PDA

View Full Version : problems with FAV solenoid - no work, but getting very hot! ::fixed::


softic
07-06-2012, 08:02 PM
After over 7 months of flawless function my HFS-3 kit seems to have problem now - there is no flow in the system ((((
Symptoms are following:
There is no flow in the system under boost The FAV solenoid gets very hot instantly after system power-up (broken schematics??) The flow sensor shows full bar under boost although there is no flow at all The impedance if my solenoid compared to new one is the same - 12.5 Ohm (I did not try to install a new solenoid as I am afraid of braking it)


Definitely I will try installing backups FAV and its solenoid to test if there is issue in them, but if schematic of the main board is broken could it probably damage the new solenoid?
Also the FAV solenoid says about 6V DC operating voltage while HFS-3 documentation says about 12V/1A output - is it normal?

Should I anyhow check the actual output on FAV solenoid wiring?

Richard, I need your help, please....

softic
09-06-2012, 05:31 PM
ok, some new info:
the voltage on FAV solenoid is always around 13,6V I have tried installing a new FAV and solenoid assembly (from second kit) - no luck at all tested the pump - it works perfect (when disable jumper FAV and enable jumper SYS)


please, help!! how can I test that the mainboard is not broken?
today it worked for few minutes but later on stopped again ((((( should I check voltages on the mainboard diag pins?

Richard L
09-06-2012, 08:34 PM
I have been thinking about your problem for a while.

I could not think of anything obvious. It appears the power transistor on the main board is blown. Before that, I would like to try a few questions with your help.

1. When you first switch on at pre-crank key position, does any one of the led on the controller lights up?

2. When you pull off the FAV jumper, does the FAV still gets hot?

softic
10-06-2012, 09:14 PM
Richard,

I have did both checks and seems like your guess is right...

1. At pre-crank key position the water level light blinks for a second and then switches off.
Here you will see he result of check: http://www.youtube.com/watch?v=fFQw5h9UQGQ
2. As to the FAV jumper - regardless of its position there is constant 13.9V output on the wire.

Regards,
Alexey

softic
10-06-2012, 10:06 PM
Okay, seems like this case is resolved - I took my HFS-3 to home and then in a bright light saw the blown up MOSFET transistor.
I still can't believe this has happened... why? shortcut on the FAV wire?

below is the picture of schematic with a clearly seen melted plastic.
how can I check that it is really dead one?

http://aklepikov.ru/wp-content/uploads/2012/06/Moskva-20120611-00016.jpg

can it be replaced by any other high-voltage/frequency MOSFET transistor?
seems like it is quite hard to get exactly this one in Russia - what are the key parameters to look for (I would buy new one tomorrow)?

Richard L
11-06-2012, 12:25 AM
The most obvious cause of a blown transistor is a short circuit. Other possibility is static electricity.

The transistor used is quite common, made by STMicroelectronics. STN3NF06L. It is good up to 4A. Logic level gate drive. You should not try to put a more powerful device in there. The next short-circuit means a burn out circuit board, not that easy to repair.

Here is the specification:

- Device name: STN3NF06L
- Vdss....... 60V (minimum)
- Id ............. 4A (minimum)
- Vgs(th) ............ 5V logic level
- Device outline package .............SOT-223
- Rds < 0.1 ohm.

You can search google for availability in Russia.

softic
12-06-2012, 04:31 PM
Richard,

I have replaced transistor with a new one - but right after switching mainboard on it has burnt again.
What are other areas to check? Any voltage reference points to watch for?

Also recently my car was on a global maintenance and potentially someone in the service could brake something in the system setup?
I am really very much worried now as transistor replacement did not help (((

Regards,
Alexey

Richard L
12-06-2012, 04:53 PM
Unplug the red harness and test it again. It will identify where the short circuit is. Cable or within the circuit board.

Richard

Richard L
12-06-2012, 09:02 PM
Can you check the 12v supply to the unit is not wired to the ignition coil (+).

softic
12-06-2012, 09:26 PM
what is the red harness?
how can I find what causing the short circuit if the transistor is already dead?

softic
12-06-2012, 09:30 PM
Can you check the 12v supply to the unit is not wired to the ignition coil (+).No, it shouldn't be because I plugged this wire to the general power supply.
But I will check again, because after recent service which included whole rebuild of cylinder block anything can be changed...

Richard L
12-06-2012, 09:42 PM
what is the red harness?
how can I find what causing the short circuit if the transistor is already dead?

The red harness is the one that goes to the FAV.

You need to change the transistor first. It should not cost more than $1.00 a piece.

Richard L
12-06-2012, 09:45 PM
No, it shouldn't be because I plugged this wire to the general power supply.
But I will check again, because after recent service which included whole rebuild of cylinder block anything can be changed...

Good. This leaves out one possible problem.

softic
13-06-2012, 10:22 AM
The red harness is the one that goes to the FAV.

You need to change the transistor first. It should not cost more than $1.00 a piece.Richard, the FAV harness was actually disconnected!
I did not connect it to the solenoid before my tests - and it immediately showed 13.0V on the harness... seems like transistor was blown immediately after I start the system.

p.s. and you won't believe, but they sell such transistors here in Russia for $3.00 (while purchase from manufacturer at $0.75 - I know the real price)!

Richard L
13-06-2012, 01:49 PM
Next step...

Can you check the voltage of the controller ground (bottom left of the test points), against chassis ground.

softic
13-06-2012, 03:42 PM
sure, will do it later today once I'm back home.
should I measure impedance between ground and accumulator (-) chassis as well?

Richard L
13-06-2012, 06:37 PM
Just to measure the controller ground against chassis ground. This is to check the system wiring.

Chassis ground is any exposed metal on the car body.

What is the accumulator?

softic
14-06-2012, 07:44 AM
the voltage on 'ground' reference point is 0.00V when off and 0.01V when switched on.
the impedance is 0.0 Ohm when off.

sorry for using wrong term: by (-) accumulator I mentioned (-) on car's battery which is linked to chassis on my car.
what are next steps? I see lots of diagnostic points on the mainboard - should I check them?

also I have checked again the information from fuel injectors (I have direct injection car) - it is correct and there are no problems here.
the same about information from flow sensor and trunk harness...

Richard L
14-06-2012, 08:11 AM
This is good, no wiring problem there.

Please check the voltage of the following test points (against controller ground):
(with the ignition on, engine not running)

1. test point 5 injector(+)
2. test point 8 injector(-)
3. test point 23 injector(+12)
4. test point 1 FAV drive (check resistance. not voltage)
5. test point 16 (IDC%)
6. test point 15 internal 5V reference 1
7. test point 11 internal reference 2
8. test point 30 internal reference 3

That is all for the time being.

I would also like to have a good picture (well focused) of the controller circuit board (top side and bottom side). Like to see the settings and configurations.

softic
14-06-2012, 12:41 PM
Is it okay if I will test all reference points without FAV solenoid connected?
Since it is always under voltage it is getting hot which is not good, I assume - ok?

Also, what to you mean by 'test point 15 internal 5V reference 1' - what is reference 1?
Should I check all voltages and impedance (resistance) vs. 'ground' control point on the mainboard?

softic
14-06-2012, 03:00 PM
ok, here are the answers:

1. test point 5 injector(+) +13.9V
2. test point 8 injector(-) +13.2V
3. test point 23 injector(+12) +13.9V
4. test point 1 FAV drive (check resistance. not voltage) 650-680 Ohm
5. test point 16 (IDC%) from 0.2V on idle up to 1.2V on load
6. test point 15 internal 5V reference 1 +4.97V
7. test point 11 internal reference 2 +9.07V
8. test point 30 internal reference 3 from +3.35V on idle to 7.66 on load (after the threshold is reached and FAV should be opened)

this all was on running engine with LED gauge switched on.
sorry, I forgot to take results on switched off engine, because when engine is OFF the LED is also OFF... will make new measurements 1.5 hours later.


the bottom of the mainboard:
http://aklepikov.ru/wp-content/uploads/2012/06/IMG-20120614-00021.jpg

the top of mainboard:
http://aklepikov.ru/wp-content/uploads/2012/06/IMG-20120614-00019.jpg

Richard L
14-06-2012, 05:37 PM
There is enough information there for me at the moment, no need to make more measurements.

As far as the functions of the system, everything is perfect. The only exeption is the pin1(FAV) drive. there shoud not have a resistance of 650-680 ohm.

Is this test done on a blown transistor or a new one?

softic
15-06-2012, 04:07 AM
Richard, the test was done on a blown transistor.
I can replace it with a new transistor, but I am 90% sure it will blow up even before I connect FAV solenoid (((

Richard L
15-06-2012, 07:40 AM
I really cannot understand what is going on, everything tested normal.
Can you remove the blown transistor and do the same test #4?

Richard

softic
15-06-2012, 08:05 AM
sure I can! should I just remove transistor, not replace with a new one?
I think that by removing the transistor this would imitate the 'closed transisto / FAV-off' situation.

Richard L
15-06-2012, 01:38 PM
I like to see if the resistance is still there when the blown transistor is removed.

softic
17-06-2012, 12:38 PM
Richard,

I have just checked test pin #1 (FAV control) after removing broken transistor and have following results:
- when power is OFF, but ignition key is ON, the impedance is 8.5 MOhm, which should be fine
- when the power is ON - the voltage on idle is 0.1 and on load / throttle up to 1.4V

Let me know, what are my next steps.
Maybe I can connect some external transistor based solenoid driver? I assume you use PWM, right?

Regards,
Alexey

Richard L
17-06-2012, 03:04 PM
I have a question on pin 1 test: Is the FAV+ cable plugged into the controller during your test?

The strange part of this test is: pin1 should not see any voltage until the FAV is plugged in. When it is pin one should see 12V (battery voltage) all the time, regardless of idle of load.

We might have found something.

The next test:
(still without the transistor)

There are three pins for the transistor, can you check all three. The center pin is the same as test pin 1. Theone on the left is power ground (black molex wire). The one on the right is PWM drive signal. At idle it should read below 1V and over 5V at full load. This is donew with the RJ connectors facing away from you.

softic
17-06-2012, 03:58 PM
Richard, the FAV harsess was disconnected at all tests.
Also I do not connect this harness to a FAV solenoid to reduce any extra load on the system

What is our next test?

Richard L
17-06-2012, 04:48 PM
The strange part of this test is the ~12V signal on pin 1. It should not be there when the FAV is not plugged in. It is especially strange it changes between idle and loaded.

Can you rememer if there is any sudden change conincide with system problem. Other than the engine rework. Are any of the wires been disturbed.

The next step would probably be unplugging all the RJ connectors (including the gauge), just leaving the grey harness plugged in.

I would like to concentrate in the test point 1 with only grey harness plugged in. Then retest pin 1 as follows:

1. Just the molex
2. Just the pump harness (blue)
3. Just the FAV sensor harness (yellow)
4. Lastly the gauge.

I like to know where this strange 12V comes from. If the 12V is present before with only the grey harness plugged in, we have found the culprite.

softic
17-06-2012, 05:20 PM
oooh... Richard, I am very sorry - I was typing my replies from Blackberry mobile and mistyped the voltage!
The voltage on PIN 1 is 0.1V on idle and from 0.5V to 1.4V on throttle after threshold is reached (I lowered down the threshold for more easy testing).

So there is NO 12V, but 1.4V maximum... sorry for this misleading typo:o:o:o
What to do in this case?

Do I get it right, that you use PIN 1 as transistor opening control voltage?
Or it is something else?

Richard L
17-06-2012, 06:00 PM
OK, delete the other tests on my last email. Pin 1 is the PWM switch to ground by the transistor.

So pin 1 is 1.4V maximum after threshold.

Can you check the rest of the unsoldered pads on the circuit board where the blown transistor was. Voltage of pad 1, 2 and 3. 2= test pin1

http://www.aquamist.co.uk/HFS3/H3-forum/mos.jpg

softic
17-06-2012, 06:36 PM
Duplicate message from mobile (((

softic
17-06-2012, 06:36 PM
Ok, if all harnesses connected and solenoid is connected as well - I see 13.8V on trnsistor pin #2, 0V on tr. Pin #1 and some small (looks like transistor control PWM voltage) 0.2-1.8V on trnsistor's pin #3 (right one on your PCB picture)
What if I will just connect transistor's pins 1 and 2?
I assume this will power the solenoid and open the valve constantly, right? Should I check this?

Richard L
17-06-2012, 08:15 PM
If you connect transistor pin 1 (power ground) and pin 2 together, the FAV will switch on. You can check this.

Can you check the black wire of the molex is connected to chassis ground securely.

softic
18-06-2012, 07:10 AM
Richard,

When I connect pins 1 and 2 - I got system fully functioning, with maximum flow.
Also I have checked both black and white wires on molex - there is 0.1Ohm impedance between them and chassis ground.

I can try to solder one more transistor and check results.
Also, when I shortcut pins 1 and 2 the system works on full load for few seconds and then the 'water level' warning appears - why?

Regards,
Alexey

Richard L
18-06-2012, 08:13 AM
The system ground is wired very well.

As far as the signal tests of the controller is concerned, there is no reason why the transistor should blow.

If you bridge 1 and 2, it produces full flow. But if this flow is outside the WL and WH, it will trigger the fail safe, hence the yellow led comes on. For the time being, just bridge the DFS jumper (disable failsafe) until we have sorted out the transistor.

What is the gauge bar reading when you bridge 1 & 2?

softic
18-06-2012, 09:14 AM
Richard,

Gauge bar shows all 8 bars when pins 1 and 2 connected and threshold is exceeded.
Also sometimes it shows 1 bar on idle, but most likely this is because of vacuum pulling methanol from the opened FAV.

Any other checks before soldering (and probably burning) a new transistor? :-)

Richard L
18-06-2012, 10:55 AM
If you see 8 bars across, you would have tripped the WH failsafe. I thought you would know that already over seven months of use.

When you do the test, you must put a nozzle or two and calibrate the SC to suit.

I still do not know why the transistor failed. Let me have an update. I suggest putting a new transistor in. Before careful. Mos transistors are very sensitive to static electricity. Make sure the soldering iron is grounded.

softic
18-06-2012, 11:17 AM
good note about grounding - will do it this time.
actually I have checked resistance between pins 1 and 2 last time after soldering a new transistor - it was okay... can it be because of too big current?
I estimate the current through the transistor should be no more than 1A (assuming we have 14V on it and 16Ohm load on solenoid).

so I will solder a new transistor this evening and will check the result.

Richard L
18-06-2012, 12:42 PM
The FAV draws about 1A at 100% DC.

The transistor will take 3A quite easily. After soldering the new transistor, power the system up without the FAV connected. If it doesn't blow, then plug in the FAV.

softic
18-06-2012, 10:21 PM
new transistor soldered with soldering iron and my hands grounded.
test pin 1 vs. ground pin resistance is around 16MOhm now, which seems to be correct.

tomorrow will do the test - in which sequence should I plug harnesses to find the problem?
first DI (user) and then FAV, right? or just FAV right after molex? anything to test after just molex connected?

sorry for soooo many questions - I have another track day on Thursday and want to fix my HFS-3 kit before I go there :o

Richard L
18-06-2012, 10:34 PM
Connect the FAV last. Keep checking the pin1 and ground.

We are very close to get it to its original state. If everything works, not sure why the transistor failed.

I am here to help. I am very glad you are able to do all the electronic work.

Richard L
18-06-2012, 11:25 PM
One thing you might like to check, the FAV cable is not damaged along the length.

softic
19-06-2012, 07:14 PM
ok, checked with new transistor.
all worked fine for some period of time, but after few minutes the system started behaving in very strange way:
- the flow sensor showed some flow in accordance with current fuel IDC
- the pump motor in the trunk worked only for 1 second every minute
- the actual temperature on the intake was not reducing so I assume there was no actual flow (((

this all is very strange because right after installation of newly soldered the FAV worked perfect and there was a REAL methanol flow causing the real intake temperature decrease (from 40 deg down to 3-5 deg) and motor of the pump was working constantly during flow (I hear its noise which is quite loud)

now I took my mainboard back to home, disconnected it and measured impedance of some reference points:
- resistance between mainboard test pin #1 and ground is around 12MOhm
- resistance between transistor pins 2 and 1 is around 12MOhm
- resistance between transistor pins 2 and 3 is around 10-12MOhm
- resistance between transistor pins 1 and 3 is around 4.6kOhm (which I assume is very bad)

I'm getting a bit crazy of all this, especially when the system shows high methanol flow (up to 8 bars) while actual flow is not as big (if there is any).
Should I try buying and using electronic oscilloscope to check for some more diagrams?

Richard L
19-06-2012, 09:49 PM
4.6Kohm between 1 and 3 is normal. It appears everything is looking good. You might have some air in the system or the FAV seal is playing up

I suggest purging the system, manual testing the system on the windshield. Please go one step at a time.

softic
20-06-2012, 05:36 AM
good point, I will do both checks!

Richard L
20-06-2012, 08:47 AM
I am staying with you as long as you need me.

The system has been off for a while with other works in the car. So good a good system check is advisable (mechanically). May be all you need to do is Purging the system and do a windshield spray test.

softic
21-06-2012, 09:40 AM
Richard,

I am very glad and thankful for all your support during these days!
Finally the system seems to be functioning and I really see the water flow / temperature drop!

Today I have a very long circuit race event (3 hours track-day), so I will have all chances to test the fix.
After that is all is fine - we close this issue.

Thanks!
Alexey

softic
22-06-2012, 06:00 AM
OMG, the result of properly working FAV is just amazing!
last week the lap time without WMI was 1:21.8, but yesterday it was 1:16.0 - very good achievement!

Richard, I would like to thank you for all your help!
the result I have achieved is worth of all the pain with diagnostics - you are the best!

Richard L
22-06-2012, 04:43 PM
I am very glad the system is now working and reduced your lap time. I would also like to thank you for being very patient in staying with me tracking down the problem.

Able to solder, tracking down transistor is a real asset.

I still would like to know what actually caused the blown transistor. The most obvious reason is a intermittent short circuit along the FAV harness. It might even be the the little whisker of wire strand crossing over inside the RJ connector.