[Success Story] 1997 Ford F-150 5.4 E4OD

Your chance to introduce yourself and your vehicle
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

The yellow trace is set to zero offset, so GND should be the middle division. So it looks like the noise is indeed going through ground, but centered maybe 0.5v above it. Taking that into account, the max voltage must be about 5v.
blundar
contributor
contributor
Posts: 141
Joined: Tue Jan 07, 2014 4:38 am
Location: Cincinnati, Ohio
Github Username: blundar
Slack: Dave B.
Contact:

Re: 1997 Ford F-150 5.4 E4OD

Post by blundar »

Talking with dbh on slack, I have a rather strong feeling that the lack of a gate drive resistor in the circuit is to blame for component death.

6A (specified by MCP1407) is a *lot* of gate drive current for casual switching. The MOSFET doesn't specify Rgate so ASSuming it to be around 1ohm puts the gate driver into potentially dangerous current consumption around 6V supply. If the driver was not internally current limited, inrush current could be enough to make it POP maybe?

Ringing is also a real issue. I don't have enough energy to type about it. TL;DR: MOSFETs are real devices that have capacitance that can contribute to non-ideal behavior.

App note that covers it decently: https://toshiba.semicon-storage.com/info/docget.jsp?did=59456
Youtube video that was a little verbose but ok:

TL;DR: always use a gate drive resistor, even if you're using a MCU pin to directly drive the gate without a dedicated gate driver. (in this case, the gate drive resistor is what will prevent you melting the I/O pin!)

TL;DR#2: If you have EMI from voltage spikes that result from the FET switching faster than a flyback diode can "capture" the inductive kick, use a bigger gate drive resistor to slow down the transition from on to off, spreading the inductive energy dissipation over a longer time period (but at a lower peak voltage). But beware that slower switching speeds will put a higher thermal load on the switching FET and increase switching losses.
blundar
contributor
contributor
Posts: 141
Joined: Tue Jan 07, 2014 4:38 am
Location: Cincinnati, Ohio
Github Username: blundar
Slack: Dave B.
Contact:

Re: 1997 Ford F-150 5.4 E4OD

Post by blundar »

Another note:

High-speed Schottky diodes are generally a good choice for the flyback diodes for relatively low switching speeds. Reverse voltage only needs to be margin of error > Vsupply which will give a great selection of parts. They're generally pretty fast.

Be mindful of both PEAK and AVERAGE current. Average current through the diode will be fairly low because it will only see current for the duration of the inductive kick. Peak current can be as high as you can switch the FET quickly, i.e. you could have several times the average on state current of your solenoid as instantaneous flyback current.

Also note: if the flyback diode is too slow and the FET switches too fast, you can have crippling unclamped inductive spikes that have the potential to destroy the FET (if they exceed its Vds rating) or cause nasty EMI.
blundar
contributor
contributor
Posts: 141
Joined: Tue Jan 07, 2014 4:38 am
Location: Cincinnati, Ohio
Github Username: blundar
Slack: Dave B.
Contact:

Re: 1997 Ford F-150 5.4 E4OD

Post by blundar »

Part of the reason half bridges are used for higher speed switching vs. FET+diode is that the non-switched FET can act as an active flyback path, which is often termed synchronous rectification. The losses from a fully-on FET are current * Rds(on) whereas losses from diode are current * Vf. Combine that with the speed at which FETs can switch and the case for half bridges / synchronous rectification is more clear at higher switching speeds.
mck1117
running engine in first post
running engine in first post
Posts: 1268
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: 1997 Ford F-150 5.4 E4OD

Post by mck1117 »

You want a Schottky because they work via a different mechanism than plain PN silicon diodes, so they have essentially zero reverse recovery time. It's purely capacitively determined, but that current flows through the capacitor (until the diode is forward biased) anyway.
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

Thank you both for the lessons in electronics! I ordered some more drivers from Digikey. When those arrive I'll replace the one that's toast, add an 100ohm resistor, and replace the 1N2001 diode with a SB5200. That diode should work, no? It's what I happen to have. It's rated at 5A average, 20A repetitive peak.
https://diotec.com/tl_files/diotec/files/pdf/datasheets/sb520.pdf
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

Well it didn't work.
I added a 100 ohm resistor to the gate, and replaced the 1n4001 with a SB5200. It failed in the same manner as the first one. The gate driver outputs about 4.5V, no matter the state of the input.

Theories:
1: My MOSFET is somehow broken. I tested it on the ATX power supply I use as a bench supply, and it switched a LED on and off successfully. I would guess that it could fail in ways that this test doesn't catch, but IANAEE.
2: Arcing between solder blobs.
3: Maybe the diode isn't fast enough?
4: Maybe the 22ga wire I'm hooking this all up with isn't skookum enough for the flyback diode to keep the voltage drained something something

So what am I going to do? Give up and hook up my Proteus, of course.
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

I changed my mind about giving up, and decided to try again. New gate driver, new MOSFET.
And I found a problem: I had the diode wired to ground instead of 12V. I don't think that would kill the gate driver, but I decided to give it another try anyway.
It worked perfectly - for maybe a minute. Then the gate driver bit the dust again. I felt the gate driver with my finger probably 15 seconds in, and it wasn't getting warm.
I have no idea what's going on, but judging from the history of this project, the answer is most likely that I am an idiot and have something wired wrong.
mck1117
running engine in first post
running engine in first post
Posts: 1268
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: 1997 Ford F-150 5.4 E4OD

Post by mck1117 »

How many more gate drivers do you have? If it's a few, I'd replace the gate driver, and remove the mosfet entirely. Scope the driver output, turn it on, and see what happens.
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

I'm fresh out. Time for my nth Digikey order I guess.

In other news, I've got all the pins crimped for a Proteus adapter harness, just need to stick them in the connector.
Here's the pinout and Proteus mapping if someone wants to look it over. My pinout is column E.
https://calc.holdemanenterprises.com/eecv-table
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

Sooooo....
image.png
image.png (30.48 KiB) Viewed 2729 times
See those capacitors on Vdd? Turns out that following the circuit in the datasheet is a good idea.
Thanks to @blundar for suggesting that this was probably the problem.
It was.

Interestingly, the behavior of the trans line pressure is completely different than when I was driving the MOSFET with a bare digital output. 50% does roughly what 90% did before. I say roughly, because the pressure now varies more with RPM. I haven't got that one thought through yet.
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

Pic for context:
IMG_20210706_150942.jpg
IMG_20210706_150942.jpg (1.85 MiB) Viewed 2614 times
Notably absent are the block and crank, which are off at the machine shop.

- junkyard dave
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

Block and crank have returned
IMG_20210716_170008.jpg
IMG_20210716_170008.jpg (1.75 MiB) Viewed 2526 times
User avatar
AndreyB
Site Admin
Posts: 12794
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: 1997 Ford F-150 5.4 E4OD

Post by AndreyB »

Any updates, how is it going? :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
dbh97
contributor
contributor
Posts: 82
Joined: Sat Jul 19, 2014 10:43 pm
Location: 67867
Github Username: chuckwagoncomputing
Slack: dbh97

Re: 1997 Ford F-150 5.4 E4OD

Post by dbh97 »

Current status of the new engine is that the block is painted and the crankshaft is installed.

I haven't been working on it much because some other stuff has taken priority. I agreed to swap the transmission in my parent's car with a junkyard pull. I drained the fluid from the junkyard pull and it looked even worse than the fluid from the original transmission, so they decided to have me rebuild it instead. Add procrastination and depression to this picture, and I've had their car for about 2 years now. I've decided that I must finish it before I tear into my pickup again - which needs to be done before I can do much in the direction of making the transmission shift itself, because this transmission has a pre-existing condition or two.

Another new development to add to the picture is that the pickup blew a coolant hose and it was at night and raining so I didn't notice for a mile or two, and now it's using coolant and the coolant tank smells like fuel, so it blew a headgasket and/or cracked a head. It still runs though, and I still drive it quite often when my wife is using my one reliable vehicle.

My plan right now is to rebuild the transmission, and get the engine running on rusEFI at the same time.
Then develop the code to get the transmission shifting properly.
Then (probably quite a while later) swap the engine out. This step will take quite a while as I'll have to custom-build all the turbo plumbing.
Post Reply