hehehe... Ask and ye shall recieve.
This is a design based on your suggestion.·
I had to dropped from 42 to 30 SSR.·
The difference is the 30 additional H11L1m's
The basic idea / schmatic·is attached as a pdf...
It's times like this I wish I had more education in electronics!
All I know is the SSR requires 5V's to trigger, and the propeller output's 3.3.
And, the SSR doesn't care + or - on how it's control lines are connected, so
To multiplex them in an effecent mannor, Sink & Source the catadodes and anadodes, ie charlieplexing.
The SSR's are going to need to have the ability to turn on and off rapidly, so speed is of some consern.
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller + Hardware - extra bits for the bit bucket =· 1 Coffeeless KaosKidd
Post Edited (Kaos Kidd) : 7/21/2006 7:26:37 PM GMT
The max switching frequency should be about 30~40 hz or better.
I just read through the data sheet, and it's not the one I'v printed! It's the same numbers, but it's a completly different item!
This one's got diodes in the inbound, 3-6V input (solves all my issues here!)
I'm telling you guys, this is not the data sheet I have printed for the same number from digi key.
Im not crazy... old yes... unknowing at times yes, but I... I sware it wasn't there about 2 weeks ago when I printed it...
Grrrrr....
[noparse]:)[/noparse]
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller + Hardware - extra bits for the bit bucket =· 1 Coffeeless KaosKidd
·
Those are pretty slow devices. They don't have to be "microsecond fast", though, since they're switching 60Hz AC. Even with a 5V input requirement, the optoisolators would have been overkill, and a simple NPN open collector driver would've sufficed. But with the input spec ranging down to 3V, you probably don't even need that.
'Looks like those 42 SSRs are gonna fit after all!
Phil:
What I'm trying to do is switch strings of xmass lights. Originally it was for 64 strings for a display. Due to the costs of the SSR's I'v cut the number of circuits down to 32. SSR's seemed like the best way to go, espically considering my lack of electronics information. After reading about charlieplexing to get the pin count down, and the propellers fast speed, I was sure I had the right plan. The original SSR's (the one that sparked this conversation in the first place) didn't have + & - on the controll lines and speced at 5V controll. They were priced at 3.20 each. The ones I found on friday have a internal diode, marked 3~6 V, and PIV of 6V, well exceeding any spec's I needed, but the price for these are 4.80 each. And now you introduce the speed issue. I'm whirling in all this information, almost ready to tell my Grandfather that I just cant do the project yet. Do you think this will work?
KK
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller + Hardware - extra bits for the bit bucket =· 1 Coffeeless KaosKidd
·
Bernd, I need to interface a bidirectional signal between a Stamp and the Prop.· Do you have a schematic showing how to connect the 4050?· I think all I need to do is drop the Stamp to Prop signal from 5V to 3.3V.· The Stamp should operate OK with a 3.3V input signal.
Thanks
Sid
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ Sid Weaver
Don't have VGA?
all you nead is in the data sheed !
example[noparse]:([/noparse]5V to 3.3V)
In to Out(Pin3-> Pin2, Pin5->Pin4...) and VCC· 3,3V at Pin 1·, GND at Pin 8
(It's only a high to low level shifter.)
Regards
Bernd
Post Edited (MC-Bernd) : 11/13/2006 2:02:16 PM GMT
Anyone know of any bidirectional level shifters like the max3378 mentioned in this forum in a DIP package? I'm doing some work for a senior design project. I'd like to prototype on a breadboard.
Nearly all new chips made by manufacturers are not availible in DIP packaging. That said you can get adapter boards, one that will work with the max3378 tssop package is availible here: http://www.sparkfun.com/commerce/product_info.php?products_id=498 also pick up a strip of machine pin headers to use with it because they plug into a breadboard much easier than square post headers. Be sure to have some flux on hand and follow the technique shown in this video when soldering: http://www.heinc.com/xytronic/Images/Drag%20Soldering%20107.wmv
I have a 9.2V supply powering a Propstick. To avoid loading the on board regulator as much as possible, I am supplying dry contact switches with the 9.2V through a 5.1V zener to drop the voltage on the supply to 4.1V. I have 4.6k resistors between the switches and Prop pins A0-A7. When the switch is closed, I see 9.2V on the supply, 5.1V across the zener, and 3.98 to 3.75V on the Prop IO pins relative to Prop GND. I expected a lower voltage.
Before (Monday) I was switching the regulated 3.3V onto the pins.
What am I missing, please, and am I cooking the pins?
Hmm, I'm not sure I understand; what do you think you're missing? That is, what voltage do you expect to see on those pins?
The voltage you describe is probably what I would expect. Measure the voltage across the resistors and this will give you an
idea of the current you are feeding into those pins. You may be pushing current back into your 3.3 regulator.
How do I set up the Boe-Bot Digital Encoder circuit on the Propeller Protoboard? Do I connect Vdd to +5 volts or +3.3? And, if the answer is +5, does that change the part of the circuit where the black wire connects to the resistor and Vdd, the digital encoder, and the Basic Stamp pins P11/P10?
Why did you expect it only to be a few tenths of a volt? Where did you expect the rest of the voltage drop to be,
across the resistor? With V=IR, you should only get this if you had current flow, and you shouldn't have steady-state
current flow on an input, unless the voltage is higher.
You still may be feeding current back into your supply though, depending on what the prop is doing.
Just for complete this repository of 5V solutions.
Beau Schwabe said...
"would a simple resistor be suitable for bi-directional communication between a prop and Parallax's GPS module" - for receiving the RAW signals yes, a 4.7K resistor would be fine.
For sending it depends on what the input threshold voltage is for the 5V device. (<--This applies to any 5V device connected to a lower voltage interface.) ... I could not find this
information on the GPS datasheet, and I no longer have my setup connected. But if I recall, the input voltage threshold is around 4V. You could try this trick using a diode to the
GND connection of the 3.3V regulator supplying the Propeller. Effectively this level shifts the Propeller I/O so that it outputs .6V for a logic LOW and 3.9V for a logic HIGH.
I haven't tried this with the Propeller, but I have used this method with other processors running at 2V. It might be just enough to get you within the input trigger threshold.
First: I just discovered this marvel in my distributor's catalogue. I didn't think much of it - mostly because I thought there was a mis-print somewhere. On second thought, I realised that it COULD actually be true, so I visited the Parallax site and fell in love immediately. Last time that happened was when I learned about Forth some twentyfive years ago. I have downloaded the pdfs and I am still in love - will get a demo board tomorrow.
Now, the discussion.
There seems to be some consensus on 1 k or 10 k. Even 100 k has been mentioned and my new hero Chip says that zero ohms would be OK as well.
I am a bit concerned about the 1 k. This is a fast chip. And we can get as much as 100 pF capacitance in a signal wire if it has some length and if a few inputs are connected to it. That would give us a time constant of 100 ns - that's a lot with a processor capable of producing pulses as short as 12,5 ns. A lot of unnecessary problems are likely to pop up if keeping to the 1 k series resistor. The more obvious is that many clock pulses will never reach threshold levels and cause frustrations for users that are not used to this kind of considerations.
I would go for 100 ohms, or 47 ohms. That would keep time constants· down in the 10 - 5 ns range and probably also be a better match to the line impedance (less reflexions). Remember that the 5 V parts you are going to interface with sometimes have nasty properties (like creating extra "quirks" in a slow edge and other mysterious things) and that a fast edge always is preferred over a slow one (OK, emission and ground bounce, I know, but...).
What is the general opinion here? Wouldn't it be good to have a standard way to handle this situation? And a standard that is good both to the chip and to the functionality? Like a 100 ohms series resistor.
Need to say it once more: I am in love again·- with a chip, not a chick!
Time constants don't play as much as a role as it would seem to in this situation. Resistors are only needed when the Propeller has an input which is 5V and you place the resistor as close as practical to the Propeller. So there is a pi filter:
Propeller Pin --+--/\/\---+-- 5V signal
| |
| |
6 pF === === X
| |
--- ---
- -
The capacitor to ground on the 5V side is the line capacitance, but this capacitance is charged at the slew rate of the 5V device, it is not current limited by the resistor. The RC filter which does need to be accounted for is the 6 pF which is the input capacitance of the Propeller, but it's order of magnitude is small enough that it's not much of a problem.
There can't really be a universal value, because the goal of an individual design may need to be optimized in a particular direction. For instance a product which must pass FCC EMI regulations will want to limit the slew rate to minimize RF radiation.
I want to hook up one of my propellers to·some CMOS dynamic ram chips. Reading this thread, it sounds like all I need is 1k resistors in series for the lines I will read from, and that even though the high treshold on the address inputs etc on the dram is 2.4V I should be fine.
My question is regarding the data lines... will the 1k resistor in series be enough? Or should I also pull the·dram side high to at least 3.3V?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔ www.mikronauts.com - a new blog about microcontrollers
Mike Green said...
Any suggestions for an appropriate device for the level shifter?
I am looking into using the VN10LP for level shifting the i2c bus. It has a gate source threshold voltage of 0.8V min. to 2.5V max. The on resistance is low at 5 ohms, and The turn on/off time is 10ns. There is a large supply available at Digikey. I have attached the datasheet,·for anyone who is·interested in seeing if they agree or disagree·that this would be a good choice.
You really have to read MOSFET datasheets carefully. The gate-source threshold is the voltage at which the FET begins to conduct, and you might be pushing your luck if you get one with VGS(th) = 2.5V. You stated that the on resistance is 5 ohms, but look at the conditions under which that's measured: VGS = 10V. What you want is a "logic MOSFET", but it's not easy to find logic MOSFETs compatible with 3.3V logic that come in a through-hole package. (A search on DigiKey turned up none.) Could you use a surface-mount device?
OTOH, if it's a one-off project, you could just try a few and see if they work. After all, it's not like you're sinking a lot of current!
Comments
This is a design based on your suggestion.·
I had to dropped from 42 to 30 SSR.·
The difference is the 30 additional H11L1m's
The basic idea / schmatic·is attached as a pdf...
It's times like this I wish I had more education in electronics!
All I know is the SSR requires 5V's to trigger, and the propeller output's 3.3.
And, the SSR doesn't care + or - on how it's control lines are connected, so
To multiplex them in an effecent mannor, Sink & Source the catadodes and anadodes, ie charlieplexing.
The SSR's are going to need to have the ability to turn on and off rapidly, so speed is of some consern.
·
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller + Hardware - extra bits for the bit bucket =· 1 Coffeeless KaosKidd
Post Edited (Kaos Kidd) : 7/21/2006 7:26:37 PM GMT
I assume the SSRs are optoisolated? How rapidly do they need to switch? How rapidly do they switch, given a fast input?
-Phil
heres the data sheet for the SSR's
http://rocky.digikey.com/WebLib/Sharp/Web Data/s101n11.pdf
The max switching frequency should be about 30~40 hz or better.
I just read through the data sheet, and it's not the one I'v printed! It's the same numbers, but it's a completly different item!
This one's got diodes in the inbound, 3-6V input (solves all my issues here!)
I'm telling you guys, this is not the data sheet I have printed for the same number from digi key.
Im not crazy... old yes... unknowing at times yes, but I... I sware it wasn't there about 2 weeks ago when I printed it...
Grrrrr....
[noparse]:)[/noparse]
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller + Hardware - extra bits for the bit bucket =· 1 Coffeeless KaosKidd
·
Those are pretty slow devices. They don't have to be "microsecond fast", though, since they're switching 60Hz AC. Even with a 5V input requirement, the optoisolators would have been overkill, and a simple NPN open collector driver would've sufficed. But with the input spec ranging down to 3V, you probably don't even need that.
'Looks like those 42 SSRs are gonna fit after all!
-Phil
What I'm trying to do is switch strings of xmass lights. Originally it was for 64 strings for a display. Due to the costs of the SSR's I'v cut the number of circuits down to 32. SSR's seemed like the best way to go, espically considering my lack of electronics information. After reading about charlieplexing to get the pin count down, and the propellers fast speed, I was sure I had the right plan. The original SSR's (the one that sparked this conversation in the first place) didn't have + & - on the controll lines and speced at 5V controll. They were priced at 3.20 each. The ones I found on friday have a internal diode, marked 3~6 V, and PIV of 6V, well exceeding any spec's I needed, but the price for these are 4.80 each. And now you introduce the speed issue. I'm whirling in all this information, almost ready to tell my Grandfather that I just cant do the project yet. Do you think this will work?
KK
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller + Hardware - extra bits for the bit bucket =· 1 Coffeeless KaosKidd
·
Let's move this discussion to a separate thread, okay? You can start it. We've gotten way off-topic here &mdash especially for a sticky thread.
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Propeller + Hardware - extra bits for the bit bucket =· 1 Coffeeless KaosKidd
·
www.semiconductors.philips.com/acrobat_download/datasheets/P82B96_5.pdf
The P82B96 comes in Big sizes (P-DIP) and costs about $3
Regards
Tim
a good high to low level shifter is the 74HC4050 I used it in other projects and it works very well.
Regards
Bernd
did you test the HEF4050 or the 74HC4050 ?
I use the 74HC4050 with 5 MHz· (hardware SPI) 5V to·3.3V·and it works great with no change in the pulswide.
Have a look at the attached scope-picture it shows·the 5 MHz clock.
(I think it will work also with higher frequencies !)
Regards Bernd
Now a test with 10MHz...
It works also·with 10MHz and the shape looks good.
See the second picture.
Regards
Bernd
Post Edited (MC-Bernd) : 11/12/2006 11:09:18 AM GMT
Thanks
Sid
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Sid Weaver
Don't have VGA?
Newzed@aol.com
·
all you nead is in the data sheed !
example[noparse]:([/noparse]5V to 3.3V)
In to Out(Pin3-> Pin2, Pin5->Pin4...) and VCC· 3,3V at Pin 1·, GND at Pin 8
(It's only a high to low level shifter.)
Regards
Bernd
Post Edited (MC-Bernd) : 11/13/2006 2:02:16 PM GMT
Thanks
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer
Parallax, Inc.
Before (Monday) I was switching the regulated 3.3V onto the pins.
What am I missing, please, and am I cooking the pins?
dira for pins 0-7 = 0
Thanks, Tom Perkins
The voltage you describe is probably what I would expect. Measure the voltage across the resistors and this will give you an
idea of the current you are feeding into those pins. You may be pushing current back into your 3.3 regulator.
http://www.parallax.com/dl/docs/prod/datast/28107.pdf
Thank you, Tom P.
across the resistor? With V=IR, you should only get this if you had current flow, and you shouldn't have steady-state
current flow on an input, unless the voltage is higher.
You still may be feeding current back into your supply though, depending on what the prop is doing.
Just for complete this repository of 5V solutions.
Dro.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
in medio virtus
Post Edited (inservi) : 3/3/2007 5:29:00 PM GMT
How to attach an images visible in this Topic ?
dro
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
in medio virtus
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer
Parallax, Inc.
Now, the discussion.
There seems to be some consensus on 1 k or 10 k. Even 100 k has been mentioned and my new hero Chip says that zero ohms would be OK as well.
I am a bit concerned about the 1 k. This is a fast chip. And we can get as much as 100 pF capacitance in a signal wire if it has some length and if a few inputs are connected to it. That would give us a time constant of 100 ns - that's a lot with a processor capable of producing pulses as short as 12,5 ns. A lot of unnecessary problems are likely to pop up if keeping to the 1 k series resistor. The more obvious is that many clock pulses will never reach threshold levels and cause frustrations for users that are not used to this kind of considerations.
I would go for 100 ohms, or 47 ohms. That would keep time constants· down in the 10 - 5 ns range and probably also be a better match to the line impedance (less reflexions). Remember that the 5 V parts you are going to interface with sometimes have nasty properties (like creating extra "quirks" in a slow edge and other mysterious things) and that a fast edge always is preferred over a slow one (OK, emission and ground bounce, I know, but...).
What is the general opinion here? Wouldn't it be good to have a standard way to handle this situation? And a standard that is good both to the chip and to the functionality? Like a 100 ohms series resistor.
Need to say it once more: I am in love again·- with a chip, not a chick!
Propeller Pin --+--/\/\---+-- 5V signal | | | | 6 pF === === X | | --- --- - -The capacitor to ground on the 5V side is the line capacitance, but this capacitance is charged at the slew rate of the 5V device, it is not current limited by the resistor. The RC filter which does need to be accounted for is the 6 pF which is the input capacitance of the Propeller, but it's order of magnitude is small enough that it's not much of a problem.
There can't really be a universal value, because the goal of an individual design may need to be optimized in a particular direction. For instance a product which must pass FCC EMI regulations will want to limit the slew rate to minimize RF radiation.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer
Parallax, Inc.
I want to hook up one of my propellers to·some CMOS dynamic ram chips. Reading this thread, it sounds like all I need is 1k resistors in series for the lines I will read from, and that even though the high treshold on the address inputs etc on the dram is 2.4V I should be fine.
My question is regarding the data lines... will the 1k resistor in series be enough? Or should I also pull the·dram side high to at least 3.3V?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
www.mikronauts.com - a new blog about microcontrollers
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
You really have to read MOSFET datasheets carefully. The gate-source threshold is the voltage at which the FET begins to conduct, and you might be pushing your luck if you get one with VGS(th) = 2.5V. You stated that the on resistance is 5 ohms, but look at the conditions under which that's measured: VGS = 10V. What you want is a "logic MOSFET", but it's not easy to find logic MOSFETs compatible with 3.3V logic that come in a through-hole package. (A search on DigiKey turned up none.) Could you use a surface-mount device?
OTOH, if it's a one-off project, you could just try a few and see if they work. After all, it's not like you're sinking a lot of current!
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
-Phil
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔