Using a MOSFET driver to run the BIP373

[Peter]

I was trying to open the BIP373s gate through the supplied 330 Ohm resistors from diyautotune, and I didn't like the results I was getting. Right now I'm using a TC427COA on an adapter with a 12V/battery power suppling it, and a 220 Ohm resistor between it and the gate. I'm getting way better results so I figured I should post what I'm doing, and maybe help some other people out while seeing if anybody knows why the BIP373 doesn't really work on 5V. The three specs from the datasheet that I find interesting are:

Collector Current Limitation I_C-L = 11.5A when V_CE = 6..10V
Base Emitter Saturation Voltage V_BE-SAT = 3.2 to 4.85V when I_C = 7A & I_B = 12 mA
Input Current I_B = 200mA max

If I understand what's going on correctly, I'm surprised that anybody gets the BIP373 to work at high loads. Really it depends on the amperage you're trying to run through it, and my coil has an internal resistance of 0.7 Ohms. The running battery voltage is about 14.7, so say 14.7A through the coil, and there's no way that the BIP373 is going to work properly on 5V. Although you could probably make it open, but I doubt that all three gates would be fully saturated.

A rough sketch of my circuit:


Any ideas, criticisms, jokes, or whatever are welcome.

[jsmcortina]

Long ago I tried to drive a BIP373 from a 12V source and got very weird behaviour. It seemed that anything over 6V would put the BIP into shutdown mode. I aborted that idea and reverted to driving it with a simple resistor. No problems (for me) since.

James

[SymTech Laboratories]

Bruce noted that one should take care to ensure Vbe is sufficient, but like James, we haven't had any problems using only a resistor:

http://www.microsquirt.com/viewtopic.ph ... 87#p147112

[Peter]

Long ago I tried to drive a BIP373 from a 12V source and got very weird behaviour. It seemed that anything over 6V would put the BIP into shutdown mode. I aborted that idea and reverted to driving it with a simple resistor. No problems (for me) since.
James

I'm just going to guess that the heat sink capacity is going to have to be fairly large to successfully run 14A with 12V at the gate. With 5V it's probably running a lot less current, because the gates aren't fully saturated. I never even considered mounting the BIPs in the ecu, because I didn't want the high current noise in there. So they got mounted on the 1/8" thick piece of right angle that I mounted the coil on, and then I put the fet drivers close to the ecu. I'm curious about what you had for a heat sink, and what circuit you used.

Bruce noted that one should take care to ensure Vbe is sufficient, but like James, we haven't had any problems using only a resistor:
http://www.microsquirt.com/viewtopic.ph ... 87#p147112

It looks like to me that he has also at least setup the circuit so that the BIP is driven by 12V. I'm a little concerned about the switching time, and importantly the consistency of the switching time.
2N3906 <-pdf datasheet
Delay time: 35nS
Rise time: 35nS
Storage time: 225nS
Fall tiome: 75nS
All of these times are with V_cc at 3V
2N222A <-pdf datasheet
With Vcc at 30V. These aren't very good specs for our purposes, but they'll do since they already suck and only get worse at lower voltages.
Delay time: 10nS
Rise time: 25nS
Storage time: 225nS
Fall time: 60nS
Then you put the two together? I wouldn't expect very good results. If I was desperate it would work, but it's not the right way to drive a gate.
TC427 <- pdf datasheet
Delay time rising: 50nS
Delay time falling: 75nS
Rise time: 30nS
Fall time: 30nS
These specs are all plus or minus 10nS with a 100kHz square input. It's made to drive gates consistently.
I've also tried a TC4427, and a BTS3408G. The TC4427 works fine, but the BTS doesn't. The circuit would have to be quite a bit different, because the BTS pulls the circuit to ground. I also have some MAX4427s that I haven't tried yet, but I think should work really well.

Why are you guys so set on running the BIP gate with 5V? I know it works for some people, and I know it wouldn't even start my truck. Just because you can get away with 5V driving it doesn't mean that you should, or that the BIP is working properly. It's probably barely working at best.

[jsmcortina]

That's the circuit I tried and it didn't work for me.

James

[Peter]

I'm stunned that that dodgey circuit doesn't work.

[jsmcortina]

As soon as the "base" voltage reached 6V, the BIP went barmy.

James

[Peter]

How hot was it getting?

[jsmcortina]

It wasn't hot at all as far as I recall.
EDIT: However, this was a while ago and I don't have good records of my testing. What I do know though is that when I dumped the additional circuit and changed to base resistors it all worked great.


James

[Peter]

That's kinda strange since the datasheet shows the max input voltage at 14V. Do you know anybody that has used that circuit successfully?

[Peter]

It wasn't hot at all as far as I recall.
EDIT: However, this was a while ago and I don't have good records of my testing. What I do know though is that when I dumped the additional circuit and changed to base resistors it all worked great.
James

It sounds almost like the BIP isn't capable of supporting the amount current that your coils can use, and your solution is to drive the BIPs softly which results in your coils being driven poorly.

[jsmcortina]

The coils run great with a strong spark. (BIPs need less dwell than VB921.)
These are Ford "EDIS" coilpacks with ~0.8R primary.

Also though... "everyone" else is able to use BIPs ok, yours is the first report of a problem in the four (?) years we've been using them.

James

[Peter]

Also though... "everyone" else is able to use BIPs ok, yours is the first report of a problem in the four (?) years we've been using them.

Their posts have probably been removed from the forum. If I would have come here and said the BIP373s don't work, what would your response be? "You must be doing something wrong" Would be my guess, because it couldn't be a sub-par choice in hardware unless you remember the VB921, but I'm sure there was a good excuse for that. I found out today that the eight pins on that register have to share a total of 20mA so if you use all eight pins you only have 2.5mA per pin, and have the possibility of overloading and bricking it.

The coils run great with a strong spark. (BIPs need less dwell than VB921.)

Ya, I've heard of several people not even being able to run a timing light off of a VB921. So what you're arguing is that the MS designers originally picked something even worse? That doesn't really do anything for the BIP373 except claim that it's better than a complete pile. Have you ever put a scope on the coil to see how strong of a spark it is?

In order to extract the most out of the driver it is important to provide sufficient gate drive.

I should probably just talk to Bruce about it since he realized and tried to fix the problem four years ago.

EDIT: Grammar

[jsmcortina]

How did this suddenly become accusations of forum censorship and slagging off the Megasquirt designs?

James

[Peter]

Because it happens, and I don't like it.

[jsmcortina]

Proof required.

James

[Peter]

The only thing that'll do is get my thread deleted.

[Peter]

Not the one I was thinking of, but close. Hidden deleted same thing. I remember that a couple of engines got burned down, and the thread was like 47 pages long. The next day 5 pages or something. I'll keep looking for the screen shot of it.
http://www.msextra.com/forums/viewtopic.php?t=26366

[Peter]

502 posts


497 posts today hum I wonder where them other posts got off to.
http://www.msextra.com/forums/memberlis ... ile&u=6805


So can we get past the "accusations" since I'm sure you know better than I do what goes on here, and maybe get back to trying to improve the hw?

[jsmcortina]

So that was five years ago? I've looked in the admin log and it only goes back four years. I do recall a case where an external user had admin rights and may well have been the one who deleted some posts there.

The current admins (James + Ken) very very rarely delete posts unless they are dupes or spam. Trolls will likely get their topics locked and their user-id banned. The user Mad Menace(?) springs to mind, but even then, the posts will remain.
The MS2/Extra sequential topic was, I believe split into two topics. I do not recall deleting any posts.

Regarding the BIPs, I'm relating my personal experience. They work well in the situations I've used them in and that proposed circuit did not appear to work for me. I exchanged mail with Bruce on it and the conclusion was not to use it.

James