hi kenyan,
What format is the drawing and what size??

Sometimes it appears to take a long time.

__________________
Eric
"Good enough is Perfect"

PIC tutorials:
Gramo's: www.digital-diy.net/
Bill's: www.blueroomelectronics.com/

 

Ok here is the schematic that i submitted. Vcc is 5v

 

hi,
Looked over the circuit I dont seem any major problems in the PT to AM88 ADC inputs.
Its a pity you didnt post this dwg earlier, I could have given a more precise answer.

As you now say the supply to the PT detector is only +5V and I expect the AM/ADC inputs are able to
accept 0 thru +5V, then you will not over drive the ADC inputs.

If you examine the datasheet for the A88 device it will give you a maximum acceptable value of the ADC inputs source impedance.
For example the PIC 16F877 has a max recommended input impedance of 10K0, so if you have a load resistor greater than about 10K it could give you acquistion time problems in the ADC part of the A88.

Ideally I would have chosen to have an opa between the PT detector load and the A88 input, this would have given a greater control over the detectors output range.

You may find that overall the output from the PT has a low voltage swing.

Does this help?

EDIT: downloaded the ATMEGA88 datasheet, its not very helpful at all, unless I have got a simplified version??

__________________
Eric
"Good enough is Perfect"

PIC tutorials:
Gramo's: www.digital-diy.net/
Bill's: www.blueroomelectronics.com/

 

Yea ive gone through all this already and it also mentions an input impedance of 10K is what its designed for. But does what ambient is saying make sense because i observed the behavior he describes when i was testing the sensors in the lab without the atmega88. So does it mean that if i implement the circuit mentioned by ambient with resistance higher than 10k i will have problems with the ADC?

 

hi,
Ref to 'ambient' dwg, if you have a fixed resistor of 10K0 in series with say another 10K variable, then if you turn the pot to say max, total 20K0 then you are working outside the input impedance limit for the AT88.

This will not harm the AT88, but will effect the ADC conversion values to some extent. If you chose a higher value pot it makes it worse.

The voltage across the combined fixed/variable will of course increase.
If you can accept the effect of the acquisition time change ???
its no problem??

EDIT: Out of interest, why are you using the ADC to detect a 'pulse' from the PT detector when the satellite is detected???

__________________
Eric
"Good enough is Perfect"

PIC tutorials:
Gramo's: www.digital-diy.net/
Bill's: www.blueroomelectronics.com/

 

Well satellite's with Infrared emmitters are placed around the room and the satellite is hung from a swivel from the roof. The unit rotates until it detects an infrared signal from the satellites and then stops to fire a laser at the target. The sensors have to be constantly checked to determine if it is still on target because we get more points for a bulls eye. So i assumed the best method was to input the voltages from the sensors into the ADC unit and convert them to a numerical value. If this value is above a certain threshold then we know were are lookin at the target and then can fire the laser.

Is there a bettter way of doing it?

 

also what do you meen by "acquisition time change" and how does this affect the values i would recive if i hooked it up to display the acdc values in a hyperterminal?

 

The ADC conversion value will be reduced if the specified source impedance is exceeded, due to the longer time it takes to charge the internal sample/hold capacitor. It will not be a 'big' reduction, bit does effect the conversion accuracy.

Is there a bettter way of doing it?
I would have chosen to feed the outputs from the 3 PT's into opa/compartors
Then used the comp outputs as digital signals to one of the ports and generate an interrupt. Much faster and simpler that 3 ADC conversions,

__________________
Eric
"Good enough is Perfect"

PIC tutorials:
Gramo's: www.digital-diy.net/
Bill's: www.blueroomelectronics.com/

 

Yea that sounds like it would have been a better way of doing it, but how would you account for different distances ( lower voltages) or is there a way to change the compare value of the opamp/comparator from the atmega88?

 

I know the circuit I am using works just fine with a 10k. I get a 30mV peak-peak signal. I simply decoupled the signal using a 10uF cap from the emitter to an op-amp. After that some simple filtering was done and then more gain, allowing me to get a 0-5V signal using 5V for Vcc.

 

hi ambient,

The problem kenyan has, there is no amplification or level shifting between the PT and the AT88 ADC input.

The IR signal, I suspect will be very small. He could adjust the ADC Vref say down to +2.5Vref, this will double the ADC conversion value by increasing the slope of the ADC.

As he states the ADC input for the AT88 states 10K0 max source impedance.

If kenyan can give us some Vir signal values we may be able to suggest ideas.

__________________
Eric
"Good enough is Perfect"

PIC tutorials:
Gramo's: www.digital-diy.net/
Bill's: www.blueroomelectronics.com/

 

At the distances he is talking about it will be a small signal. I believe I am using a gain around 50 for my application, and that is only designed for a 2 foot range to saturate the signal.