PIC versus comparator/op amp hardiness

[Thunderchild]

I'm designing a circuit which may work at some temperature, I am making its main crontrol with a pic as it reduces component count and is more flexible and tamper proof. However a comparator chip or op amp might be able to do the same job.

At the end of the day this has to be a robust design so which of the two options will be most suitable ? a pic can work up to 125 C so it should be ok but then I have found that most component specs for temperature tend to amount to: at this temperature the thing will survice alone but its usaility under these circumstances is negligeable. the pic will be having but a few mA going through it so won't make much heat in itself

 

[Nigel Goodwin]

If the job is easily done with a simple opamp, then use it - there's less to go wrong.

 

[Thunderchild]

well the PIC will be more flexible and requires less components, this is for my vintage car regulator project, if I am to build a number of them I can easily alter current limitation values by altering the program, with a comparator I will need to ajust small trimmers and have the trimmers in the first place, basically just how durable is a pic ?

 

[Nigel Goodwin]

PIC's are VERY durable - now you've actually said what it's for, I would say go for a PIC.

However, make sure you apply suitable in-car supply filering and protection, no matter what you use.

 

[Mikebits]

A PIC or comparator circuit will only be as good as the reference voltage over temp. Have you looked at the PIC ref V at 125C? Does the PIC allow for an external REF V?

 

[Thunderchild]

Thanks Nigel, being a beginner in pics and knowing very well just what a pain computers can be for stability i was just a bit worried, being a car power regulator if it fails the car may stop short of of power from a drained battery or worse if regulation fails to happen 40 V can be put into the car

 

[Thunderchild]

A PIC or comparator circuit will only be as good as the reference voltage over temp. Have you looked at the PIC ref V at 125C? Does the PIC allow for an external REF V?

The ref voltage will be the 5 V supply, so its a case of how stable the 78L05 reg is, a slight variation of output will not be an issue (give or take 0.5 V). having checked the TL082 datasheet I think its made for 70 C versus the 125 of the PIC.

 

[Mikebits]

If accuracy is important, you are better off not using supply voltage as reference voltage. The Ref voltage will give a better threshold accuracy regardless of Vcc. And I am confused, if you need to operate at 125C how will a part rated at 70C survive, or is regulator in cooler environment? At any rate precision references are cheap and will ensure a more robust design, I think anyways.

 

[Nigel Goodwin]

If accuracy is important, you are better off not using supply voltage as reference voltage. The Ref voltage will give a better threshold accuracy regardless of Vcc. And I am confused, if you need to operate at 125C how will a part rated at 70C survive, or is regulator in cooler environment? At any rate precision references are cheap and will ensure a more robust design, I think anyways.

I would always tend to use an external reference - but in this case consider what it's for, replacing a crude mechanical regulator - it's not a precision application.

 

[MrAl]

well the PIC will be more flexible and requires less components, this is for my vintage car regulator project, if I am to build a number of them I can easily alter current limitation values by altering the program, with a comparator I will need to ajust small trimmers and have the trimmers in the first place, basically just how durable is a pic ?

Hi,


Not necessarily true, as you can put an external comparator in the
loop just as well as the internal comparator simply by sampling the
output of the comparator, or use it to trigger an on change pin event.
Thus, the external can be used very close to how the internal works,
which means programming can change features.
There is a catch of course, and that it requires more pins, and
possibly an extra pin for a simple D to A converter (PWM) to adjust
the comparator threshold. The internal comparator adjustment isnt
that good anyway, with the available voltage levels being very limited
to very unusual step increments.
What i dont know is just how much the internal comparator offset
voltage drifts with temperature, while with the external one you can
select just about anything out there.
In any case, with a design like this usually some rigorous real world
testing is in order...build a small test oven to test the final design.

 

[Mikebits]

I would always tend to use an external reference - but in this case consider what it's for, replacing a crude mechanical regulator - it's not a precision application.

I think I missed that post about the vintage regulator thingy. Thanks for pointing that out.

 

[Thunderchild]

As I'm using the original regulator case heat dissipation will be a problem so I'm trying to design for a hot environment and use componenets that will make as little heat as possible.

i will be carrying out real world testing at my friends house, he is going to rig a dynamo to a lathe and we will run it all set up like it is in the car, I will try and measure the internal temperature whilst putting the system under load. I could also integrate thermal protection with an LM335 to lower output if the temp goes high but as a last resort (there are some nutters that think racing in very hot weather is fun)

 

[Mikebits]

Wait. you must explain this dynamo to lathe thing. How big is this lathe? This is a real car right?

 

[Nigel Goodwin]

Wait. you must explain this dynamo to lathe thing. How big is this lathe? This is a real car right?

I think he's just using the lathe to test the dynamo, it's easier than doing it in a car, and the lathe allows you to change speeds.

However, it's not a very good test - and proper testing on a car would be required as well - but for preliminary testing it's not a bad idea.

 

[3v0]

The ref voltage will be the 5 V supply, so its a case of how stable the 78L05 reg is,.

There are many regulators engineered for use with the harsh automotive electrical environment. I highly recommend you use one of them. They are priced about the same as other regulators.

Try regulator automotive in your favorite parts suppliers search engine.

regulator automotive

3v0

 

[Thunderchild]

I think he's just using the lathe to test the dynamo, it's easier than doing it in a car, and the lathe allows you to change speeds.

However, it's not a very good test - and proper testing on a car would be required as well - but for preliminary testing it's not a bad idea.

yes thats right, as a dynamo out of control can develop 40 volts we decided to set up a test bed to make sure it works before putting it into his car

 

[Mikebits]

But a lathe will hardly have the torque of a car engine, guess I just don't understand.

 

[Thunderchild]

well a car engine needs most of its torque to move the car, a minor part of torque drives the dynamo, we will rig a spare dynamo to the lathe to thatwe can have a simulated working environment to see if the circuit works, if the dynamo outputs 22 amps then thats 14 X 22 = 308 watts considering say 80 % dynamor efficiency thats 385 mechanical watts, I'm sure his lathe is at leat 2 KW more than enouth to run the dynamo, the only real world situation we may no be able to replicate is fast acceleration. basically this is a circuit I cannot simulate with software due to the mechanical envolment, i have verified the correct working of the control circuitry and now the next step is connecting it to a running dynamo.

in favour of pics: i'm having a hard job finding op amp chips that can handle high temperature 85 C being the tops where as pics are rated at 125 C

 

[Mikebits]

You should Google military grade op amps. Just a quick search showed this part.

https://www.electro-tech-online.com/custompdfs/2009/09/AD845.pdf

Most likely not what you need, but just to give you an idea.

 

[Thunderchild]

yes that looks more like it, having found all other semiconductor devices that will withstand 125 - 175 C I'd rather keep the minimum spec of 125 C, basically my reasoning is that if I make it bomb proof it should at least be durable, I think someone will be just a tad anoyed with me if i make them a regulator that fails whilst they are a couple of thoudand miles from home for silly reason like poor componenet selection. for the same reason i want to avoid electrolictic condensers