redesign needs some extra eyeballs to critique

kittmaster

I "think" I've completed the redesign of the Ft232bm. Using the datasheet here:

www.ftdichip.com/Documents/DataSheets/DS_FT232BM.pdf

On page 22 of 25 configuration within that sheet.

Can anyone see any issues to this point? It seems like it should work, I looked to see if the unconnected pins were any type of problem, they appear not to be any issues.

I know the demo board recreation is no complete, I'm only interested in the power sequence which receive +5V from the main 8051 board with 1 amp of current. The 317T drops the 5V to 3.3 volts.

Anything stand out?

Thanks.....[noparse]:)[/noparse]

Ale

kittmaster,

The second picture has some problems, parts are missing.

The FT323BL uses ~ 22 picoF caps for the crystal, not 33 nanoF.
The LM317 needs a bit more current in the adjust pin, check the manual at national, but I think has to be in the several mA order, that will mean resistors in the hundred of ohms range... like 120 and 200 ohm. Why you do not get a LM1117-3.3, like a 7805 but for 3.3.

Peter Jakacki

I'm scratching my head as to why an older chip like the BM? Even FTDI say concerning the FT232R:

FTDI said...

The FT232R is the latest device to be added to FTDI’s range of USB UART interface Integrated Circuit Devices. The FT232R is a USB to serial UART interface with optional clock generator output, and the new FTDIChip-ID™ security dongle feature. In addition, asynchronous and synchronous bit bang interface modes are available. USB to serial designs using the FT232R have been further simplified by fully integrating the external EEPROM, clock circuit and USB resistors onto the device.

The FT232R adds two new functions compared with its predecessors, effectively making it a "3-in-1" chip for some application areas. The internally generated clock (6MHz, 12MHz, 24MHz, and 48MHz) can be brought out of the device and used to drive a microcontroller or external logic. A unique number (the FTDIChip-ID&#8482[noparse];)[/noparse] is burnt into the device during manufacture and is readable over USB, thus forming the basis of a security dongle which can be used to protect customer application software from being copied.

The FT232R is simpler but I guess you could do it with the BM but normally you don't use superseded tech for new designs.

Ale is correct about the 317, your adjust current levels are way too low. Besides the 317's dropout means that 3.3V out does not give it any headroom to regulate properly. Use an LDO regulator like the one Ale mentioned or what I use, the LP3985 or similar.

*Peter*

Phil Pilgrim (PhiPi)

I guess the crystal loading depends on which app note you read. The one I have says 27pF. It probably depends more on the individual crystal. Usually a loading spec is given in the crystal datasheet.

Vcc1 and Vcc2 on the FT232 each need a 0.1uF bypass cap located near those pins. The Prop should also have two 0.1uF bypass caps located near its Vdd pins, as should the EEPROM. Also, the recommended pullup for an I2C bus is 4.7K.

-Phil

darco

Oh, oops.... I made the same mistake on my projects... (using a 10k for the i2c pullup)

Any dire implications for doing this?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
—
darco
www.deepdarc.com/

kittmaster

I'll drop the 33's to 22, the crystal load variance shouldn't be that much.

I know the second image has stuff missing, I mentioned that in the preface, I'm really only interested in the serial eeprom, and composite and vga out and sound.....everything else is unimportant.

As I've mentioned in my previous posts, I went with the BM because of its lead package of LQFP. The R version only come in QFN and TSSOP, something I don't relish trying to solder or cut out on my routing machines. They are just to small and I'll have to risk hand soldering 100+ board if I go with that series. So the BM, for now, is my only answer.

Since the propeller and eeprom is the only chip needing 3.3V, and I don't have any other 3.3V sources on hand, the 317 is a cheap quick solution which I do have on hand to play with this. If there is loading effect, I'll just through a pot on R2 and dial in 3.3v under load.

I did size the resistor too high, that's a my bad, should be more like 120 ohm and 200 ohms.

Anything else jump out, I'm really more interested if the BM is set up properly per the data sheet and what the propeller may need. While I don't hold anyone hostage to yes its right, no its bad.......any input on its configuration would be helpful....[noparse]:)[/noparse]

deSilva

darco said...
Oh, oops.... I made the same mistake on my projects... (using a 10k for the i2c pullup) Any dire implications for doing this?

Understanding what the "Pull-Up" does helps you answering your question yourself..
(a) In a static state with high impedance at all parties, the Pull-Up connects the bus to Vdd.
(b) During comunication the transmitter drives the line, overriding the Pull-Up; to allow this the Pull-Up should be > 1k.
(c) When the transmitter leaves the line HIGH before going Tri-State everything is fine, and the Pull-Up can have any value above 1k.
(d) If not, or if the transmitter does only drive LOW (and leaves HIGH for the Pull-Up), then the Pull-Up has to deliver all the current to switch all capacities on the bus. This is around 5 to 10 pF per party, so maybe some 100pF have to be filled. As I2C signals normally run at 400kHz your RC should be lower than 1 µs which leads to R<10k.
(e) So this is your range: 1k to 10k.
(f) Note that this is a nearly worst case scenario with many parties, long lines and unintelligent driving behavior. Simpler configurations as Propeller and some EEPROM will flawlessly run with a 100k resistor.

Paul Baker

Looking at your USB section as per your request, the 33nF load caps for your crystal, is that the value specified by the crystal manufacturer? I don't see any bypass caps for the chip's main supply (there is a 10uF and two 0.1uF (though I think you can get away with just one 0.1uF)). Due to the Tx and Rx LEDs typically being a different color (and therefore a different Vf), it is recommend to use seperate resistors (recommended value 220&#937[noparse];)[/noparse]. Otherwise for the USB portion of your circuit everything checks out.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

kittmaster

Actually the capacitors are on page 13 of 25 listed as 27pF, what's 5pF among friends?....LOL...I just have to fix the schmetic....that was a copy and paste thing......

I didn't put any bypass caps, the main power regulator has a 100uF cap for power supply stablization. I'll add some bypass caps for the eeprom and propeller....np.....

I plan to use the same led color and style for both RXD/TXD leds, so that is why I wired it that way. I'll adjust the resistor the forward voltage of led. I used that value resistor based on what I saw in the schematic pdf .......which I assumed dropped 2v across the led for a 3.3v supply.

Sweet.........

This is a gonna be pretty wild to get up and running. Thanks for reviewing it guys.....

Chris

Paul Baker

I would still place the 10uF bypass capacitor, bypass capacitors must have minimal resistance to the chip it's bypassing to be most effective.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

kittmaster

I'll add it in...NP

Is this IC that sensitive or have that much inrush current that a huge 10uF for a power reserve right on the pins?

I understand the .1uf to kill any high frequency noise, just seems like a big value for a bypass.

Its moving right along....

Thanks!