Twin Seq. Dry sumping for beginners :D

[jkarran]

They look seriously cool... but you're missing two

I presume you're you going for butterfly valves rather than roller barrels, I guess at 48mm the butterfly wont make much difference! I'm going to make a 23mm rollerbarrel throttle as a progressive blower bypass partly because I'm worried about the integrity of a butterfly as the flow goes supersonic (or... doesn't) at about 8K RPM.

jk

[Turbo-Brown]

Yeah, those are half a set I made for a V8 recently.

They are indeed normal butterfly throttles, only with proper sealed ball bearings this time rather than the plain bearings I used on my current set.

Gonna be tricky making an accurately mated set of rollers.

[jkarran]

A mated pair would be very tricky especially as you need to get gears to mesh snugly so they open together without fluttering. I'm just going for a single roller running in ballbearings, I don't want anything there the blower could injest due to the chances of absurdly high airspeed ripping something loose in the longrun.
jk

[Turbo-Brown]

What are you reckoning on the air speed being out of interest?

If it's getting all that high then would it be an idea to make the flow paths larger so as to reduce the speed?

[jkarran]

The just using very simplistic flow vol/area calculations, bypass flow will hit 330m/s if the throttles shut at ~8Krpm and the bypass is fully open. I suspect above that engine speed shockwaves in the bypass pipes will prevent much of an increase in flow though I'm guessing a bit, I know next to nothing about fluid dynamics. There'll be an additional dump valve to cope with high rpm lift-offs. Basically, I'm limited by available space and port sizes, there's no point fitting a big bypass valve to a small port, it will only flow marginally better.

I want the bypass so I can map the boost to give smooth throttle response under cruise conditions but it'll shut down completely at high throttle openings for max boost. If I have switchable boost/ignition maps I should be able to get a well mannored road car and a tyre shredding drag car from the same setup ... that or I'll eventually get bored of the tame map, cut a bigger pully and just shut the bypass valve for maximum torque

jk

[Toby_Unna]

Alex 48mm seems pretty big dude

i mean, big given the standard port/valve sizes. i can see why it would work fine under steady load, but will it be ok under transient throttle conditions?

i don't know the answer! but the general consensus is that the 45mm m5 throttles are too big for a standard 12v M30 because it doesn't breathe well enough

[Turbo-Brown]

It would probably be too big to just use on an N/A engine, or even one with low boost but I think for high boost and all out power they're about right.

I'm not too concerned about low throttle openings as I'm running Alpha-N so the bobbly pressure signal won't cause any probs and lastly I'll be making a veeeeery progressive linkage so as to give the right foot lots of resolution.

[Toby_Unna]

fair enough

i can still see it stuttering and not doing a lot if you jump on the throttle at low revs, and with the potential not to be able to do a lot about it

still, i'm sure you'll prove me wrong and get it running nicely!

and i suppose if you're going for really big power you can't expect micra bottom end driveability

[Turbo-Brown]

My confidence in it working is backed up by the fact that Brown's already got 48mm runners which only taper down to 40mm about 50mm from the face of the head.

I reckon it'll be alright and if not, let us never speak of it again

Anyway, I cut the end tanks off a 2wd cossie IC today to see where it could be fitted under the bonnet. Looks like it will fit on top of the engine as I'd hoped but I need to come up with a way of removing it easily when tappet adjustment time comes around.

There's loads of lovely room on the nearside of the engine bay for the cooler tank and pump to live.

I've gone off the idea of making a glass fibre bonnet after talking to a mate about the idea and costs involved.

I have however come up with a design for some stiffer engine mounts which lower the engine ever so slightly and which also have a fail safe which prevents the engine from moving more than 5mm even if the elastomer fails. Wonder if it's worth scaling the design down to make transmission mounts too.

Hmmmm, will have to think about that one.

Anyway, last bit of news is that the cheque's gone off to my friendly metal supplier so I should have some materials to work with soon

[Turbo-Brown]

And finally today, I mocked the plenum and charge cooler up using the cooler core and a load of thick card.

Assuming I can do a reasonable job of welding it up it should look pretty nice in polished ali.

Can just about access the spark plugs with the assembly in place, but it'll all need removing to get to the rockers

[jkarran]

It might be worth finding a scrap cooler and having a go welding that.

I've abandoned my plan to convert a CooperS core into a charge cooler. The brazing material (presumably an aluminium alloy) used to join the end tanks to the core appeared to be blowing out of the joint while I was tacking it all up. It might have been ok but considering the risk of it blowing apart under boost or of the engine ingested the cooling water I thought better of it. Eventually I ordered a new bar and plate core form Paceproducts, price is ok but the leadtime is a ballache

It's just a word of caution, the Cossie one might be a totally different construction so you could be fine.
jk

[Turbo-Brown]

I do have concerns about the construction of the core too.

I reckon it'll be alright on the night though as the end tanks are TIG welded on as standard so it can't do the core too much harm.....I hope!

I'll weld the new water caps on the ends first and then pressure test it to see what's what. The vast remainder of the welding is not on the core so hopefull it'll be OK.

[jkarran]

My concern with welding to the core directly was melting the internal brazing that holds the fins in place. Probably just paranoia! I did however decide to glue the water jacket on instead. Araldite 2014 has a shear strength ~20N/sq mm on clean aluminium and is pretty well impervious to heat, fuel, hot water and common solvents. Of course if you're passing charge air over (rather than 'through' as it origionally would have gone) the core and running water lengthways through it you'll pretty much have to weld it all up. Fun!

Should look cool polished
jk

[Turbo-Brown]

he he, I left some of the end tanks in place so I should be able to avoid the core with the TIGgy torch I reckon. Does mean welding a little bit of cast ali but "how hard could that be?!"

Looking forward to polishing it up if it all works

Just been on the phone to order some 3" ali tube for it's inlet and some 2" tube to let the air back out to the throttles again.

Ordered a nice 2mm thick slitting saw so that I don't have to spend hour upon painstaking hour machining slots in the butterfly spindles like I've done in the past.

Really getting into this project now!

Think that Brown's gonna have to buy it when the insurance's due for renewal though just to conserve some cash so I'll be without turbo power for some time Might buy a Volvo 760 turbo or something and connect the ECU up to that though as an interim measure

[Andy335Touring]

Those mounts sound interesting, any picies ?

[Turbo-Brown]

They're just on the drawing board at the moment but I'll be knocking up a set when the materials arrive

[Andy335Touring]

[Turbo-Brown]

Hmmm, piccies seem to have stopped working

Anyhoo, I've had a really good think about the control strategy of the system and I believe that I can control the two wastegates using the PWM controls (there are two tables using throttle position and rpm as their inputs) and then make a relatively simple extra PWM controller which just looks at the pressure between the turbos and opens the air divertor valve as a function of that pressure, so it's just a 2D thing rather than a 3D thing....gy.

Also, after seeking to make all this torque off idle, I'm now having thoughts of holding the little turbo's wastegate open at idle and progressively closing it as the throttle opens, only for it to open fully again when the large turbo's started doing it's bit. Might also be a good use of Emerald's multi-map thing as wet weather driving might be a real issue.

[Turbo-Brown]

Ah haaa!

After three loooong hours on the internet looking for information on PIC microcontrollers and how on God's green Earth you program them I suddenly found that Microchip Technology sell a kit with instructional CDs, prototyping board, USB programming thingy and a chip which has both ADC and PWM (see, I'm learning all the jargon now! Apparently ADC means Analogue to Digital Conversion )

Anyway, think it's a step in the right direction for the princely sum of £32

[Andy335Touring]

Cool but a bit complex !

[jkarran]

Definitely a step in the right direction.

The PIC won't be able to drive a solenoid directly, it'll need a transistor output and some protection from the solenoid. Have a look at megasquirt idle valve driver circuits for inspiration. For RPM measurement look at the megasquirt optically-isolated tacho input for inspiration. A quality automotive power regulator is essential, something like the MAX5024 http://www.maxim-ic.com/quick_view2.cfm/qv_pk/3633 or something in a non surface mount package www.maxim-ic.com has excelent data and does free samples. Programming should be relatively simple in assembly language, there's loads of aplication notes on the microchip website with handy code examples. Are you planning to use a differential sensor or a pair of absolute pressure sensors sensors? Either will work fine and be easy to code for. A simple lowpass resistor/capacitor filter on the pressure input will remove a lot of the glitches/noise you will otherwise get from the sensors.

http://rswww.com
www.farnell.co.uk are both quick reliable sources of electronic parts.

If the valve is just to open when the pressure one side is greater than the other would a lightly sprung flap or even reed valve not suffice? No electronics needed then. That said, I'd be wary of a reed valve in case it breaks up. Even a butterfly driven by a modified wastegate actuator could give you the desired action.

jk

[Turbo-Brown]

Hey JK,

That's some very useful info! I had wondered about the power supply, was thinking just some resistors would do

Don't think I'm gonna need the rpm input as the valve just needs to open when the boost from the large turbo gets going.

However, there's a couple of important reasons why I need electronic control over the valve:

1) I don't know how it's going to react to being able to change it's parameters with a laptop is a distinct advantage over changing spring rates etc.
2) (The most important reason) As the large turbo starts to boost, it starts feeding pressurised air to the small turbo. Now what turbos do is multiply their inlet pressure. Because the little turbo will have some pressure ratio or other which is multiplying the boost of the large turbo, the pressure on the other side of the divertor valve would always be higher than that between the turbos so the valve wouldn't open.

The mechanics of it will be a butterfly with an actuator though

The all of the actuators in the system are going to be driven by a vac pump (got a pulley driven one on the way) with the consequence that any of the valves can have a programmed position regardless of boost. That also means that I can make the system fail safe so if a signal line pops off or a controllor or solenoid fails, the valve in question will open up fully so only minimal boost will be made.

If the pump fails on the other hand......

Anyway, I've been searching on RS for pressure transducers (the PIC kit is on it's way from them too ) and it seems you can get a 0-30psi transducer for about £10 or so. Doesn't do vacuum but that's not an issue as I don't need to sense that.

Quite looking forward to learning how to program these little chip things!

[jkarran]

Cool,

I wish I had your enthusiasm! You could probably do this just as well without any programming using a few op-amps. At least it's an option if you get sick of coding...

If you want any tips on the circuitry or to run it past someone for a 2nd opinion drop me a line, it's (theoretically) what I do for a living.

jk

[Turbo-Brown]

Wicked, that'd be awesome!

I'l probably run straight into a brick wall with it, but fingers crossed that the tutorials are numpty proof

[Turbo-Brown]

Oh, apparently I'm not allowed to link from that page anymore.

But on the plus side, here are some piccies of the charge cooler I've been working on.

Nobody's allowed to laugh at my pitiful attempts at aluminium welding though:

Here's the secondary plenum where cold air gathers ready to be fed into the throttle bodies. There will be 6 outlet stubs welded on the more vertical face in the photo at some point.



Here's a closer-up of some poopy welding:



These are the water tanks which cooling water is fed to and from at either end of the core.



Here are most of the components which go to make up the cooler. The big black thing is the intercooler from a 2wd Sierra Cosworth with it's end tanks sawn off:



And the bits arranged as they'll eventually be welded up. The great big 3" pipe at the back of the photo will be cut down so that it lines through with the angle of the top and bottom plates. The pipe lives on the exhaust side and faces forward in the engine bay.




And lastly, here's a little pic of the PIC microcontroller kit which arrived the other day, computer nerd-dom awaits me!

[Andy335Touring]

Nice work mate, i wish could TIG weld

[Toby_Unna]

+1

[Turbo-Brown]

+2 the piccies make it look better than it is!

Cheers chaps Hope it all hangs together under boost!

[jkarran]

Cheers chaps Hope it all hangs together under boost!

I know that feeling. The welds look pretty good in the pics, it's hard to get them looking really pretty but I don't think it makes any difference to the strength so long as the metal was clean of oxides and you get a decent puddle on both sides of the join. Mine's much the same, probably worse, it's just hidden by rubbing down with a wire brush

[Turbo-Brown]

Blimey dude, that looks a lot better than mine!

How did you go about pre-heating the metal? I've been getting pretty good, neat welds with good penetration on smaller parts that I've done, but having such a large area to get hot seems to take ages (using the gas stove this is )

[Toby_Unna]

i really want a tig welder now

is it difficult to get to the standard you two are at, assuming lots of practise with a mig welder on steel beforehand?

those welds all look pretty nice

[ric325i]

fits like a glove

[Turbo-Brown]

I think a lot of the difficulty I had getting started with TIG was in my mind Toby. People said it was difficult so I was dreading giving it a go.

It takes a while to get used to using both hands and one foot to operate the machine, and there's a tendency when you're starting to just move your foot to one position and then leave it there rather than concentrating on the formation of the weld, but once you get used to it you start to relax into the procedure.

I quite enjoy doing it now

[Ant]

TiG rules, anyone with gas welding experience would take to it easy, nice welds gents, alooy is shite to weld as all the slag pools and contaminates the weld you're trying to lay down.

TBH a cheap TiG is a bad investment, one with slope in/out and pulse is a must have I'd say for any serious welding.

Dont forget( for those that dont know) you can MiG Stainless and alloy, but the welds are chunky to say the least

Nice work Alex, still ant get my head around why the PIC is required dude, surely you have some spare 5V logic outputs on the Emerald you could use ?

Example, use the 5V logic "boost control" output to map the divertor between snails , that would allow full 3D maping of the changeover using a TPS driven or rpm driven algo( obv a relay or FET would need to run the actuators side....) TPS driven seems the obvious choice

if there's no spare PWM outputs that is ??

[Turbo-Brown]

I've only got two PWM outputs spare unfortunately Ant.

One will be assigned the task of controlling the small turbo's WG which is also the Exhaust gas Divertor Valve (EDV ) and the other will be in charge of the external WG for the large turbo.

These are both mappable against throttle position and rpm and as they're independent of eachother, I should be able to tweek the crossover until it's perfect.

Also, as I've chosen to actuate everything using vacuum instead of boost pressure, I can give the small turbo a rest at idle by just holding the EDV open, and have a map which gently feeds the boost in as the throttle opens giving the car a 'big engined' feel rather than wheel spinning in and out of every parking space Think that will have the peculiar effect of making the exhaust note become really quiet as I pull away though, will have to play when the time comes.

Anyway, that leaves the small matter of controling the Air Divertor Valve (ADV). I believe that control of this will have more to it than an open or closed control system.

I think what needs to happen is for the ADV to progressively open as a function of the pressure between the two turbos (the small one feeds from the output of the large one) as it opening fully once a certain pressure or throttle position threshold is reached will, I reckon, lead to a sudden loss of boost.

Anyway as a result of all that, I started looking for another PWM device and found that these PIC things can take an analogue signal from something like a pressure sensor and give a PWM output based on that, so all I need do is plumb a solenoid valve into the signal line to the ADV actuator and Bob's yer uncle...........I hope