having a progressive spring rate, I can see that as being an advantage in some applications, but not in a regulator.... The FORCE isn't constant anyway, even with a constant spring rate the disc springs develop a lot less force when the regulator is open (ie the output pressure is low) than when it is on the seat.... Add in the fact that once you drive the bellevilles to flat you damage them, and I fail to see the point in using a mixed stack....

If I've got it wrong, please enlighten me....

Bob

[Bent]

For those that think a regulator can respond during the 1-2 mSec it takes for a shot to occur, I would like to see some proof.... I have used regulators with a gauge on the output side, and you can see the needle drop during the shot and take up to a second to reach full pressure again.... I simply don't believe it happens....

Regarding the idea of having a progressive spring rate, I can see that as being an advantage in some applications, but not in a regulator.... The FORCE isn't constant anyway, even with a constant spring rate the disc springs develop a lot less force when the regulator is open (ie the output pressure is low) than when it is on the seat.... Add in the fact that once you drive the bellevilles to flat you damage them, and I fail to see the point in using a mixed stack....

If I've got it wrong, please enlighten me....

Bob

You should have around twice the volume of port and barrel as the regulated volume. Larger regulated volume means you can go down in pressure >> more shots >> higher "effiency" of the airtank.
A larger volume also secures that after shot taken, there is enough pressure in the regulated area to shut of the valve quick and consequent >> not "spilling" air.
Sorry if some terms aren't correct english.. 

[rsterne]

You should have around twice the volume of port and barrel as the regulated volume

Sorry, I don't understand how it relates to the barrel volume.... I always strive for about 1 cc of volume between the regulator and the valve seat (outlet) for every FPE the rifle is expected to produce.... Less than that and you have to run more pressure, hence fewer shots.... I agree that having lots of volume mya help the valve close quicker and/or avoid bounce as well....

Bob

[Rescue35]

I believe he is referring to the "chamber" area of the barrel. Behind the pellet and in front of the bolt face. That was my take.

[lloyd-ss]

Bent,
Making regulators is not easy.  Did yours work out ok?  I am curious about the groove/gland configuration for the o-rings on the moving parts.  It seems the grooves work better when they are narrower and deeper, compared to a groove for a static o-ring.  What did you find that worked best? Are you using the tiny 1mm cross section o-rings on the moving parts?

[Cal]

For those that think a regulator can respond during the 1-2 mSec it takes for a shot to occur, I would like to see some proof.... I have used regulators with a gauge on the output side, and you can see the needle drop during the shot and take up to a second to reach full pressure again.... I simply don't believe it happens....

Regarding the idea of having a progressive spring rate, I can see that as being an advantage in some applications, but not in a regulator.... The FORCE isn't constant anyway, even with a constant spring rate the disc springs develop a lot less force when the regulator is open (ie the output pressure is low) than when it is on the seat.... Add in the fact that once you drive the bellevilles to flat you damage them, and I fail to see the point in using a mixed stack....

If I've got it wrong, please enlighten me....


Bob

If one did not want the regulator to screech, buzz and whine,  he might not care about spring rate tuning.
 True also,  that some conditions might allow straight rate without these undesired side effects.

Regarding response during a firing cycle,   Such rapid response, if attempted, will surely be the death of most regulation valving seals, certainly resulting in deformation that results in undesired outlet pressure change. (creep)

[Bent]

Bent,
Making regulators is not easy.  Did yours work out ok?  I am curious about the groove/gland configuration for the o-rings on the moving parts.  It seems the grooves work better when they are narrower and deeper, compared to a groove for a static o-ring.  What did you find that worked best? Are you using the tiny 1mm cross section o-rings on the moving parts?

It works, but still need fine tuning. First i used POM/DELRIN for the valveseat but half the time it didn't close before the pressure was about 25 bars higher than setpoint. After changing to PTFE,  which are a litttle softer it's maybe one of ten where it goes 15-20 bars to high. Haven't had time since I made it, but I think I polished/rounded the valvepibe to much... I made it just 3 weeks ago, and Christmas and familytime is rated higher right now... 
The O-rings are 11,28x1.18(the same as I use for the portpart on my HW100), so it's a tight fit, and the little hanging could also come from this. Even if everything got a little molykote, wiped dry afterwards.

[lloyd-ss]

Bent,
 I think we are similar in our development.  Mine is working ok, but can be better.  Thanks for the idea about the valve seat material.  I will try some different ones.  Also, I forgot to put a vent hole in the side of mine, ha, ha. I will fix that too.
Have you seen this site?
http://www.zdspb.com/tech/misc/maxflo.html
LOTS of drawings of regulators.  Also, look at their paint ball gun animations.  They are complex and beautiful! (techno-geek)
Lloyd-ss

[Bent]

Llloyd,
Sounds like we are..  Learning by failing....  And I'm still failing a lot.
I'm rather new to PCP's first time I shot a PCP was in March this year.
I'm used to working with spreadsheets and development, and in PCP I can combine hobby, shooting, developing and some calculation all in one....

The reason for the narrow groves was to avoid the O-ring from "rolling" in case it sticks a little to the wall.

Never seen that page before, great description and graphics.

[rsterne]

Lloyd, I notice you have the thinner Bellevilles in the middle of each set.... I did mine the other way and it took a long time (like ~30+ seconds) to reach the setpoint.... ie it crept up the last 100-200 psi.... Do you know if it should make a difference?.... I'm using 0.032" and 0.022", with flat loads of 260 lb. and 105 lb. for a total of 365 lbs.... Are you using any lube on the washers, and/or did you polish the rims?....

Well, I just swapped them around.... They used to be Thin Thick Thick Thin for each (()) double-paired set.... Now they are Thick Thin Thin Thick.... and the problem went away.... Interestingly, the setpoint didn't change, just the response time to balance.... I have a total of 14 Bellevilles and the adjustment range I have is 1200-2700 psi, but it takes only about one flat on the adjusting bolt (1/6th turn) for a ~100 psi change.... Anyway, works fine for my purposes....

Bob

[lloyd-ss]

Bob,
That is interesting about putting the thicker Bellvilles on the outside of the quads gave a faster response time.  I remember reasoniing through which should be on the outside, but honestly, I can't explain the logic.  The idea of using some lube in the stack is interesting.  Maybe an extreme pressure lube, like lathe center point lube, or hypoid gear oil?  I think I might try some lube in the stack.
Lloyd

[rsterne]

I would think it might be worthwhile polishing the edges of the bellevilles with 600 grit and I'd like to try moly paste burnished in.... Anyway, I was using it today and swapping them around was a HUGE improvement in the balance time....

Bob

[lloyd-ss]

Bob,
Thinking about it, that certainly makes sense. Regulators that have coil springs don't have the friction of all the Bellville's sliding past each other, but you need the Bellville's to get the required pressure and travel. The coil springs are basically frictionless in a reg.  And then when the spring stack is almost at equilibrium, any little bit of friction becomes a big player in the final settling of the reg. It is making more sense all the time.
Lloyd

[rsterne]

I guess that would speak in favour of using single, heavy Bellevilles (eg. the 0.047" with a 600 lb. flat load) instead of pairs of thinner ones.... when high forces are required.... The only downside is they have less travel....

Bob

[Motorhead]

Makes perfect sense reducing friction at the contact areas of each and every washer so there movement ever so minor is done so with the least amount of parasitic drag/friction.