rusefi.com

Top holes are 65mm apart, bottom holes are 53mm apart.

We can also use thermal paste/glue under it in addition, and get a lot of thermal dispassion

Official https://www.molex.com/molex/products/datasheet.jsp?part=active/5007620481_PCB_HEADERS.xml has arrived - it snugs into the case quite nicely!

Plastic edge position should translate to pins position I assume?

Note how 8 thicker pins are on one side of the connector, this probably would tells us to place TLE8888 on the same side of the board?

I think this is close to the shape ?

(No components are placed or routed)

New first post inserted (I've learned how to change post date) with a link to https://github.com/rusefi/hw_microRusEfi

Open question if this crimper crimping tool would help us

https://www.amazon.com/gp/product/B00OMM4YUY

wow! that's micro indeed!

kb1gtt wrote: ↑

Mon Apr 08, 2019 5:00 pm

Pins / sockets and the crimp tool for the harness side of that connector?

Datasheet: https://www.mouser.com/datasheet/2/276/5007620481_PCB_HEADERS-179151.pdf

This 48 pin Molex 500762-0481 header mates with

8x1.5mm & 40x0.635mm terminals


0.635mm terminals are 64322 0643221029
https://www.molex.com/webdocs/datasheets/pdf/en-us/0643221029_CRIMP_TERMINALS.pdf


1.50mm CMC CP Female Terminal 64323 0643231029
https://www.molex.com/webdocs/datasheets/pdf/en-us/0643231029_CRIMP_TERMINALS.pdf
https://www.molex.com/molex/products/datasheet.jsp?part=active/0643231029_CRIMP_TERMINALS.xml


No idea if used crimp tool 90302-1 is relevant here or not.

@ says we will be using this plastic between PCB and back of the board

https://www.amazon.com/Push-Plastic-Kapton-Sheets-5-pack/dp/B00J4Q688Y
or
https://www.amazon.com/Gizmo-Dorks-Kapton-Tape-Printers/dp/B00LAJNOXK

@ is making progress

Top post now has iBom link https://rusefi.com/docs/ibom/microRusEfi_latest.html

Also some relevant goodies have arrived

So few components, it is wonderful. This may inspire a few of us to build some very small PnP RusEFI now
Very nice work

@Abricos is the daddy of pleasant surprises. Pictures is our Molex 48 plugged into 2009-2014 NISSAN 370Z COUPE Z34 OEM VQ37VHR 6MT ENGINE WIRE WIRING HARNESS LOOM

Also pictured 2010 Nissan 370Z Roadster ECU.

Next question is which is the most junk yard available Nissan or else with this connector

Junk yards here in the US do not have many 370Z but they do have many Sentras and Altimas. On Sentras and Altimas ECU harness used brown (keyed differently) version of the 48 pin only but transmission controller uses our 5007620481.

On 4 cylinder Sentra it's much easier to grab this pigtail, on 6 cylinder Altima engine bay is pretty packed. On both vehicles the connectors are exposed to elements and usually are pretty nasty

Fun fact: smaller Nissan crimped terminals/wires go into Chinese clone of the harness plug. Bigger crimped terminals thou do NOT go into harness side clones.

PS: they say https://github.com/rusefi/hw_microRusEfi is ready for first batch!

Boards are being fabbed!

We've made it all the way to e-test, they'll be in the mail soon!

looks great... my reflow oven is on standby ready for a successful prototype.

@mck1117 reports some progress!

Parts arrived yesterday, and, well...

So far everything I've tested works:

• Power supplies: 12v, 5v, 3.3v
• USB
• SWD debug interface
• Ignition outputs (5v)
• Injector outputs (which means the tle8888 is alive!)
• LEDs
• 5v/12v aux outputs (the tc4427 driven ones, not the tle8888 aux outputs)
• Flash the already-building f767 firmware, connect with Tunerstudio, configure ECU

Stuff I haven't tested:

• Other tle8888 outputs (high current low-side, 12v push-pull)
• Analog inputs (I screwed up on ordering the BOM, so I'm missing some resistors to be able to test this)
• VR/hall inputs
• Onboard NAND flash. I have the chip, but didn't install it yet as we don't even have a driver for the firmware to be able to do something with it.

Things are coming along nicely!

Nice ...

Very nice...
More pics, more vidéos and news

eBay has "100pcs AMS1117-3.3 LM1117 1A 3.3V SOT-223 Voltage Regulator" for $4 shipped, i.e. $0.04 for each. Would this replace the $1.50 926-LM2937IMP3.3NOPB from BOM https://github.com/rusefi/hw_microRusEfi/blob/master/microRusEfi.csv?

AMS1117 is cheap chinese clone ...

It is very bad copy of LM1117 ...

JRD McLAREN wrote: ↑

Thu Jun 27, 2019 7:09 am

AMS1117 is cheap chinese clone ...

It is very bad copy of LM1117 ...

Any URLs? Any specific concerns with AMS1117? Does it smell wrong? Does it produce too much heat? What is not good about it?

I have some boards https://www.ebay.co.uk/itm/DC-5V-Out-6V-to-12V-In-AMS1117-5-0-PCBA-Step-Down-Linear-Regulators-Module/173217961880?epid=19006476589&hash=item2854985798:g:e58AAOSwTmJaqjZ~

some of them burn after "higher" load, aprox 450-500mA .. or less
datasheet tell something about 800-1000mA ...

if you have "little bit" ripple input, then you have ripple output also ...
(but why ..??? it is voltage regulator .. or not .. ?? )

Using mystery Chinese parts will save maybe $3-5 per board. Absolutely not worth it for the risk of boards having issues.

This board dissipates 3-5 watts, and could be mounted somewhere that's already hot, like the engine bay, so it's going to get pretty toasty. The lm2937 is rated up to 125c - and I trust that rating from TI. That eBay one might have misbehavior that we may only see occasionally - but the 2937 is backed by TI that it will not, under the rated conditions, misbehave. For all we know, the $0.04 ones might be production rejects that went in the "trash" but found their way to eBay instead.

Also - that's our analog supply. Error and ripple on the output will directly affect accuracy of the analog inputs.

mck1117 wrote: ↑

Thu Jun 27, 2019 6:21 pm

Using mystery Chinese parts will save maybe $3-5 per board. Absolutely not worth it for the risk of boards having issues.

Yep, looks like not with the linear regulator. Gotcha.

russian wrote: ↑

Thu Jun 27, 2019 8:41 pm

Yep, looks like not with the linear regulator. Gotcha.

Don't get me wrong - I'm totally cool with it for passives, but it gets iffy for silicon. It's pretty hard to screw up a resistor, but it's very easy to screw up an IC.

Even something like the TLV75533PDRVR, which is around 1/2 the cost and considerably smaller might be worth investigating. Looks like overall better performance than the LM2937, too. Even a regular LD1117 would work very well, and is only $0.20 at qty 100.