Topic: DIY Flow Bench w/VIDEO (Read 31138 times)

HiProfile · « **on:** November 14, 2011, 08:13:36 PM »

I've been perfecting this for a while, and the bench is 99% done. I just have to run some fluid through to remove any shavings/etc. I have an adjustable FPR that goes on it and returns it to the tank, but it will pump into a jar until cleaned.

It's just a D16 rail tapped where the stock fpr is, with washers and threaded rods to hold the injectors in. Later I'll be getting 14mm adapters for k-series size injectors. The two pumps are just 79lph civic pumps, but I have 139lph dohc pumps availible. It pumps into the delrin bulkhead to a single outlet, then to a stock CR-V filter, through the rail, and back to the delrin bulkhead.

I used threaded rod to make bulkhead electrical connections - ground is shared, while possitive is seperate for each pump. Later if I step up to larger pumps, I can run one or both depending on the injector size. These two should pump enough for four 700cc injectors at 43.5psi, and the GSR pumps might do >1200cc.

I have some picaxe microcontrollers on order, and I'll eventually add those for multiple pulse widths and frequencies JUST to please JD (idle flow).

92CXyD · « **Reply #1 on:** November 14, 2011, 08:22:23 PM »

Kool when are ya ready to flow test our shit?

HiProfile · « **Reply #2 on:** November 14, 2011, 08:28:11 PM »

I just have to finish the wiring, make the injector clips universal, and it's go time. I won't be able to get super-accurate until I get the microcontroller set, but I can test it against a stock 240cc and get very close. With the picaxe I have plans to also check for latency, then get a bench PSU to try 11v to 14.5v.

92CXyD · « **Reply #3 on:** November 14, 2011, 08:30:18 PM »

Sweet, can not wait to see this bench get some action.

Vids and lots of pict. for the write-up when ya do the wiring, please.

Joseph Davis · « **Reply #4 on:** November 14, 2011, 08:45:51 PM »

Nice Accord motor.

HiProfile · « **Reply #5 on:** November 14, 2011, 08:57:40 PM »

I'm baller like that. I only bought it for it's injectors & clips though...

The wiring will be straight forward. Each injector will be grounded and fed battery voltage via an EV1 clip with the outside shroud removed. All injectors have the same pin size and spacing. I also have a resistor box to swap in for low-z injectors, but my main goal is high-z injectors.

For now I'm just doing raw flow tests to determine size.

crxvtec91 · « **Reply #6 on:** November 15, 2011, 04:01:22 PM »

Fuck yea!

HiProfile · « **Reply #7 on:** November 21, 2011, 07:01:31 PM »

Now for the controller. I would normally program lots of presets into it, but the chip limits me to 15 presets (sub procedures) max. I'll need to use input variables to get what I want.

I'm thinking of having a simple input dial (values 0~10) for it and 3 input-select buttons. It will let me select injector duty and frequency (rpm) in steps that will be useful (so no 2% at 15000rpm). Zero will select a static flow test, since there 0 rpm and 0 duty wouldn't flow fuel...

I'm thinking I should make duty cycle type selectable, as in select percentage OR milliseconds. I would set the dial to a number, say 8, then press a button which inputs that value into the chip. 3 different buttons could let me use the number either for 80% duty (8 x10%), 8 milliseconds flow time, or 8000 rpm (8 x1000rpm) for each spray window, it just depends on which button I press. The 4th button starts the program.

This will give me various modes for cleaning, checking flow, and letting me calculate injector dead time.

Any other ideas?

HiProfile · « **Reply #8 on:** November 21, 2011, 10:31:30 PM »

Don't mind this, just fooling with code and need a way to "mail" it home:

Code: [Select]

main:
	high b.0				; light up LED
	let b1 = 0			; set variable DUTY percent
	let b2 = 0			; set variable DUTY milliseconds
	let b3 = 0			; set variable FREQUENCY rpm
	goto blinkled			; call main procedure


blinkled:
	gosub dialin			; call sub procedure and return	
	if b1 > 0 then
		for b0 = 1 to b.1
			for b0 = 1 to 50		; loop 50 times
				high b.0		; light up LED
				pause 10
				gosub dialin
	low b.0				; dim LED
	if b2 > 0 then
		for b0 = 1 to b.2
			for b0 = 1 to 50		; loop 50 times
				high b.0		; light up LED
				pause 10
				gosub dialin	low b.0
	pause 100
	if b3 > 0 then
		for b0 = 1 to b.3
			for b0 = 1 to 15		; loop 15 times
				high b.0		; light up LED
				pause 10
				gosub dialin
	low b.0				; dim LED
	

dialin:
	if pinC.1 = 1 then		; if DUTY.per pin is on
		pause 100
		readadc C.5,b1
		let b2 = 0
	if pinC.2 = 1 then		; if DUTY.milli pin is on
		pause 100
		readadc C.5,b2
		let b1 = 0
	if pinC.3 = 1 then		; if FREQUENCY pin is on
		pause 100
		readadc C.5,b3
		b3 = dial.pin
	if pinC.4 = 1 then		; if START pin is on
		pause 100
		goto Flownow
	return

fuse · « **Reply #9 on:** November 23, 2011, 10:56:52 PM »

Nice work cant wait to see vids. What will you run threw them to clean?

HiProfile · « **Reply #10 on:** November 23, 2011, 11:24:47 PM »

Mineral spirits for flow, then 50/50 spirits + Chevron cleaner for dynamic flow/ultrasonic cleaning.

DasPoop · « **Reply #11 on:** November 23, 2011, 11:49:33 PM »

Damn, This is going to be bad ass. Jd likes goats. Baaaaaa. Baaaaaaa.

HiProfile · « **Reply #12 on:** November 26, 2011, 06:41:40 AM »

New code is written, just need to x-fer to PC, test, and alter again. I'm adding a ultrasonic cleaning mode where it will click them on/off for 1-10 minutes but keep the pumps off.

HiProfile · « **Reply #13 on:** November 26, 2011, 07:39:33 PM »

The code is finished.

Its not very pretty, but it simulates w/o problems. Now I just have to test it on a breadboard with LED's then wire it up.

You turn the dial to a number, then press either Duty button (milliseconds OR percentage), then change the dial to rpm x1000 number and press the Frequency/Mode button, then press start. Pump primes and injectors pulse for 6 seconds.

Turn the dial to Raw Flow, press Frequency/Mode button, next press the start button, then watch it prime the pump and flow all injectors for 6000 milliseconds. Multiply by 10 to get raw CC/Min.

Turn the dial to how many minutes your ultrasonic bath is set to, press either Duty button, next turn the dial to Ultrasonic and press Frequency/Mode button. Once you hit the start button it will pulse them in alternating 10, 25, and 50ms loops.

HiProfile · « **Reply #14 on:** December 10, 2011, 10:56:39 PM »

I have the bare bones working atm. I'm testing it with a chip that has less inputs/outputs, so I just reprogram it when I want a different mode. Right now it has a 6 second wide-open test for static flow and a 10/25/50 millisecond pulse mode that has them open for 6 out of 12 seconds.

It tests fine with at least 7 amps (1 peak/hold and 3 saturated), but I have a resistor box that plugs in by changing 3 connectors. I just have to clean out the fuel rail and pump block, then I'm set.

I will try to get vids monday or tuesday.

The next step is modding the power supplies so I can test latencies at different voltages, then setting up the full-featured chip.

HiProfile · « **Reply #15 on:** December 11, 2011, 02:02:41 AM »

I'll think about it, but I need to make sure it's stable. The controller is easy to replicate, it's just the hardware that's time consuming. Especially if you want it to look nice vs just functional.

Right now it can clean any 12mm top feed injector, and 14mm top feeds would need an adapter. Actually the 14mm fit if you put on two 12mm o-rings. Side feeds would just need a fuel rail & a FPR adapter from ebay.

HiProfile · « **Reply #16 on:** December 11, 2011, 02:55:11 AM »

I just found someone who was serious about getting an ASNU "Junior", and his quote was $7100. That's fucking crazy. The only thing that I know I can't match in functionality is the ultrasonic cleaner, mine was $30.

I can also reprogram mine. Anything can be done, it's just a matter of time and materials.

The pulse side of things is solid. This thing can control down to 10 millisecond pulses (1/100,000 second).

HiProfile · « **Reply #17 on:** December 13, 2011, 08:16:30 PM »

So between a pump being locked up, another with a bad relief spring, and the power supply I tried only giving me 10 volts, it wasn't a good day.

With one integra pump it makes 55psi max, and integra+civic pump it makes 40psi max.

Tomorrow I'm going to do some testing with just the single pump, and I'm ordering a ching-chong 255lph pump tonight.

HiProfile · « **Reply #18 on:** December 14, 2011, 07:42:38 PM »

Got a bit more power with the different PSU, but there's a good 2 volt drop between the psu and pumps. I might just go back to a car battery.

Video:

http://youtu.be/mIb0PreP48A

Btw
Unknown bike injector #1: 290cc
CBR 06-07 1000RR injector #2: 250cc
Scion tC injector #3: 320cc
Modded Denso, 00-ish Protoge #4: 220cc -> 475cc

fuse · « **Reply #19 on:** December 15, 2011, 12:42:42 AM »

Holy shit that is so cool. That one injector at the end looks to be out flowing the other

HiProfile · « **Reply #20 on:** December 15, 2011, 10:27:22 PM »

I think mineral spirits has enough difference to toss off my raw flow numbers, but for matching purposes it will be fine. I'll just have to measure weight using specific gravity to get the correct CC/min.

DasPoop · « **Reply #21 on:** December 15, 2011, 11:52:32 PM »

I dig the roundy's water bottles. That looks bad ass. Going to have to check it out in person and bring some injectors.

HiProfile · « **Reply #22 on:** December 16, 2011, 12:12:42 AM »

Quote from: DasPoop on December 15, 2011, 11:52:32 PM

I dig the roundy's water bottles. That looks bad ass. Going to have to check it out in person and bring some injectors AND YOUR $50.

Sounds like a plan noggle.

snm95ls · « **Reply #23 on:** December 16, 2011, 12:24:41 AM »

That's some solid work noggle.

Wouldn't mind making a similar setup some day.

HiProfile · « **Reply #24 on:** December 16, 2011, 12:45:26 AM »

You'll have to wait 10 yrs to use water bottles. My patent details them and is active for that time period.

snm95ls · « **Reply #25 on:** December 16, 2011, 12:48:08 AM »

Quote from: HiProfile on December 16, 2011, 12:45:26 AM

You'll have to wait 10 yrs to use water bottles. My patent details them and is active for that time period.

I had actually planned on spending a few bucks and using plastic graduated cylinders. Notice I didn't opt for the spendy glass/Pyrex/whatthefuckever graduated cylinders.

Joseph Davis · « **Reply #26 on:** December 16, 2011, 07:57:06 AM »

Quote from: HiProfile on December 15, 2011, 10:27:22 PM

I think mineral spirits has enough difference to toss off my raw flow numbers, but for matching purposes it will be fine. I'll just have to measure weight using specific gravity to get the correct CC/min.

You can rough and sleazy it by using a control injector of a known rate.

Teg2boo · « **Reply #27 on:** December 16, 2011, 08:20:02 AM »

That's really cool! What are you waiting to build a flow bench? I would like to build one, but I have no free time

HiProfile · « **Reply #28 on:** December 16, 2011, 12:07:22 PM »

It took very little time to rig it, but forever to get it how I wanted it. And I'm still not done. Since I have ADD I can only do an hour of work per day.

I'd not recommend plastic cylinders, since most are polyethylene and gasoline or spirits will eat it up. There is a surplus science place close by, so I get "chiyrex" for a decent price. I definitely recommend getting the plastic anti-roll rings for them, that woulda saved mine.

Quote from: Joseph Davis on December 16, 2011, 07:57:06 AM

Quote from: HiProfile on December 15, 2011, 10:27:22 PM
I think mineral spirits has enough difference to toss off my raw flow numbers, but for matching purposes it will be fine. I'll just have to measure weight using specific gravity to get the correct CC/min.

You can rough and sleazy it by using a control injector of a known rate.

That's what I did before I had my injector driver. A clean 240cc, and it worked out darn close. I want to send out a handful to just get flowed so I have a good baseline to compare to.

snm95ls · « **Reply #29 on:** December 16, 2011, 12:53:51 PM »

I was attempting to make a cheeky joke, but failed.

$:-\$

News:

Author Topic: DIY Flow Bench w/VIDEO (Read 31138 times)