Help! I Need A Serial Controlled PWM Chip or Chips That I Can Get 24 Outputs Fo
DigitalDj
Posts: 207
Hello!
I Need A Serial Controlled PWM Chip or Chips That I Can Get 24 Outputs For RGB L.E.D. Control! If i can adress each chip it would be great! It would be nice if it had other features for RGB control also.
Thanks,
Kevin
·
I Need A Serial Controlled PWM Chip or Chips That I Can Get 24 Outputs For RGB L.E.D. Control! If i can adress each chip it would be great! It would be nice if it had other features for RGB control also.
Thanks,
Kevin
·
Comments
Here is the PAK-V PWM coprocessor by Al Williams, which offers 8 channels per chip, and 2-32 additional chips can be daisy chained for up to 256 channels, using only one Stamp pin:
http://www.awce.com/pak5.htm
Regards,
Bruce Bates
Thanks! I guess i should have mentioned that this is what i started using but i am having some problems with the address lines that actually caused the chip to fail twice and i can't get Al from A L Williams to give me any answers as to why this is happening. I have narrowed it down to pin 22 (address pin) that is causing the problem and they won't enlighten me as to what may be causing the problem.
Has anybody used this chip and may be able to help?
Thanks,
Kevin
·
The PAK-V is an 18 pin chip, so I'm not sure what you're looking at when you mention pin 22. Care to check the documentation again?
Regards,
Bruce Bates
Maybe i should have said PAK Vc! Here is the website for the chip!
http://www.awce.com/pak5.htm
Thanks,
Kevin
·
Anyone know if something is wrong?
Shawn
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Shawn Lowe
Remember - No matter where you go
There you are.
This is the latest email that i sent him!
Al,
I was setting here thinking and i have a question to what you mentioned about a ground loop, i have 2 wires from the computer for the serial out 1 is the signal and 2 is the ground would this cause a problem? I am running a seperate power supply for the 5V and ground for the Pak.
·
I would think this problem would have shown up before any of the address pins were grounded. Unless the internal pull up resistors are allowing enough current when grounded to cuase the ground differential.
·
Let Me Know What You Think!
It's just been hard hard to get any advice from him! It actually tok about 7 or 8 emails to get him to explain to me how the addressing works!
Kevin
Al has been out of town, but I'll make sure he's aware there are some pending questions awaiting him. You're not being ingnored, honest!
Regards,
Bruce Bates
Just based on what little you've said there, I'd connect the PAK ground and the Stamp ground together. That should eliminate any potential ground loop problems.
DISCLAIMER: My answer is only based on the minimal information contained in your posting here. You and Al may have had a much more thorough discussion of the problem you are having.
Regards,
Bruce Bates
This is a very large project that is going on that i have with a digital jukebox that i have designed and almost have finished and i am also designing lighting to go in it. I have a internet deal on the jukebox for plans & kits to be done when the jukebox is completed but no deal yet on the lighting. This problem has really held me up on the programming because the programming has to be setup for addressing each chip in manual control mode and auto sequence.
Kevin
Post Edited (DigitalDj) : 2/20/2005 4:55:42 AM GMT
Two observations. If this has nothing to do with a Basic Stamp, and actually has nothing to do with anything from Parallax, this thread should be in The Sandbox instead. Read the first post in the "Support for using this forum" forum about getting it moved. You'll probably have more luck as there are people in The Sandbox that are pretty hardware helpful, but don't deal with the Stamp.
The second is that, if you haven't already, draw up a schematic and post it. It's really hard to determine what a hardware interface problem is without seeing the interface.
Jim
I know this chip is used alot with the stamps and i just figured i would get more help here, i guess i figured wrong(Sorry). I put in a request for the move to be made with the moderator. Sorry if i have affended anyone with non stamp problems!
I can Draw up a schematic or block diagram if someone wants to help!
Thanks,
Kevin
Actually, if you will be selling plans and kits for this, a schematic should really be part of that. I'm not familiar with the various schematic tools out there, but I've used Visio with some success. Others have used the free schematic tool that Express PCB has available. You can download it at www.expresspcb.com/ExpressPCBHtm/Free_cad_software.htm.
Jim
I don't have a internet deal with any of the lighting just the jukebox!
I have attached a drawing of what i have when i dicovered that pin 22 is what was causing the problem. The reason i say pin 22 is because in the beginning·with setting the address·pins i was using pin 22 as least signifcant instaed of pin 18. So when i set the chip for address 1 i left pin 22 high instead of 18, i had everything backwards just didn't have the address correct and the program was not enabling the chip. When i reversed everything and set all the address lines with 18 being high the only thing that changed was 18 now high & pin22 low and that's when the chip blew. This has happened with 2 chips and this is the question i asked Al from A L Williams as to why this would happen and he said there was no problems with the chips.
I have also attached the location on the manual!
http://www.al-williams.com/pak5c.pdf
Thanks,
Kevin
Thanks for the link to the manual. When you say you left a pin "high", do you mean you left the pin open, as in not connected to anything? The docs say a pin not connected will be high. I just want to verify that you left it not connected as opposed to forcing it high with +5vdc.
The docs say that pin 7, RXINV, should be tied low for "true serial I/O". You show this pin tied low, but also with a connection to the host serial port. Is this line what you are using to connect to ground on the serial port? If so, that would seem ok.
A big concern I have is the pin 8 connection to the host. This docs define this as TTL level RS-232. However you are connecting this to the serial port on a PC which sends -12 to +12 vdc. I don't think that is a Good Thing. You need to use a level shifter like a MAX232 chip to allow the PAC-V to see only 5 volts.
Jim
Thanks,
Kevin
·
1. Yes pin high·i left it·unconnected.
2. The RXINV is grounded and this is also the ground to the computer DB9 connector
I did notice on the old 18 pin Paks they use a 22k resistor and it does say you can connect to the DB9 connector, i may have got this confused with the new PAKV chips.
A l Williams website has a sample visual basic program that can be used to control the chip but doesn't say anything about not being able to connect it to the DB9 connector on the computer and doesn't say anything about using any other components.
I thought the DB9 connector was TTL RS232?
This page shows the PAKV on a circuit board with a DB9 connector! It looks like there is a resistor by the DB9 connector on this board not sure of the Ohms. This is not called out anywhere in the documentation as well as setting up the address lines like you were going to set them up in programming.
http://www.awce.com/gp6.htm
Kevin
Post Edited (DigitalDj) : 2/21/2005 12:09:38 AM GMT
I am basically searching at this time because i do not know what would cause the chip to fail just by the address lines!
Help!
Kevin
·
The purpose of the 22k resistor on the serial line is to drop the voltage from the PC's serial port down to TTL level (0-5vdc) for the PAK-V chip. You'd probably want to confirm it with Al, but I'd be willing to bet that straight RS-232 voltage levels into the chip would cause the magic smoke to leak out.
Jim
In general to prevent things like this from happening, (IE this is very good SOP (standard operating procedure)) don't ever tie an input directly to either ground or supply. Instead use a SIP bus resistor, these are resistors which have one end tied together and the other ends are availible for connection, so an 8 resistor pack will have 9 pins, 1 common and 8 other terminals. A SIP is a single line package with the same spacing as a DIP so it breadboards easily. For your application, you would tie the common pin to ground and tie a seperate pin to each PAK input pin you want to tie low. If your pins are input, the low current draw causes the pin to see the low voltage (very little voltage drop across the resistor). If you accidently tie an output to the node, instead of causing a short circuit, the excess current is safely dissapated in the resistor. Plus it allows you to notice this problem by measuring the pin voltage with a DMM.
Paul
What·would the voltage be on the serial line when not active at the chip?
Do you think this would have an effect on the address lines.
Why wasn't it showed in the new PAKVb/c manual?
Kevin
·
I would really enjoy being able to wire up everyone's circuts and test them, but the reality is even when I was personally·running AWC every day all day, we just can't prototype everyone's circuits. We try to help people in every way we can, but short of rasing the price of each chip to $500 or something, we can't offer custom design work to everyone or even anyone.
That being said:
1) Kevin: I know you are frustrated. If you decide to stop pursuing the PAK-V, then return everything along with all your order numbers and AWC will refund your money.
2) You can't hook the PAK directly up to a PC. The +/- 12V will greatly stress the input protection diodes and bad things will happen including a hot chip that is eventually burned up. The 22K resistor in series should work fine, but directly connecting it is not fine. I suspect in retrospect that this could be where you've burned up chips.·The manual clearly states this is a TTL level input. The 22K trick is not the preferred way to do it (a MAX232 or similar translator is best) but it is possible to make it work.
I am still having terrible e-mail problems -- I'm trying to get a connect long enough to grab about 5 days worth of mail (about 150-300 mails a day ads up). So I apologize for making Kevin this offer in public, but it may still be awhile before I can read the rest of his incoming mail and reply.
·
Yes i'm not a engineer and i am a experimenter, project builder that's where the fun is and that's why there is this discussion board.
I appologize to everyone that this discussion took off on the PAKV, this was not the intention if you look at the original post. I also appologize for this being a way to settle Al and my problems. But i feel that since i was having problems and i gave as much info to begin with and a picture of the circuit·that i expected better input and questions or asked for a diagram of what i was doing to solve the problem. It's just like most busineses they get your money, answer half of your email questions because they don't read it all.
Al, The offer was great but that's not what i want, i like the chip and just want the support and changes to be made, i have alot of time on this project. I believe that sending me another chip would suffice.
Now! I feel better, but still have to get the circuit to work!
Kevin
·
I think you are being excessively harsh on Al. While I've seen sites with better examples, his are pretty good. In fact, they are better than many products offered by big name companies. His pin out table clearly states that the serial connection is TTL level. If you didn't understand what that meant, or at least the implications of it, and hooked your chip up to the incorrect voltage levels you have only yourself to blame. Ask anybody on these forums that has hacked around, even the most experienced, if they've blown a chip before. Everybody does it periodically. But they don't blame the manufacturer for their own mistakes. When you buy an electronic chip there is a certain level of knowledge that the buyer must take upon themselves.
As to support levels, I've never purchased from Al's company so I don't know how their support is. But support costs money. Large quantities of support comes only from the money that huge volumes can provide, or that very high margins pay for. If you want a small company to review your proposed design, then provide full engineering support on a rapid turn-around basis, you should expect to pay significantly more money for the product than is being charged for the PAC-V.
Just my $0.02 worth.
Jim
Kevin