Another ECU...

Hardware inside and outside of the ECU
Horizenjob
Posts: 45
Joined: Mon Jun 22, 2015 8:31 pm
Location: Massachusetts, USA

Re: Another ECU...

Post by Horizenjob »

The signal swing is what I was working on. The above trace shows the signal with a pair of diodes for voltage clamps. It seems better to do the clamp off the chip for the reasons you mention, it just doesn't make sense to dump current into a little chip.

These diodes gave a .9V signal swing and that is at 550 RPM. I was thinking of using a pair of diodes in series to give more voltage swing, but this may be enough.
Rhinoman
contributor
contributor
Posts: 256
Joined: Thu Sep 24, 2015 2:14 pm
Location: Wiltshire, UK

Re: Another ECU...

Post by Rhinoman »

You must do the clamping off-chip, the datasheet is quite clear that you should not put a negative voltage (<-0.3V) into the input of the differential amplifier; the behaviour of the output is not specified for voltages outside the common mode range. A diode clamp to rail isn't going to be adequate because the voltage drop across it is too high.
Horizenjob
Posts: 45
Joined: Mon Jun 22, 2015 8:31 pm
Location: Massachusetts, USA

Re: Another ECU...

Post by Horizenjob »

I should be more clear. The diodes I am using to clamp are between the signal lines, they limit the differential voltage to a diode drop, it will never exceed something like 1.2V and maybe not reach even that.
Rhinoman
contributor
contributor
Posts: 256
Joined: Thu Sep 24, 2015 2:14 pm
Location: Wiltshire, UK

Re: Another ECU...

Post by Rhinoman »

Horizenjob wrote:I should be more clear. The diodes I am using to clamp are between the signal lines, they limit the differential voltage to a diode drop, it will never exceed something like 1.2V and maybe not reach even that.
Ah, I understand now, however that stills leaves the common mode issue, how are you off-setting the input voltage?
Horizenjob
Posts: 45
Joined: Mon Jun 22, 2015 8:31 pm
Location: Massachusetts, USA

Re: Another ECU...

Post by Horizenjob »

I'm using the Maxim 9926, so am expecting the chip to reference the common mode voltage. It appears to have an internal 100K resister setting the positive leg to the internal bias voltage.

It doesn't seem like a very strong connection to the bias voltage though. Do you think I should use external parts to generate the bias voltage and use that thru a smaller resistor to set the common mode voltage?
Rhinoman
contributor
contributor
Posts: 256
Joined: Thu Sep 24, 2015 2:14 pm
Location: Wiltshire, UK

Re: Another ECU...

Post by Rhinoman »

Horizenjob wrote:I'm using the Maxim 9926, so am expecting the chip to reference the common mode voltage. It appears to have an internal 100K resister setting the positive leg to the internal bias voltage.
It doesn't seem like a very strong connection to the bias voltage though. Do you think I should use external parts to generate the bias voltage and use that thru a smaller resistor to set the common mode voltage?

The datasheet states the Maximum input range for the pin is -0.3V - VCC+0.3V so the internal diodes are clearly for short term protection only, I would expect that any current flowing through them is likely to cause jitter. The operating range is given as:

INPUT OPERATIONAL AMPLIFIER (MAX9925/MAX9927) Input Voltage Range IN+, IN- Guaranteed by CMRR 0 to VCC +0.3V

I haven't read the datasheet fully so that may be at the pin or at the differential amplifier itself. I have an evaluation board here somewhere, from memory it has a jumper to apply an external bias voltage; it would be worth looking for the datasheet for the dev board.
User avatar
abecedarian
Posts: 386
Joined: Fri Nov 15, 2013 10:49 am

Re: Another ECU...

Post by abecedarian »

Not to knock anyone here, but I think a good, hard look at the datasheet for the MAX992x chips as well as a deep study of the evaluations kits might be in order.
You can lead the horticulture but you can't make them think.
Horizenjob
Posts: 45
Joined: Mon Jun 22, 2015 8:31 pm
Location: Massachusetts, USA

Re: Another ECU...

Post by Horizenjob »

Not to knock anyone here, but I think a good, hard look at the datasheet for the MAX992x chips as well as a deep study of the evaluation kits
I really have read the datasheet. Honest. Several times even. There is not much info in the evaluation kit information, so not so sure how deep it can be studied.

Abecedarian, is there something specific you're trying to say?

Maxim expects to do the clamping on chip, I'd rather take it off... If it doesn't work it's not like there isn't a fallback position. I'd just like to build a rugged ECU and if it costs $1 more to ruggedize a VR channel that's not bad.
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: Another ECU...

Post by kb1gtt »

I see the eval kit uses the 4X 5kohm resistors, very similar to what I did. I did this to get a higher working voltage. I see they used 1206 instead of 0805 though. That might be of interest. I used 0805 as they are pretty close to a 100% survival rate when thermally cycled. While 1206 and larger have about a 50% survival rate. The 1206 depends on what it's made from, if it's made from something with a thermal expansion rate that's similar to FR4, then it survives, if not then it fails. Seeing this board, I'm tempted to consider 3X of 3.33kohm 0805's. I would likely suggest a new design consider the same.

The datasheet doesn't say much about how much current is acceptable to flow through the clamping diodes. The 10kohms at 300V would conduct about 30mA, which is about 9W of energy to be dissipated and about 0.03W of that is dissipated by the MAX diodes. You would need a really hot VR at really fast RPM to generate 9W of energy, and even if you did have such a hot VR, not much of the energy is dissipated by the MAX chip. Most applications will only see about 50V, and the above test VR was generating 0.4mW at lower RPM. I think the big concern is if you have more than a 30mA load on the +V rail pulling it down. If you don't you run a potential over voltage risk. Horizenjob is attempting to avoid that issue by using clamping diodes in paralleled with chips inputs that clamp before hitting the 5V rails. This prevents that energy from being dumped to the + and gnd rails.
Welcome to the friendlier side of internet crazy :)
Horizenjob
Posts: 45
Joined: Mon Jun 22, 2015 8:31 pm
Location: Massachusetts, USA

Re: Another ECU...

Post by Horizenjob »

At the very top of the datasheet Maxim claims 40 mA max for IN1+, IN1-, IN2+, IN2-.

It's nice the Maxim part is auto qualified. They don't mention ESD in the datasheet though. The TVS I am using on ECU inputs are 0603 SURGX and cost about $0.09. OEMs avoid this cost, but I don't think it's a high cost for an aftermarket unit especially considering that it's inputs may be handled a lot more than a car that works for 200,000 miles after it is sold.

Those TVS don't want to see the high voltages possible from the VR though, so clamping would be nice. It also removes concerns about the input resistors and the wiring.

It is also worth some effort to reduce parts count. I'm only using 0603 10K resistors at the moment. The ones I am getting are pulse and ESD capable and rated 1/4W. The resistors in the eval kit are quite large, but not ESD rated. I don't have the experience to really know where to draw the lines here.
Post Reply