AVR-Chat

Hi All,



Could you help me on this one?



The AT90S2313 seems to be very picky about it's power supply.

Flicking an electrical load on or off randomly put the ACU into reset.

I've tried to separate all ins & outs, double buffered the PSU, used
some of those ferrite chokes.



But... still no result...



Recently I bought a Honda CBF600 tophalf.

All well accept for the headlight wiring.

Honda's policy is to turn on the lights although the engine
isn't running.

So the onboard battery, just a 6Ah, has to coop with 75W lights &
another 35W for the ignition, without being charged. The battery has to
be in very good shape getting the engine started after a pre-load
"shock" of 9Amp's.



  [B-)] So I took an AT90S2313, wired it's inputs by opt couplers to
the dynamo and high beam switch and used some N-MosFET's  to
power-up the 2 60W headlight bulbs.



  [:((] And so a reset does mean: NO LIGHT at all!  [:((]



The program has to follow a few rules:

1 – As soon as the engine is running, wait a few sec. light-up the
high-beam at 30%, wait a few sec. light-up the low-beam.

2 – Turnoff the lights as soon the starter is activated.

3 – If the engine is turned off wait a few sec., turn-off the low
beam wait half a min. than turn off the reduced high-beam.

4 – Flicking the phasser or switch the high-beam, immediately juice
up the high-beam.



[Non-text portions of this message have been removed]

----- Original Message ----- 

From: "max.motor" <imeelmax@yahoo.com>
To: <AVR-Chat@yahoogroups.com>
Sent: Monday, September 10, 2007 4:14 AM
Subject: [AVR-Chat] How to overcome brownout spikes (AT90S2313)



Hi All,



Could you help me on this one?



The AT90S2313 seems to be very picky about it's power supply.

Flicking an electrical load on or off randomly put the ACU into reset.

I've tried to separate all ins & outs, double buffered the PSU, used
some of those ferrite chokes.

-------------------------

Try a transient suppressor like an AVX TransGuard.

Leon
--
Leon Heller
Amateur radio call-sign G1HSM
Yaesu FT-817ND and FT-857D transceivers
Suzuki SV1000S motorcycle
leon355@btinternet.com
http://www.geocities.com/leon_heller

Hi, Nice project. A whole bunch of your trouble is probably coming
from using the 90S2313, I remember using those quite a few years ago
now and they were very susceptible to power supply problems. I would
definitely recommend using a newer chip like the Tiny26 or Tiny2313.
But if you have to persist with the 90s2313, try the following things:

FILTER THE POWER SUPPLY - Series inductor or resistor into a 100n cap
on all the vcc pins. 

OPTO COUPLE SIGNALS - use optocouplers on all input AND output signals.

RESET CIRCUIT - Use a 4k7 pullup and 100N capacitor on the reset line
and bypass the 4k7 with a reverse biased diode. 

Use a fast regulator to generate VCC, you may find in an automotive
application a decent switchmode regulator will work better than a bog
standard old 7805

Also use shielded cable for any runs longer than 30cm (10 inches)or so.

Hope this helps !!!
Cheers
Hein B
Auckland, New Zealand


--- In AVR-Chat@yahoogroups.com, "max.motor" <imeelmax@...> wrote:

>
> 
> Hi All,
> 
> 
> 
> Could you help me on this one?
> 
> 
> 
> The AT90S2313 seems to be very picky about it's power supply.
> 
> Flicking an electrical load on or off randomly put the ACU into reset.
> 
> I've tried to separate all ins & outs, double buffered the PSU, used
> some of those ferrite chokes.
> 
> 
> 
> But... still no result...
> 
> 
> 
> Recently I bought a Honda CBF600 tophalf.
> 
> All well accept for the headlight wiring.
> 
> Honda's policy is to turn on the lights although the engine
> isn't running.
> 
> So the onboard battery, just a 6Ah, has to coop with 75W lights &
> another 35W for the ignition, without being charged. The battery has to
> be in very good shape getting the engine started after a pre-load
> "shock" of 9Amp's.
> 
> 
> 
>   [B-)] So I took an AT90S2313, wired it's inputs by opt couplers to
> the dynamo and high beam switch and used some N-MosFET's  to
> power-up the 2 60W headlight bulbs.
> 
> 
> 
>   [:((] And so a reset does mean: NO LIGHT at all!  [:((]
> 
> 
> 
> The program has to follow a few rules:
> 
> 1 – As soon as the engine is running, wait a few sec. light-up the
> high-beam at 30%, wait a few sec. light-up the low-beam.
> 
> 2 – Turnoff the lights as soon the starter is activated.
> 
> 3 – If the engine is turned off wait a few sec., turn-off the low
> beam wait half a min. than turn off the reduced high-beam.
> 
> 4 – Flicking the phasser or switch the high-beam, immediately juice
> up the high-beam.
> 
> 
> 
> [Non-text portions of this message have been removed]
>

Thanks Hein & Leon,

Yeh, I've used a simple 78L05. It might be to slow. Have you got a
partnr. of a quick version? Or would a suppressor be enough.

  I think swapping the 90s for a Tiny would be the easyest thing to do.
Aren't both ones (Tiny26 & -2313) pin compatible with the 90s version?

About the reset circuit, would it be like this?

+                      -
|           o--Reset   |
|----|<|----|   100n   |
|---[4k7]---|----||----|
|                      |

The input signals are already opto coupled. The output signals aren't
and that would be my next step. I wasn't for sure if it would give a
sharp signal. as soon a power FET has to dissipate it'll get very very
hot. I've got a To220 with a melted backside. (Yep, It's still alive)





[Non-text portions of this message have been removed]

Hi Again, Yeah, that reset circuit is correct, it LP filters the reset
line, but still allows it to reset quickly when VCC drops out. 

Ive had endless trouble driving electric motors through fets straight
from the output of my micro controllers, noise and spikes can get to
the uc through an output! Maybe you can use a Schmidt trigger on the
back end of an optocoupler to keep the edges going to your fet nice
and fast ? Youd have to use a high speed optocoupler if you want to
PWM the fet tho. 

I believe the Tiny2313 is pin-compatible, not sure about the Tiny26. 

Yeah, its amazing how much punishment some power fets will take! And
the amount of current some of the IRF fets can pass is truly amazing,
but it's critical to keep the turn-on and turn-off times as short as
possible, I usually drive mine with a totem-pole configuration to
overcome the gate-source capacitance and have it turn on / off very fast!

Keep us informed on your progress and findings! 

Cheers
Hein B
Auckland, New Zealand. 


--- In AVR-Chat@yahoogroups.com, "max.motor" <imeelmax@...> wrote:

>
> 
> Thanks Hein & Leon,
> 
> Yeh, I've used a simple 78L05. It might be to slow. Have you got a
> partnr. of a quick version? Or would a suppressor be enough.
> 
>   I think swapping the 90s for a Tiny would be the easyest thing to do.
> Aren't both ones (Tiny26 & -2313) pin compatible with the 90s version?
> 
> About the reset circuit, would it be like this?
> 
> +                      -
> |           o--Reset   |
> |----|<|----|   100n   |
> |---[4k7]---|----||----|
> |                      |
> 
> The input signals are already opto coupled. The output signals aren't
> and that would be my next step. I wasn't for sure if it would give a
> sharp signal. as soon a power FET has to dissipate it'll get very very
> hot. I've got a To220 with a melted backside. (Yep, It's still alive)
> 
> 
> 
> 
> 
> [Non-text portions of this message have been removed]
>

At 12:41 AM 13/09/2007, you wrote:
>Yeh, I've used a simple 78L05. It might be to slow. Have you got a
>partnr. of a quick version?

Have a look at National automotive regulators LM2940.


Regards

John Samperi

********************************************************
Ampertronics Pty. Ltd.
11 Brokenwood Place Baulkham Hills, NSW 2153 AUSTRALIA
Tel. (02) 9674-6495       Fax (02) 9674-8745
Email: john@ampertronics.com.au
Website  http://www.ampertronics.com.au
*Electronic Design * Custom Products * Contract Assembly
********************************************************