A real PIC question :O)

[Newbie]

Hi,
First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.

now to the real questiion, I have asked this but didn't really get an answer, so here it is again.

**************
If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,

Input High = 2.1v - 3.3v
Input Low = 0v - 0.5v
Output High =2.2v - 3.0v
Output Low = 0v - 0.35v

Also is there any problems running the 16F876 at 3.3v 20mhz"

*********

I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)

Thanks guys.
___________________________
This message was ported from CCS's old forum
Original Post ID: 10681

[Hans Wedemeyer]

I'm reading data sheet DS30292C-page 177

It may be a clue about "the truth" of course Micrchip would never lie to us.... would they...!

On page 177 It's interesting that the tests for HS mode and 3.3V stops at 10MHz.....

I would test it. Take into consideration the temperature range in which the final product is expected/advertised to work in.

Remember everything may be fine in the lab. at room temperature!

Not sure about the BOR I only found figures for 5V operation, but I could have missed it.
Spec's for the GSM seem to be "normal" for modern CMOS device running at 3.3V. I expect if the pic runs it will interface just fine.
You have to watch for "exceeding" maximum and in this case the pic will not exceed 3.3V so it will do just fine.

Now get on with it... that's an order !:-)
___________________________
This message was ported from CCS's old forum
Original Post ID: 10683

[oscar]

:=Hi,
:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=
:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=
:=**************
:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=
:=Input High = 2.1v - 3.3v
:=Input Low = 0v - 0.5v
:=Output High =2.2v - 3.0v
:=Output Low = 0v - 0.35v
:=
:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=
:=*********
:=
:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=
:=Thanks guys.

You suspect wrong Newbie.
First of all, the 16F87X series' power supply only operates down to 4V. Second, you can only operate the crystal at 20MHz at 4.5V supply or higher.

Now, if you change your micro to the 16LF87X series you can operate it down to 2V but at a maximum xtal freq. of 4MHz.
If you do go with the LF series you will have trouble meeting the Vil and Vol specs on the GSM.

If you're intent on maintaining a 20MHz xtal frequency you have to upgrade your supply to a Vdd(min) = 4.5V and add 3.3V to 5V I/O level converters (LHX245 from philips I believe).

If you can't go higher than 3.3V on Vdd then you're max frequency on the xtal is 4MHz.

Good luck,
Oscar
___________________________
This message was ported from CCS's old forum
Original Post ID: 10692

[Newbie]

OK, so if I run the PIC at +5v and as PCM Programmer suggested in an ealier post, I use SN74AHCT1G32 with A+B tied together this should give me a level shift from the PIC to the GSM (the TX line), as for the GSM to the PIC I should not need a level shift as the PIC should see 2.2v-3.0v as a high and 0v as a low.

Would this work.

Thanks again

:=:=Hi,
:=:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=:=
:=:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=:=
:=:=**************
:=:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=:=
:=:=Input High = 2.1v - 3.3v
:=:=Input Low = 0v - 0.5v
:=:=Output High =2.2v - 3.0v
:=:=Output Low = 0v - 0.35v
:=:=
:=:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=:=
:=:=*********
:=:=
:=:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=:=
:=:=Thanks guys.
:=
:=You suspect wrong Newbie.
:=First of all, the 16F87X series' power supply only operates down to 4V. Second, you can only operate the crystal at 20MHz at 4.5V supply or higher.
:=
:=Now, if you change your micro to the 16LF87X series you can operate it down to 2V but at a maximum xtal freq. of 4MHz.
:=If you do go with the LF series you will have trouble meeting the Vil and Vol specs on the GSM.
:=
:=If you're intent on maintaining a 20MHz xtal frequency you have to upgrade your supply to a Vdd(min) = 4.5V and add 3.3V to 5V I/O level converters (LHX245 from philips I believe).
:=
:=If you can't go higher than 3.3V on Vdd then you're max frequency on the xtal is 4MHz.
:=
:=Good luck,
:=Oscar
___________________________
This message was ported from CCS's old forum
Original Post ID: 10700

[oscar]

:=OK, so if I run the PIC at +5v and as PCM Programmer suggested in an ealier post, I use SN74AHCT1G32 with A+B tied together this should give me a level shift from the PIC to the GSM (the TX line), as for the GSM to the PIC I should not need a level shift as the PIC should see 2.2v-3.0v as a high and 0v as a low.
:=
:=Would this work.
:=
:=Thanks again
:=
:=:=:=Hi,
:=:=:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=:=:=
:=:=:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=:=:=
:=:=:=**************
:=:=:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=:=:=
:=:=:=Input High = 2.1v - 3.3v
:=:=:=Input Low = 0v - 0.5v
:=:=:=Output High =2.2v - 3.0v
:=:=:=Output Low = 0v - 0.35v
:=:=:=
:=:=:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=:=:=
:=:=:=*********
:=:=:=
:=:=:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=:=:=
:=:=:=Thanks guys.
:=:=
:=:=You suspect wrong Newbie.
:=:=First of all, the 16F87X series' power supply only operates down to 4V. Second, you can only operate the crystal at 20MHz at 4.5V supply or higher.
:=:=
:=:=Now, if you change your micro to the 16LF87X series you can operate it down to 2V but at a maximum xtal freq. of 4MHz.
:=:=If you do go with the LF series you will have trouble meeting the Vil and Vol specs on the GSM.
:=:=
:=:=If you're intent on maintaining a 20MHz xtal frequency you have to upgrade your supply to a Vdd(min) = 4.5V and add 3.3V to 5V I/O level converters (LHX245 from philips I believe).
:=:=
:=:=If you can't go higher than 3.3V on Vdd then you're max frequency on the xtal is 4MHz.
:=:=
:=:=Good luck,
:=:=Oscar

Newbie,

One question: Is the Input High (Vih) to the GSM strictly 2.1 - 3.3V? Or can it be higher? If you're powering the GSM from a 3.8V supply I would assume that the Vih has a minimum of 2.1V but a max of 3.8V. Can you verify this please.

If Vih can be close to 3.8V then I recommend you use Philips' NC7S08 (2-input AND tiny gate). Tie pins 1 & 5 to 3.8V (GSM supply), pin 2 to the TX out on the PIC, pin 3 to ground and 4 to the GSM TX input.

You can't use the SN74AHCT1G32 from TI because it has a minimum supply requirement of 4.5V. If you tie this to the +5V supply of the pic then you still will be outputing >3.8V to the GSM TX input and I imagine this would destroy the TX input.

If the Vih of the GSM can be no higher than 3.3V then this is what you will have to do (your last resort). Obtain a 3.3V LDO regulator, connect its input to the +5V supply and feed the 3.3V output to pins 1 & 5 on the NC7S08 as described two paragraphs above the rest of the pins stay the same. What this will do is drive your GSM TX input with a minimum of 2.2V or so and a maximum of 3.3V. This should solve your problem for sure (if Vih on GSM is stricly 2.1V - 3.3V) but you will have to spend another $0.50 on the 3.3V LDO regulator.

Let me know how it goes,
Oscar
___________________________
This message was ported from CCS's old forum
Original Post ID: 10703

[Newbie]

OK I understand, luckily I have a 3.3v reg already on board so I can hook the NC7S08 to this, Could I not simmply use a 3.3v Reg in the TX line to bring the PIC tx voltage down or will it not respond fast enough?

I'm surprised there is not a small 5v level shifter that will take it from 5v to 3.3v?

And one last thing is the TX from GSM to RX on PIC OK without anything inline?


So

:=:=OK, so if I run the PIC at +5v and as PCM Programmer suggested in an ealier post, I use SN74AHCT1G32 with A+B tied together this should give me a level shift from the PIC to the GSM (the TX line), as for the GSM to the PIC I should not need a level shift as the PIC should see 2.2v-3.0v as a high and 0v as a low.
:=:=
:=:=Would this work.
:=:=
:=:=Thanks again
:=:=
:=:=:=:=Hi,
:=:=:=:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=:=:=:=
:=:=:=:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=:=:=:=
:=:=:=:=**************
:=:=:=:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=:=:=:=
:=:=:=:=Input High = 2.1v - 3.3v
:=:=:=:=Input Low = 0v - 0.5v
:=:=:=:=Output High =2.2v - 3.0v
:=:=:=:=Output Low = 0v - 0.35v
:=:=:=:=
:=:=:=:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=:=:=:=
:=:=:=:=*********
:=:=:=:=
:=:=:=:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=:=:=:=
:=:=:=:=Thanks guys.
:=:=:=
:=:=:=You suspect wrong Newbie.
:=:=:=First of all, the 16F87X series' power supply only operates down to 4V. Second, you can only operate the crystal at 20MHz at 4.5V supply or higher.
:=:=:=
:=:=:=Now, if you change your micro to the 16LF87X series you can operate it down to 2V but at a maximum xtal freq. of 4MHz.
:=:=:=If you do go with the LF series you will have trouble meeting the Vil and Vol specs on the GSM.
:=:=:=
:=:=:=If you're intent on maintaining a 20MHz xtal frequency you have to upgrade your supply to a Vdd(min) = 4.5V and add 3.3V to 5V I/O level converters (LHX245 from philips I believe).
:=:=:=
:=:=:=If you can't go higher than 3.3V on Vdd then you're max frequency on the xtal is 4MHz.
:=:=:=
:=:=:=Good luck,
:=:=:=Oscar
:=
:=Newbie,
:=
:=One question: Is the Input High (Vih) to the GSM strictly 2.1 - 3.3V? Or can it be higher? If you're powering the GSM from a 3.8V supply I would assume that the Vih has a minimum of 2.1V but a max of 3.8V. Can you verify this please.
:=
:=If Vih can be close to 3.8V then I recommend you use Philips' NC7S08 (2-input AND tiny gate). Tie pins 1 & 5 to 3.8V (GSM supply), pin 2 to the TX out on the PIC, pin 3 to ground and 4 to the GSM TX input.
:=
:=You can't use the SN74AHCT1G32 from TI because it has a minimum supply requirement of 4.5V. If you tie this to the +5V supply of the pic then you still will be outputing >3.8V to the GSM TX input and I imagine this would destroy the TX input.
:=
:=If the Vih of the GSM can be no higher than 3.3V then this is what you will have to do (your last resort). Obtain a 3.3V LDO regulator, connect its input to the +5V supply and feed the 3.3V output to pins 1 & 5 on the NC7S08 as described two paragraphs above the rest of the pins stay the same. What this will do is drive your GSM TX input with a minimum of 2.2V or so and a maximum of 3.3V. This should solve your problem for sure (if Vih on GSM is stricly 2.1V - 3.3V) but you will have to spend another $0.50 on the 3.3V LDO regulator.
:=
:=Let me know how it goes,
:=Oscar
:=
___________________________
This message was ported from CCS's old forum
Original Post ID: 10714

[oscar]

:=OK I understand, luckily I have a 3.3v reg already on board so I can hook the NC7S08 to this, Could I not simmply use a 3.3v Reg in the TX line to bring the PIC tx voltage down or will it not respond fast enough?
:=
:=I'm surprised there is not a small 5v level shifter that will take it from 5v to 3.3v?
:=
:=And one last thing is the TX from GSM to RX on PIC OK without anything inline?
:=
:=
:=So
:=
:=:=:=OK, so if I run the PIC at +5v and as PCM Programmer suggested in an ealier post, I use SN74AHCT1G32 with A+B tied together this should give me a level shift from the PIC to the GSM (the TX line), as for the GSM to the PIC I should not need a level shift as the PIC should see 2.2v-3.0v as a high and 0v as a low.
:=:=:=
:=:=:=Would this work.
:=:=:=
:=:=:=Thanks again
:=:=:=
:=:=:=:=:=Hi,
:=:=:=:=:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=:=:=:=:=
:=:=:=:=:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=:=:=:=:=
:=:=:=:=:=**************
:=:=:=:=:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=:=:=:=:=
:=:=:=:=:=Input High = 2.1v - 3.3v
:=:=:=:=:=Input Low = 0v - 0.5v
:=:=:=:=:=Output High =2.2v - 3.0v
:=:=:=:=:=Output Low = 0v - 0.35v
:=:=:=:=:=
:=:=:=:=:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=:=:=:=:=
:=:=:=:=:=*********
:=:=:=:=:=
:=:=:=:=:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=:=:=:=:=
:=:=:=:=:=Thanks guys.
:=:=:=:=
:=:=:=:=You suspect wrong Newbie.
:=:=:=:=First of all, the 16F87X series' power supply only operates down to 4V. Second, you can only operate the crystal at 20MHz at 4.5V supply or higher.
:=:=:=:=
:=:=:=:=Now, if you change your micro to the 16LF87X series you can operate it down to 2V but at a maximum xtal freq. of 4MHz.
:=:=:=:=If you do go with the LF series you will have trouble meeting the Vil and Vol specs on the GSM.
:=:=:=:=
:=:=:=:=If you're intent on maintaining a 20MHz xtal frequency you have to upgrade your supply to a Vdd(min) = 4.5V and add 3.3V to 5V I/O level converters (LHX245 from philips I believe).
:=:=:=:=
:=:=:=:=If you can't go higher than 3.3V on Vdd then you're max frequency on the xtal is 4MHz.
:=:=:=:=
:=:=:=:=Good luck,
:=:=:=:=Oscar
:=:=
:=:=Newbie,
:=:=
:=:=One question: Is the Input High (Vih) to the GSM strictly 2.1 - 3.3V? Or can it be higher? If you're powering the GSM from a 3.8V supply I would assume that the Vih has a minimum of 2.1V but a max of 3.8V. Can you verify this please.
:=:=
:=:=If Vih can be close to 3.8V then I recommend you use Philips' NC7S08 (2-input AND tiny gate). Tie pins 1 & 5 to 3.8V (GSM supply), pin 2 to the TX out on the PIC, pin 3 to ground and 4 to the GSM TX input.
:=:=
:=:=You can't use the SN74AHCT1G32 from TI because it has a minimum supply requirement of 4.5V. If you tie this to the +5V supply of the pic then you still will be outputing >3.8V to the GSM TX input and I imagine this would destroy the TX input.
:=:=
:=:=If the Vih of the GSM can be no higher than 3.3V then this is what you will have to do (your last resort). Obtain a 3.3V LDO regulator, connect its input to the +5V supply and feed the 3.3V output to pins 1 & 5 on the NC7S08 as described two paragraphs above the rest of the pins stay the same. What this will do is drive your GSM TX input with a minimum of 2.2V or so and a maximum of 3.3V. This should solve your problem for sure (if Vih on GSM is stricly 2.1V - 3.3V) but you will have to spend another $0.50 on the 3.3V LDO regulator.
:=:=
:=:=Let me know how it goes,
:=:=Oscar
:=:=

Newbie:
You didn't answer my question about the Vcc on the GSM, is it 3.8V? and does Vih have to be 2.1 to 3.3V strictly?

In the world of level-shifting between a 5V and 3.3V system there's always an asumption that there will be two different power supplies (one a 5V and the other a 3.3V). Your case is a bit different because you have a 3.8V supply with weird Vih requirments. This is the only solution.

You can't use a 3.3V regulator as a level shifter because regulators have a rather long power-up curve. I assume the data transmitted from the PIC to the GSM is greater than 1kHz?
Most regulators require power-up times in the milliseconds.

You asked why there isn't a small 5V to 3.3V level shifter, well the NC7S08 is a tiny SOT-23 package, you just can't get smaller....wait... you can there is a SC-70 package but you have to have rock-steady hands to solder that thing on the board. However, you will still need a 3.3V regulator because of the non-standard Vih your GSM requires.

Yes, you are ok on the GSM TX to PIC RX pins.

Is the power supply on the GSM strictly 3.8V? Can you drop that down to 3.3V?
How are you obtaining 3.8V? Are you using a variable regulator?
Hope this helps,
Oscar
___________________________
This message was ported from CCS's old forum
Original Post ID: 10723

[PCM programmer]

:=
:=I'm surprised there is not a small 5v level shifter that will take it from 5v to 3.3v?
:=
----------------------------------------------------
Did you look at the TI document on level shifting ?
<a href="http://www.ti.com/sc/docs/psheets/abstract/apps/scya006.htm" TARGET="_blank">http://www.ti.com/sc/docs/psheets/abstract/apps/scya006.htm</a>
On page 4 (page 8 of the PDF file), it has a section called:
"Voltage Translation from 5v TTL to 3.3v LVTTL/LVCMOS."
It says you can use TI 3.3v logic families that have 5v
tolerant inputs. These are LVC, AHC, LVT, and ALVT.
The PIC outputs are 5v CMOS, but because the above logic
families are 5v tolerant, they will work.

If you run the following gate at 3.3v, it will accept signals
from a PIC running at 5v, and put a Voh of 3.3v.
TI p/n: SN74LVC1G32DBVR
Digikey p/n: 296-9847-1-ND

The TI document referenced above should have solutions for
all your level-shifting problems.
___________________________
This message was ported from CCS's old forum
Original Post ID: 10726

[Newbie]

:=:=OK I understand, luckily I have a 3.3v reg already on board so I can hook the NC7S08 to this, Could I not simmply use a 3.3v Reg in the TX line to bring the PIC tx voltage down or will it not respond fast enough?
:=:=
:=:=I'm surprised there is not a small 5v level shifter that will take it from 5v to 3.3v?
:=:=
:=:=And one last thing is the TX from GSM to RX on PIC OK without anything inline?
:=:=
:=:=
:=:=So
:=:=
:=:=:=:=OK, so if I run the PIC at +5v and as PCM Programmer suggested in an ealier post, I use SN74AHCT1G32 with A+B tied together this should give me a level shift from the PIC to the GSM (the TX line), as for the GSM to the PIC I should not need a level shift as the PIC should see 2.2v-3.0v as a high and 0v as a low.
:=:=:=:=
:=:=:=:=Would this work.
:=:=:=:=
:=:=:=:=Thanks again
:=:=:=:=
:=:=:=:=:=:=Hi,
:=:=:=:=:=:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=:=:=:=:=:=
:=:=:=:=:=:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=:=:=:=:=:=
:=:=:=:=:=:=**************
:=:=:=:=:=:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=:=:=:=:=:=
:=:=:=:=:=:=Input High = 2.1v - 3.3v
:=:=:=:=:=:=Input Low = 0v - 0.5v
:=:=:=:=:=:=Output High =2.2v - 3.0v
:=:=:=:=:=:=Output Low = 0v - 0.35v
:=:=:=:=:=:=
:=:=:=:=:=:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=:=:=:=:=:=
:=:=:=:=:=:=*********
:=:=:=:=:=:=
:=:=:=:=:=:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=:=:=:=:=:=
:=:=:=:=:=:=Thanks guys.
:=:=:=:=:=
:=:=:=:=:=You suspect wrong Newbie.
:=:=:=:=:=First of all, the 16F87X series' power supply only operates down to 4V. Second, you can only operate the crystal at 20MHz at 4.5V supply or higher.
:=:=:=:=:=
:=:=:=:=:=Now, if you change your micro to the 16LF87X series you can operate it down to 2V but at a maximum xtal freq. of 4MHz.
:=:=:=:=:=If you do go with the LF series you will have trouble meeting the Vil and Vol specs on the GSM.
:=:=:=:=:=
:=:=:=:=:=If you're intent on maintaining a 20MHz xtal frequency you have to upgrade your supply to a Vdd(min) = 4.5V and add 3.3V to 5V I/O level converters (LHX245 from philips I believe).
:=:=:=:=:=
:=:=:=:=:=If you can't go higher than 3.3V on Vdd then you're max frequency on the xtal is 4MHz.
:=:=:=:=:=
:=:=:=:=:=Good luck,
:=:=:=:=:=Oscar
:=:=:=
:=:=:=Newbie,
:=:=:=
:=:=:=One question: Is the Input High (Vih) to the GSM strictly 2.1 - 3.3V? Or can it be higher? If you're powering the GSM from a 3.8V supply I would assume that the Vih has a minimum of 2.1V but a max of 3.8V. Can you verify this please.
:=:=:=
:=:=:=If Vih can be close to 3.8V then I recommend you use Philips' NC7S08 (2-input AND tiny gate). Tie pins 1 & 5 to 3.8V (GSM supply), pin 2 to the TX out on the PIC, pin 3 to ground and 4 to the GSM TX input.
:=:=:=
:=:=:=You can't use the SN74AHCT1G32 from TI because it has a minimum supply requirement of 4.5V. If you tie this to the +5V supply of the pic then you still will be outputing >3.8V to the GSM TX input and I imagine this would destroy the TX input.
:=:=:=
:=:=:=If the Vih of the GSM can be no higher than 3.3V then this is what you will have to do (your last resort). Obtain a 3.3V LDO regulator, connect its input to the +5V supply and feed the 3.3V output to pins 1 & 5 on the NC7S08 as described two paragraphs above the rest of the pins stay the same. What this will do is drive your GSM TX input with a minimum of 2.2V or so and a maximum of 3.3V. This should solve your problem for sure (if Vih on GSM is stricly 2.1V - 3.3V) but you will have to spend another $0.50 on the 3.3V LDO regulator.
:=:=:=
:=:=:=Let me know how it goes,
:=:=:=Oscar
:=:=:=
:=
:=Newbie:
:=You didn't answer my question about the Vcc on the GSM, is it 3.8V? and does Vih have to be 2.1 to 3.3V strictly?
:=
:=In the world of level-shifting between a 5V and 3.3V system there's always an asumption that there will be two different power supplies (one a 5V and the other a 3.3V). Your case is a bit different because you have a 3.8V supply with weird Vih requirments. This is the only solution.
:=
:=You can't use a 3.3V regulator as a level shifter because regulators have a rather long power-up curve. I assume the data transmitted from the PIC to the GSM is greater than 1kHz?
:=Most regulators require power-up times in the milliseconds.
:=
:=You asked why there isn't a small 5V to 3.3V level shifter, well the NC7S08 is a tiny SOT-23 package, you just can't get smaller....wait... you can there is a SC-70 package but you have to have rock-steady hands to solder that thing on the board. However, you will still need a 3.3V regulator because of the non-standard Vih your GSM requires.
:=
:=Yes, you are ok on the GSM TX to PIC RX pins.
:=
:=Is the power supply on the GSM strictly 3.8V? Can you drop that down to 3.3V?
:=How are you obtaining 3.8V? Are you using a variable regulator?
:=Hope this helps,
:=Oscar

Sorrt for not answering your question, The spec for the GSM is Its 3.4v-4.2v I'm going to use switch mode power supply as it also can peak at 1.9amp! and yes Vih is 2.1v-3.3v.

Thanks for your help...
___________________________
This message was ported from CCS's old forum
Original Post ID: 10730

[John Yaron]

:=Hi,
:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=
:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=
:=**************
:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=
:=Input High = 2.1v - 3.3v
:=Input Low = 0v - 0.5v
:=Output High =2.2v - 3.0v
:=Output Low = 0v - 0.35v
:=
:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=
:=*********
:=
:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=
:=Thanks guys.

www.fairchildsemi.com/ds/FS/FSTD3306.pdf
___________________________
This message was ported from CCS's old forum
Original Post ID: 10902

[John Yaron]

:=Hi,
:=First of all thanks to all who helped with the "Don't shoot me question" I have opted for a switch mode supply, a bit more technical but I think I can handle it.
:=
:=now to the real questiion, I have asked this but didn't really get an answer, so here it is again.
:=
:=**************
:=If I run the 16F876 at 3.3v will this interface directly with a GSM that runs at 3.8v and has the follwing I/O spec,
:=
:=Input High = 2.1v - 3.3v
:=Input Low = 0v - 0.5v
:=Output High =2.2v - 3.0v
:=Output Low = 0v - 0.35v
:=
:=Also is there any problems running the 16F876 at 3.3v 20mhz"
:=
:=*********
:=
:=I suspect the answer is "yes, everything will be fine" as the max voltage from the PIC cannot exceed 3.3v but will be high enough to be a logic 1, and vise versa. and the PIC will run fine at 3.3v (with disabled brown out detection)
:=
:=Thanks guys.


www.fairchildsemi.com/ds/FS/FSTD3306.pdf
___________________________
This message was ported from CCS's old forum
Original Post ID: 10903