Twin Seq. Dry sumping for beginners :D

[gareth]

great balls of fire!!!!!

[kam-325i]

Heres a tip for you.......

Place your filler rod in postion first, before you strick the arc..........

[Turbo-Brown]

I tend to find that just melts the rod as soon as the arc strikes though, if I put the rod on the spot I want to tack.

Anyway, little bit more progress after the belt broke on the lathe last weekend leaving me machineless All sorted now though

SO! Here's some component parts of the inlet manifolds and one which is about half done and has a throttle bolted to it.




The half done mani from the side showing the angle of attack the runner has into the port. Believe it or not, the bottom of the runner lines up exactly with the bottom of the port.




Here's a look down the throttle:




And one showing the transition from runner to head flange. The large hole will be filled by a welded plate (the thing which is kinda shield shaped in the first piccy) and the injector bosses will be welded to those plates. Advice when I was at Emerald was to move the injectors away from the head to get better mixing.




For anyone wondering why the runners are so long, it's because of that pesky thermostat housing which is very close to cyl1's port. Make the runners about 120mm long though and you've got all the room you could want

Did a trial fit of the E36 front hubs today. Looks like all that's needed is to modify the rear crud guard a bit and then make up a very simple sleeve to take the diameter of the spigot out. The E30 nut looks big enough and goes on 5 1/2 turns which is more than enough in my opinion

Think I'm going to need about a 3mm spacer behind the E46 330Ci brake discs to get enough clearance for the track rod ends but that's no great shakes. What does remain to be seen is whether the offset of the front wheels will need altering as there's about 13mm from the inner edge of the rim to the strut and, on an 8J rim I think I need the wheels as far into the arches as possible. Should just be a case of making up a set of spacers for the wheels so that the rim sits further in, will give me that nice staggered offset look too


Finally, on a more aesthetic note, I had another go at dismantling the wheels today. I tried standing in them, levering at them and was about to resort to chemicals but thought I'd give a hammer a go. Found a nice rubber headed hammer and gave the back of the spokes a twat and, after several knocks the centre dropped out







I'm really torn now between making more manifolds for the throttles tomorrow or whether to attack the rim sections with the polish!


Oh yeah, and as an aside, I'm getting a Ford style heated windscreen for the car too. How cools is that ?!!

[Andy335Touring]

Looking sweet, no pod tomorrow then ?

[maxfield]

Looking good

That Carling can reminds me

[Turbo-Brown]

Looking sweet, no pod tomorrow then ?

Cheers dude, no pod for me, I'm being a reclusive welder bee at the moment

Poor Brown, it's the last chance he'd have had too

Is there nothing Carling can't support?!

[Turbo-Brown]

Well, lack of Pod means I've now got this far with the inlet side of things:






Time to take some days off work to really get stuck into these bits

[jkarran]

Good progress, there's a huuuuge amount of work going into this, just watching from the 'web, it's easy to overlook the hours that each of these photo updates actually represents
jk

[Turbo-Brown]

Forgot to borrow the camera this weekend, but have made the great big (56mm dia) cylinder for my great big 40mm dia dump-valve.

Also made and polished the last bit of the inlet air side of the charge cooler.

[Gunni]

sweet.

How´s the vacuum control system for the twin turbo coming along?

[Turbo-Brown]

Just gotta get some more Tiggy tig gas and then I can start making the vacuum actuators.

Not really looked at the electronics despite JK kindly offering to help, will e-mail you the bits if you're still willing to have a look at them for me dude?

Other than that, got a single vee belt driven vac pump to go on.

Gonna start concentrating on actually finishing some of these bits off rather than just starting lots and lots of little bits!

Get the DV done, and them move onto the throttles. Everone likes throttles!

[Yaninnya]

Everone likes throttles!

No! Not everyone!
Nice to see that the progress is continuous on your project, Alex.
BTW. Did you think about rest of the car? Mostly about the brakes and suspension?
Jan

[Turbo-Brown]

No! Not everyone!

fair point

I'm going to keep the suspension kit I've already got which is the Bilstein Sprintline (if memory serves!) and hold on to the offset front bushes too as I like the way that Brown drives at the moment.

For the brakes, I'm using E30 M3 discs at the rear (no idea what size they are or indeed if they're any larger than 325i discs). On the front, I'm using the 325x22mm E46 330Ci discs and either the Wilwood calipers I've got on Brown (which are dust sealed) or possibly another set of Wilwood calipers I bought ages ago for the GT6 which are quite alot bigger.

Would prefer to keep the dust sealed calipers if I can though.

Think that should take care of the stopping requirements, the front discs really fill the front wheels nicely too

[jkarran]

Not really looked at the electronics despite JK kindly offering to help, will e-mail you the bits if you're still willing to have a look at them for me dude?

No problem. I'm just getting back into PICs anyway for another project so I'm a bit less rusty after a week or two reading and thinking.

jk

[Turbo-Brown]

You're still about a million miles ahead of where I'll ever be with the darned things, JK

So, this weekends spoils are.....the dump valve! Just needs a spring and a pipe connection in the top and it's done!

Here's the component parts:






And the assembled valve:





Looks a lot shinier in real life

[Speedtouch]

I have a couple of programs for the PIC 16F876 that provide PWM function (one with keyboard input and LCD display), if you're interested...

[Turbo-Brown]

Now that is interesting! Does the keyboard input allow you to adjust the look-up table settings by any chance?

Being able to have a readout and also quick access to the table would be so handy!!!

[Speedtouch]

AFAIK, the look-up table is set within the program, and the keyboard is used to select which section of the table is used, to give a corresponding change in Duty Cycle, or mark-space ratio of the PWM output signal.

Have sent copies of the files to the contact address in your 'air-in' website - hope this is OK?

Cheers,

[Turbo-Brown]

Hey Maurice, got them through yesterday. Bit over my head at the moment but will have a good hard look at trying to understand what makes them tick

Can anyone shed any light on the kind of artificial slowing down of the frequency? I send a plea for help to Microchip who said the chips can't output as slow as 15Hz and that you need to slow them down using interrupts or something.

[jkarran]

You can produce pretty much any waveform you can imagine so long as it's quite a bit lower frequency than the main clock crystal using built in timers and some simple software. You could also produce a 15Hz waveform using the hardware pwm output but you'd need to run the chip very slowly which limits you computing power so that's not the way to tackle it.

Have a think about what you want to achieve re. control strategy and adjustability then either post it up here or email and I'll go through some different possible implementations with you. Adding buttons to adjust tables and displays adds more complexity than you might imagine because you'll also need to access EEPROM memory which isn't mapped the same as the RAM though once you get into it and get the basics working (so you can see what's happening) it's not hard.

jk

[Speedtouch]

You could use dividers (flip-flops) to drop the output frequency. For instance, a watch crystal oscillates at 32,768Hz (2^15Hz), which is divided by 2 fifteen times over to give a 1Hz reference for the seconds.

Dividing the 20kHz signal of those PWM circuits by 2 ten times over gives an output frequency of around 20Hz...

[Turbo-Brown]

Been burying my head a bit with the eletronics but will need to look at them again soon I think, I'm just not getting it

Anyway, not had an amasingly productive weekend, but have managed to close off one of the inlet manifolds and get another to about 50% completion before running out of gas. Couldn't figure out why the weld just would take on the last one until I flicked the gauge on the bottle which made it drop to zero

So, here's the most complete manifold which just needs injector bosses and fuel rail ties welding on

[Onz]

looking good mate, ill look forward to seeing this dynoed!

[appletree]

Just out of intrest has there been much distortion on the plate were it will mate to the head or are you going to skim them in a milleing machine???

looking

[Turbo-Brown]

Cheers guys, can't wait to get the thing up and running myself!

I'm gonna get the manifolds all jigged up and skimmed to just make sure the angles are all the same, but amasingly the plates have all stayed pretty much dead flat.

All I've been doing is clamping them down to a plank of steel about 3/4" thick which has been doing the trick nicely

[jkarran]

All I've been doing is clamping them down to a plank of steel about 3/4" thick which has been doing the trick nicely

That'll be a nice heatsink. I wish I'd done that, mine looked like a curly wurly when I'd finished welding up my runners

jk

[Turbo-Brown]

Aah, the other thing I've been doing is jigging everything up, clamping to my plank of steel and then putting the whole lot on the gas stove for a while to bring it up to temperature. Makes welding thicker stuff to thinner stuff SO much easier!

Nice gauge of when the metal's ready to weld is touching a piece of wood against it. If the wood gets a scorch mark on it, the metal's ready to weld

[Turbo-Brown]

Ooh, had a reeeeealy bad hangover on Saturday after we threw a haloween party so haven't managed to do much this weekend.

Have badgered out the spindles, and made one butterfly. Most of the time was spend making the tooling for the spindles, but last time it took a few boring days with a tiny little slot drill to make spindles, this time it took about an hour for all six. Tooling is time well spent!









Quite pleased to have a butterfly done as they're some of the most accurate (read time consuming) parts I make. Only five left

Typed PWM CONTROLLER into the internet during my hangover yesterday too. It came up with a little 12v 555-based PWM controller which uses a variable resistor as a reference and costs about £2 to make so I might give that a try. Seems it's frequency is set by a capacitor, think you can get variable caps can't you? If so it should make the thing a little more tunable....I hope!

Basically all I want from this PWM controller is to prevent the air divertor from just banging open as soon as the large turbo starts boosting leading to a sudden drop in boost overall. Fingers crossed for the simple solution!

[Speedtouch]

Nice work! Alas, variable capacitors only come in values up to around 1500pF (1n5F) since they used to be used for tuning radio front ends prior to digital PLL tuners, hence probably won't be much use on a 555 timer circuit; in the one I think you're referring to, C1 is 0.047uF/0.1uF = 47/100nF, way bigger than variable capacitors...

http://www.mainlinegroup.co.uk/jacksonb ... 5720br.htm

[Yaninnya]

Have badgered out the spindles, and made one butterfly. Most of the time was spend making the tooling for the spindles, but last time it took a few boring days with a tiny little slot drill to make spindles, this time it took about an hour for all six. Tooling is time well spent!

Hi Alex,
IMO you shoud reduce the diameter of spindle in area where the throttle is. It don't need to be so big and you will reduce the drag (you can reduce it near by half I think; what is the diameter of spindle?).
Jan

[Turbo-Brown]

Aah, is there some way of gangint capacitors to increase their overall value? or is that just resistors?

Hey Jan, how's it going?

There's a fair amount of work to do to the spindles before they're ready to use, including drilling and tapping them for the butterfly bolts, welding an arm onto each one and machining them down to reduce restriction as you say.

The spinldes are 10mm diameter which is quite large! The reason behond that is I designed the throttles to have one sided spindles like this (in the bottom right of the picture):



Only problem being that I don't think brass is strong enough to do that with so I made the ones pictured there from stainless steel which is a pain to machine and I can't be bothered to go through it all again, especially as I'm not paying myself to make these ones

[Speedtouch]

Yeah, you used to be able to buy ganged variable capacitors; in theory just connecting them in parallel will increase the overall capacitance. Snag is, they're not ideally suited to an automotive environment unless well sealed in a casing because the vanes (on air-dielectric type) tend to collect dust, metal particles, etc.

[Yaninnya]

The spinldes are 10mm diameter which is quite large! The reason behond that is I designed the throttles to have one sided spindles like this (in the bottom right of the picture):

That is exactly what I mean. In part open throttle the bigger width of spindle doesn't matter. The flow is restricted then by the butterfly.
Jan

[jkarran]

If you can dedicate some time to the PIC solution it is your best option (besides finding extra PWM OP channels from your ecu). It should however be easy enough to make an analogue controller though it'll never be as adjustable and versatile as the PIC could be.

All variable components in a car are a potential sorce of problems down the line, they don't like harsh environments and tend to be quite fragile and unreliable in real world situations. The best bet is to set up with variable components, measure their values then substitute fixed devices when you're happy. PWM generators are easy to make and can be tuned by varying resistances (variable R is much more common, versatile and robust than variable C) instead of varying capacitors.

Is the controller basically looking at 2 pressure sensors, when the big turbo one reads higher than the little turbo one you want to start opening the butterfly? In principal this is easy to do but the problem will be stability, it'll need careful tuning to make sure it doesn't continually oscilate between open and shut yet opens quickly enough to do the rest of the engine justice. Do you have a pressure sensor in mind, if so pop up the part no.? Also, what PWM freq do you need, I guess pretty low for the solenoids, 30-50Hz? Will increasing or decreasing duty cycle (basically controller output power) open the butterfly valve?

<edit> Additional consideration will need to be given to making sure the butterfly behaves correctly once both turbos are up and running then shuts again when the throttle shuts since the pressure will be equal either side. It's an interesting and complex problem.

jk

[Speedtouch]

Agreed, variable resistor adjustment is preferable to variable capacitor, being more durable and cheaper.
Maybe you could adapt an E34 heater control since it provides a PWM output to open and close solenoid operated heater valves at around the frequency you're aiming at.

http://cgi.ebay.co.uk/BMW-E34-5-SERIES- ... dZViewItem

http://cgi.ebay.co.uk/BMW-5-series-E34- ... dZViewItem