Injector Control Options

From DIY contraptions to sophisticated FreeEMS-specific designs! Plus general hardware development!
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Injector Control Options

Post by jharvey »

My gut feel is that 100ma sensing would be enough, this feed back is really to ensure your current is correct-ish, not as an active feed back right? Such that if you drop to low, you know you need to need to increase your PWM slightly. So you can technically set your 100ma at the low side, then after you dial something in to the range, you really don't need it's feedback unless something changes.

Also if a PSoC is used (or op-amp conditional), you can add some analog signal conditioning before the A/D, so you can scale that 8 bits to just the current range you're looking at. Such that 8 bits can range from .5 amps to 1.5 amps.

I just found this datasheet about active clamping from IRF. I'm still quite ignorant about terms "active clamping" or "bi-level snubbing" To me it looks like an OV protected drive is an active clamp, and is a different term than "bi-level snubbing". So this diode may be a redundant or an alternative snubbing approach. I'm not quite sure what to think about it. Right now, I don't see a benefit of that circuit. Perhaps it's additional benefit is that it allows you to use a larger variety of FET's, not sure. Should we pick a drive chip and design around that, or should we pick a series of drive silicon and use that. I'm tempted to say pick a OV MOSFET and design around that.

http://www.irf.com/technical-info/designtp/dt99-4.pdf

I'd like to learn more about the features you gain by using both techniques, so I can better comment about that feature.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Injector Control Options

Post by Fred »

IMO we should purely be using logic level FETs and individuals can choose whether they want cheap ones or ones with thermal protection and so on.

Current feedback wise, I'd tend to think you don't need it at all. IE, you can set it up to the parameters that you want in an open loop way. It's a nice feature to have, but probably good that it's left optional in code/hw. Turn it off in code and bridge the resistor in hw. We're looking at 4amps peak current, unlimited and 1 amp hold, I guess, 100mA detection gives you 10% which is probably close enough for government work to steal a yanky phrase ;-)

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!
TonyS
LQFP112 - Up with the play
Posts: 192
Joined: Mon Jun 21, 2010 4:18 pm

Re: Injector Control Options

Post by TonyS »

I haven't disappeared, but this circuit has required quite a bit of "thinkn".
What initially seemed like an easier design, has taken me much, much longer than I estimated. I spent the better part of today becoming familiar with the HITFET line of smart MOSFETs from Infineon, before determining that they do not have a part that can be PWM'ed at the frequencies I need.

Just wanted to let you guys know that I'm still working on it (I have to, it's part of my job).

- Huff
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Injector Control Options

Post by jharvey »

Great and keep us posted, I'm sure several of us will find interest in what you come up with.
ufotrusov
TO92 - Vaguely active
Posts: 1
Joined: Thu Dec 04, 2008 12:39 am

Re: Injector Control Options

Post by ufotrusov »

hhh
Attachments
inj_drvr_d.pdf
(12.34 KiB) Downloaded 708 times
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Injector Control Options

Post by jharvey »

To me it looks like the posted circuit has an active clamp, that is nearly functionally the same as the OV protection of an autofet. What advantages does this circuit offer that we can't buy off the shelf for a couple bucks?
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Injector Control Options

Post by Fred »

Judging by the email address, you might get a better answer if you google translate it into Russian? Just a stab in the dark, but... :-)
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
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Injector Control Options

Post by jharvey »

Per ravage thread, lets add this to the list.

http://octopart.com/vns14nv04-stmicroelectronics-335427
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Injector Control Options

Post by jharvey »

[edit]
This is outdated, see below post for updated simulation there is an error in this picture and the results are inacurate!!!!
[/edit]

There has been some discussion about the 70V OVP MOSFET's vs the 40V variety, and which is better. I did some simulations and here are the results.

The 40V, appears to dissipate about .1watts total, and the delay from when it's turned off, to when it actually off, is about 1 mS.
Image

The 70V, appears to dissipate about .5watts total, and the delay from when it's turned off, to when it actually off, is about 1/2 mS.
Image

I think the 70V dissipates .4 watts more heat because of the difference in current in the injector. I think the 40V causes the injector to dissipate the extra energy. They both have very repeatable off times, and the delays are relatively insignificant. I expected both were dissipating the same uH energy, so I expected both to be about the same. I double checked my math, and it seems to be about right. If anything, it might be .6mS not .5mS, which would cause the ECU energy to go up. Does anyone see anything wrong with my math, simulation, or anything that conflicts with empirical data? Based on this simulation, I'm still going to encourage folks to use the VNP28N04 or VNS14NV04. However if you can dump the watts at the ECU, the 70V might not heat your injected fuel quite as much, however the 40V allows for less ECU heat, and smaller packaging.

Anyone one have any thoughts about why the difference in heat dissipation? The best I can think of is that it's gotta be dissipated at the injector. Perhaps installing a snubber cap to capture that energy at the end of the pulse, and dumping it at the beginning of the pulse would be helpful. Such that it's dissipated via saturated current/RDS, not OVP transition currents. After all, we're using 14 to 28 amp drivers to drive 4 to 1 amps of current. So we could dump some extra current when it fires the next time around.

Is 12ohms about right for a HighZ injector? I estimated it by using 1 amp, at 12V.

Also note, my above simulation uses a Zener to simulate the OVP, so the real parts could be a bit different. I believe the decay curves and such would be about the same, but technically the OVP MOSFET will vary the resistance, that I've ball parked at a fixed .05 ohms.

[edit]
The picture text for the 40V (It's a bit hard to read on the picture)
Heat created during OVP phase. 12,000 RPM is about 200 cycles per second. The transition per cycle is about 1mS long, so about 200mS per second, at very fast RPM's.
40V at .01 amp is .4 watts.
So .4(.2)=.08 watts of OVP heat.
A .035ohm RDS could be used as much as .8. So 1A^2(.035)=.035watts for .8 of the time ='s .028 watts resistively.
Total heat is .028+.08=.108watts per injector channel.

The picture text for the 70V
Heat created during OVP phase. 12,000 RPM is about 200 cycles per second. The transition per cycle is about .6mS long, so about 120mS per second, at very fast RPM's.
70V at .07 amp is 4.9 watts.
So 4.9a(.12s)=.49 watts of OVP heat.
A .035ohm RDS could be used as much as .9. So 1A^2(.035)=.035watts for .9 of the time ='s .0315 watts resistively.
Total heat is .0315+.49=.5215 watts per injector channel.
[/edit]
Last edited by jharvey on Mon Sep 12, 2011 11:28 am, edited 1 time in total.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Injector Control Options

Post by Fred »

Which exact part numbers, all the units ending in N07 are 70V and the different current rating units have different RDS figures. Good research though.
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!
Post Reply