Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

From DIY contraptions to sophisticated FreeEMS-specific designs! Plus general hardware development!
MotoFab
1N4001 - Signed up
Posts: 307
Joined: Thu May 29, 2008 1:23 am
Location: Long Beach CA

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by MotoFab »

Looks like a good part, and the price is too right if you factor in assembly costs compared with doing the same thing with discretes. The big difference between this part and the LM1815 is the differential input.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by Fred »

MotoFab wrote:The big difference between this part and the LM1815 is the differential input.
Jim, I'm stretched for time, can you spoon feed me the ins and outs of this please? :-)

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!
MotoFab
1N4001 - Signed up
Posts: 307
Joined: Thu May 29, 2008 1:23 am
Location: Long Beach CA

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by MotoFab »

Fred wrote:Jim, I'm stretched for time, can you spoon feed me the ins and outs of this please? :-)

Fred.
I'll give it a go.

A single-ended input includes the chassis ground, which isn't necessarily a bad thing. But the output of a VR sensor winding is an alternating sign. Tying one end of the signal to chassis ground creates an input that swings both positive and negative. This isn't a bad thing if the input amp is using a dual voltage (pos and neg) supply. A single voltage supply scheme is difficult because it requires a 'pseudo ground', or call it a 'floating bias'. With a single voltage supply (just pos and zero), the ever changing input bias can become an issue.

Using a specialized part like an LM1815 makes it easier. Sometimes is the only thing that makes it possible.

Some discrete single-supply VR signal conditioning circuits provide an inconsistent output signal to the controller. The circuit is a straightforward design when using discrete components and a dual supply. When using a single supply however, if the design isn't right, setting the bias can become a real tail-chaser. I've noticed that software must be written to accommodate the inconsistent or 'lost' VR signal. It seems to me that makes good sense only if there's no other hardware option.

A differential amplifier input isn't referenced to zero volts. And because one end of the VR sensor isn't connected to ground (zero volts), the input doesn't swing below zero volts. The input does change polarity, but isn't referenced to an external voltage (zero volts). Differential is the big winner for VR use because the bias level is set independent of the input.

Say the amp is biased to 1/2 the supply voltage. Also say that the VR sensor output changes polarity. You know, sometimes A>B, and sometimes B>A. In the case when A=B, the output of the amp is sitting right at its bias voltage. When A becomes greater than B, the output swings above the bias voltage. Conversely when B becomes greater than A, the output swings below the bias voltage.

If a subsequent comparator stage is comparing the amp output to the bias voltage of 1/2 the supply voltage, you can see that the signal to the controller will be a very nice square wave.

Besides all that goodness, a differential amp rejects common-mode voltage. Meaning that noise which appears equally on both inputs will necessarily be of the same sign, and will not get amplified. The noise doesn't change the voltage difference of the two inputs, because the noise is the same on both inputs. Mathematically, it's like adding the same number to both sides of an equation.

Boy I've been talking for a while here, heh heh. Hope that answers some questions for you, Fred.

- Jim

Edited a bit for clarity, not content.
User avatar
jharvey
1N4001 - Signed up
Posts: 1607
Joined: Tue Jun 10, 2008 5:17 pm

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by jharvey »

Good write up.

I agree that twisted pair and diff amps have been know for good things in noisy environments. One concern I might have is that with out a ground ref at either side, isn't there a chance the spinning wheel can build up a charge kind of like a Tesla coil. Do you think we can create a ref voltage or something to prevent such a possibility, or perhaps you can explain a reason why that doesn't seem to happen. I don't recall RS485 or CAN as having a ground path, but on the other hand, these aren't typically near moving parts. With the diff amp, I think we can simply ground one side to keep it at a known voltage, or known ish voltage.
MotoFab
1N4001 - Signed up
Posts: 307
Joined: Thu May 29, 2008 1:23 am
Location: Long Beach CA

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by MotoFab »

jharvey wrote:Good write up. Thanks.

I agree that twisted pair and diff amps have been know for good things in noisy environments. One concern I might have is that with out a ground ref at either side, isn't there a chance the spinning wheel can build up a charge kind of like a Tesla coil. Do you think we can create a ref voltage or something to prevent such a possibility, or perhaps you can explain a reason why that doesn't seem to happen.

I don't recall RS485 or CAN as having a ground path, but on the other hand, these aren't typically near moving parts. With the diff amp, I think we can simply ground one side to keep it at a known voltage, or known ish voltage.
LVDS signaling like CAN, have a resistor between the pair, or the bias voltage connected between the pair, or other such configuration.

About your other questions, are you talking about a discrete component conditioning circuit? Or the integrated solutions from Maxim or ON Semi?

p.s. I may not get a reply back til much later today.
User avatar
jbelanger
LQFP144 - On Top Of The Game
Posts: 387
Joined: Sat Feb 23, 2008 8:58 pm
Contact:

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by jbelanger »

I use this chip on a board I'm currently developing (the 2-conditioner version) and I still need to test it under different circumstances but there is also the advantage that it will work with a 5V or 12V square wave signal such as the one from a hall or opto sensor. That means you don't have to duplicate the input circuit for the different sensors but can use the same circuit for all. You simply need to connect the sensor to one input and leave the other empty.

I just want to add that I'm not happy if any comment I make turn out to help the design of the 5554 project (due to the attitude as I have mentioned elsewhere). But I'm trying not to be petty about it and not to impact FreeEMS because I'm happy if my comments can be of use in that case. And Jared, I think it's great that you're doing all of this but up to now your work is the only thing that has been open about that project because besides making demands and making grand declarations of what it will and won't be, there hasn't been anything concrete openly and publicly shared about this project. If you're comfortable with this then good, but don't get exploited (or bullied) for a single person's own personal benefit. And you can also just tell me to mind my own business.

Sorry for the off-topic comments. But as I mentioned, he's brought something bad to the forum. And hopefully that will be my last comment about it.

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

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by Fred »

MotoFab wrote:Boy I've been talking for a while here, heh heh. Hope that answers some questions for you, Fred.
Thanks! :-) You left out one key thing though, which part has the differential input? I'm assuming the 1815 has it, but your original statement was ambiguous in that regard and Jean stated that the new part has "another" advantage implying that it was the one with the differential input. Help.
Edited a bit for clarity, not content.
It's all good :-)

Jean, the best approach is to just ignore him/it. This makes you the better person (which you almost certainly are). One of three things will happen with that project:
  1. It will flourish into something wonderful
  2. It will stumble into something average (most likely from current trends)
  3. It will fade into a distant memory
In any case, why should we care? If it has more merit people will run to it, if it has less they will run to FreeEMS. Nature should take care of resourcing just fine :-)

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

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by Fred »

MotoFab wrote:Some discrete single-supply VR signal conditioning circuits provide an inconsistent output signal to the controller. The circuit is a straightforward design when using discrete components and a dual supply. When using a single supply however, if the design isn't right, setting the bias can become a real tail-chaser. I've noticed that software must be written to accommodate the inconsistent or 'lost' VR signal. It seems to me that makes good sense only if there's no other hardware option.
Jim, is that a reference to what I think it is? If so, can you spell it out for those who didn't notice so that it's a public record of why that option isn't the best choice? I think error handling software for RPM inputs is essential regardless, however.

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!
MotoFab
1N4001 - Signed up
Posts: 307
Joined: Thu May 29, 2008 1:23 am
Location: Long Beach CA

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by MotoFab »

Fred wrote:Jim, is that a reference to what I think it is? If so, can you spell it out for those who didn't notice so that it's a public record of why that option isn't the best choice? I think error handling software for RPM inputs is essential regardless, however.
It's one man's opinion anyway, but I was referring to the sometimes 'difficult' VR hardware of the MS board. It's documented on their forums, certainly the difficulties are known.

And sure, I didn't mean to imply no error handling, that's a funny one :lol2: I meant that error handling above and beyond the usual was written to accommodate sketchy hardware. I don't think that makes much sense. But kudos to the MS programmers that make it all work in spite of any hardware difficulties.

- Jim
MotoFab
1N4001 - Signed up
Posts: 307
Joined: Thu May 29, 2008 1:23 am
Location: Long Beach CA

Re: Inputs, Conditioning Ignition/Timing signals for CPU (56k??)

Post by MotoFab »

Fred wrote:You left out one key thing though, which part has the differential input? I'm assuming the 1815 has it, but your original statement was ambiguous in that regard and Jean stated that the new part has "another" advantage implying that it was the one with the differential input. Help.
The Maxim parts, and one of the ON Semi parts, have the differential inputs. It sounded like Jean was saying he's testing a new board using one of the dual input Maxim parts. His current dual VR board uses the LM1815 and that is a very good part too. It's been around for along time and probably will remain on the scene for a long time to come. Doing the single-ended signal conditioning with an integrated solution assures the circuit will operate properly.

I think Jean and I are a little bit soft on using Maxim parts, not because they aren't top performers, but because it's always a little bit of a crapshoot as to how long a Maxim part will stay in production. The automotive and industrial parts, like the VR conditioners we're talking about, aren't usually discontinued, because sensing for motor control and positioning remains as popular as ever. (Same with the thermocouple parts we were discussing in the other thread, they will likely be around for a long time.)

- Jim
Post Reply