Separating the power and signal grounds on a V3 board

Discuss MegaSquirt, VEMS and other non-free hardware and software here.
gearhead
LQFP112 - Up with the play
Posts: 120
Joined: Sun Feb 03, 2008 9:30 pm
Location: Chicago, USA

Separating the power and signal grounds on a V3 board

Post by gearhead »

I was thinking about the grounding issue with MS V3 boards. It appears from a read of Bruce's lecture notes on grounding and noise, pages 23 and 24:

http://www.bgsoflex.com/megameet2008/me ... gnding.pdf

That a properly placed saw cut could resolve the high current ground/digital ground connection that adds noise to the MS2 system. With HC8, it is not so critical as it is only 8 bit and the noise usually does not cause a problem, but with the 10 bit inputs on on the HC12 chip it does show up. I have a spare V3 board and may try this:

Cut through the board between the trace from the VB921 pin 3 and the DB37 screw lug hole with a narrow kerf saw up to the center of the hole. Make sure that you have not dragged material between the 4 layers by checking continuity between the grounds. After you have done this, then re-purpose the DB37 pin 36 as a ground.

I mean, Come on, none of us would even consider running our ignition on the ground plane, anyway, right? So, this pin should be free on the board. If you use a V2.2 board, this pin is just terminated and not connected to any thing, so a similar mod can be done to ensure compatibility with a V2.2 board, but you will not need to cut through the board. Just cut through a trace to feed out the ground from the tip42 as well as the ground you are using for the PWM idle and boost control circuits. In reality, the FP relay should also be on this power ground, but I do not think it will cause that much of a problem as it only cycles on startup and shut down.

Anyone see a problem with this?

Gearhead
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Separating the power and signal grounds on a V3 board

Post by Fred »

Image

Image hosted for Gearhead :-)
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Separating the power and signal grounds on a V3 board

Post by Fred »

One issue, even with the cut across there, all the Ground pins will still be attached to the 5v ground plane. SO, you will need much more than one db37 pin to properly ground them externally. However that's a small issue that can easily be resolved with a second plug and more spare pin usage.

Image

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
gearhead
LQFP112 - Up with the play
Posts: 120
Joined: Sun Feb 03, 2008 9:30 pm
Location: Chicago, USA

Re: Separating the power and signal grounds on a V3 board

Post by gearhead »

Fred,

That is why I suggest using the pin 36 on the DB37 as the power ground by soldering a jumper from somewhere on the power gnd trace to pin 36 and because as you show in that graphic, you cut through the trace here as well. (Pin 36 is used for the single channel ignition output according to the MegaManual and it is not connected to anything else)

Pin 17 and 16 are the ground return for channel A and 15 and 14 are the ground for channel B that you cut through. A and B will nave to be connected with a jumper between the left hole of R37 and R38 (we do not install these, anyway), then this will need to be connected to pin 36 such that the power ground is no longer on the board.

Or, this is how I read those graphics.

Gearhead
gearhead
LQFP112 - Up with the play
Posts: 120
Joined: Sun Feb 03, 2008 9:30 pm
Location: Chicago, USA

Re: Separating the power and signal grounds on a V3 board

Post by gearhead »

Or, with more thought... remove R37 and R38 or the jumpers. Connect left hole of R37 and R38. Then jump this to pin 36 on the DB37. Make sure that the PWM transistor ground is routed to the RHS of D36. No cutting required.

Glad I thought through this before I cut anything.

Gearhead
gearhead
LQFP112 - Up with the play
Posts: 120
Joined: Sun Feb 03, 2008 9:30 pm
Location: Chicago, USA

Re: Separating the power and signal grounds on a V3 board

Post by gearhead »

Just tried that. Doesn't work on the stim. It is still getting ground from somewhere.
GrowlingandBiffo
QFP80 - Contributor
Posts: 56
Joined: Sat Apr 26, 2008 7:42 am

Re: Separating the power and signal grounds on a V3 board

Post by GrowlingandBiffo »

Perhaps some one ought to redesign the MegaSquirt V3, as it sounds like it is a defective product. Are there any other problems like this?

Mistakes and inadequacies of the software are quite easy to attend to, by nearly any one but hardware redesign is a major commitment.

Of course there would be no legal or indeed IP issues, under English Law as you are effecting a repair. No copyright issues, as you are not copying.

Not to mention the fact that it all started off as an Open Source product...
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Separating the power and signal grounds on a V3 board

Post by Fred »

GrowlingandBiffo wrote:Perhaps some one ought to redesign the MegaSquirt V3, as it sounds like it is a defective product.
That is exactly what they are doing, who knows when it will come out though, but I bet a *million dollars* (monopoly) that they sell the entire stock of 3.0 boards before they release it. The 3.57 has these issues fixed by the way.
Are there any other problems like this?
See the thread below this one. Perhaps link the other thread in this one for a cross reference.
Not to mention the fact that it all started off as an Open Source product...
Well, the source code sort of did, but the hardware never was, you can look at the wayback machine and see them saying "our lawyers are better than yours" about copying the boards.

Fred.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
GrowlingandBiffo
QFP80 - Contributor
Posts: 56
Joined: Sat Apr 26, 2008 7:42 am

Re: Separating the power and signal grounds on a V3 board

Post by GrowlingandBiffo »

Fred wrote:
GrowlingandBiffo wrote:Perhaps some one ought to redesign the MegaSquirt V3, as it sounds like it is a defective product.
That is exactly what they are doing, who knows when it will come out though, but I bet a *million dollars* (monopoly) that they sell the entire stock of 3.0 boards before they release it. The 3.57 has these issues fixed by the way.
They are still making the V3.0 so it is still a production item, and so any one could redesign it or indeed modify as above, under English Law with out infringing any ones IP
Not to mention the fact that it all started off as an Open Source product...
Well, the source code sort of did, but the hardware never was, you can look at the wayback machine and see them saying "our lawyers are better than yours" about copying the boards.
Their lawyers would contact their pals in England, whom would inform them that they have no actionable case.

If it was the same it would come under "a replacement component made to like and or matching standards" Armstrong Patents vs British Leyland 1977 House of Lords

(Both defences have been used and tested on PCB's)
Last edited by Fred on Tue Oct 25, 2011 8:09 pm, edited 3 times in total.
Reason: Fixed quote tags and spacing
gearhead
LQFP112 - Up with the play
Posts: 120
Joined: Sun Feb 03, 2008 9:30 pm
Location: Chicago, USA

Re: Separating the power and signal grounds on a V3 board

Post by gearhead »

Part of the reason for no change is that Bruce does not believe that the star ground point should be on the engine, but on the MS board. Upon further looking at this, the mod suggested will pull out the drive currents off the board. The problem is Q3 the flyback pin 2 is getting ground from the 'power ground' plane on the board. The other deal is D6 which I do not know where it is getting ground. It looks like there is a trace on the bottom of the board linking it to the 'power' ground plane. To get by this, a cut will probably be needed. As for why it flashes on the stim, I am not quite clear. Once R37 and R38 are pulled, it should not flash, I would think.

edit: It will not actually open an injector, but it must be a characteristic of the stim that it actually flashes. The remaining problem is the ground for the flyback (Q3)...

Gearhead
Post Reply