Integrating the "prop plug" circuitry into my own board?
Mahonroy
Posts: 175
in Propeller 1
Hey guys,
I am wanting to build a device that has the prop plug circuitry built into it, so the entire device can plug into a USB port and use serial communication (full duplex serial library to communicate with a piece of software on a computer).
I came across this thread:
http://forums.parallax.com/discussion/135764/ftdi-usb-interface-to-propeller-circuit-variations
And it has this circuit schematic:
Can I just integrate this into my board using those parts, and that same windows driver thats used for the prop plug and be good to go?
What do you guys recommend?
I am wanting to build a device that has the prop plug circuitry built into it, so the entire device can plug into a USB port and use serial communication (full duplex serial library to communicate with a piece of software on a computer).
I came across this thread:
http://forums.parallax.com/discussion/135764/ftdi-usb-interface-to-propeller-circuit-variations
And it has this circuit schematic:
Can I just integrate this into my board using those parts, and that same windows driver thats used for the prop plug and be good to go?
What do you guys recommend?
Comments
-Phil
-Phil
Here's my reference to all of the various Serial to USB circuits that I've seen Parallax use. Hope it can be of help to others doing this.
One thing I don't recommend that is found on most of the Parallax schematics is a capacitor connection directly from the DTR output of the FT chip to the base of the transistor that provides the reset. It does work, but it severely overdrives the transistor and creates a nasty transient. Use a resistor in series with the capacitor to the base, as is shown on the circuit in post #1. The capacitor can be quite small, ~100pF. I don't use the buffers, but to forestall phantom power to the FT chip, I do check to see if anything is connected to the USB port and only start the serial port object if so, plus resistors in the rx and tx paths.
Please don't let this be the last word on what should be a reference design. As mentioned the transistor should have a current limit resistor for best design practice and in mine I actually use a digital transistor with built-in resistors. Also there is no provision for powering the Prop from the USB power and all it needs there is one extra pin while still being compatible with the standard 4-pin arrangement. In this case I typically use my 4+4 layout with I2C and +5V on the other pins or a 1x5 arrangement with the +5V next to the ground. Although I don't use I2C from the USB I do use I2C to read and program the EEPROM directly with a clone dongle or from the Bluetooth module. Sometimes the "prop plug" does double duty as a connector for an external display etc.
In the same form-factor as the 4+4 arrangement I also have a Bluetooth module which can plug in in place of the USB although swapped 180' so that it is kept internal.
btw, since I tie the USB +5V directly to my board it also means that if I powered my board externally it would also supply 5V onto the USB port. In the USB spec this is a no-no but in all the many years I have done this on all kinds of equipment I haven't had any problems with doing so plus I reason if gimmick manufacturers can sell USB coffee cup warmers and so on, then the USB spec is already being abused anyway. My way of doing it also means that I never have a problem with phantom power reset issues. Seriously, I don't have USB problems.
Here's a circuit of my prop plug although the resistor network isn't normally loaded plus the DDTD transistor has built-in resistors plus I actually use microUSB connectors with through-hole mounts.
I use a SOT23 transistor with internal series resistor to the base, and a 10K between base-emitter. There are a few different versions with 470R and 1K series base being the easiest to get. I use an external 10K pullup on the prop /Reset pin.
I now use a microUSB connector with thru' hole frame mounts.
I use a 1x6 connector where the centre 4 are the standard GND,nRST,SI,SO. The pin before GND is +5V from USB but I also have a link in this line. I don't use the last pin following SO.
My latest designs don't usually have the USB on the board. I provide the transistor reset circuit on my pcbs with 0.050" links than can be soldered or use the rather expensive 0.050" pins and a link. Same goes for the EEPROM - I have a 10K pullup on the nWP pin and a 0.050" link to ground. This gives me the option to write protect the eeprom.
I now use the cheapie eBay CP2102 board (less than $2 shipped) that has the DTR brought out (not all cheapies bring this pin out). For reference, there are 6pins: DTR,RXD,TXD,+5V,GND,3V3. Do not use the 3V3 as there is not enough current for the Prop circuits.
For this reason, I have changed my boards plug pinouts to suit these USB boards. On my P8XBlade2, unfortunately I just could not squeeze the *RST pin next to the SO so it is necessary to bend the DTR pin to the next position.
Note I cut off the 3v3 pin from the CP2102 board.
(see my P8XBlade2 link in my signature for the whole circuit)
"use this design" is more than an opinion or personal preference. Parallax do indeed know how to make their own prop plugs but that does not necessarily mean that the person who designs it is necessarily the best designer or their best design. Please take into account that Parallax are not the originator of USB to logic level serial converters, there are many of these converters out there, all based on FT232 or CP2102 or CH304 etc. All these manufacturers know how to make their chips and have recommendations for their use as well. In the end we have to be the best judge of what is suitable though.
This in no way says anything against Parallax products, they make good quality products, it's just that they could be and I think should be better. Besides the OP's question was to the forum and the reply is from the forum. That original circuit looks overly complicated with the gates etc yet it doesn't even have pullups on the RX line and there needs to be bulk capacitance across the VCC as well, usually 4.7uF.
-Phil
The problem of the FT232 being parasitically powered from the SO/P30 prop pin is likely typical of a lot of chips. It is just that they are powered from the same power supply. The issue is the clamping diodes to protect from static and human body discharge. Many years ago, chips died for apparently no reason but when they were decapped static discharge was an extremely common problem.
To me, the issue is that the FT232 is an overpriced chip, making the PropPlug an expensive product. That in turn makes prop boards expensive.
Originally that was probably not the case, but it certainly is these days, and hence why we see other so many cheaper alternatives. Heck, they are even inbuilt in many micros these days.
Design inertia for new designs, or what?
I think it boils down to propagation of copy/paste schematics. The schematic you copy may not be the newest/best design, but I think that is what happens sometimes.
FTDI too is on a learning curve and their products evolve and improve to keep up with technology and standards.
The FT231XS will work fine in exactly the same circuits as the FT232RL, cut and paste, original or bettered. One only has to adjust the pin numbers. I choose it largely because it has far better support for compliant USB->LiPoly battery charging.
Just to be clear, I'm with you. I look around for improvements to circuits before copy/pasting, but I think the "don't re-invent the wheel" concept is ingrained in us and we sometimes fall victim to it.
"Why is anyone still using FT231XS, the CP2102N is a better chip, smaller package (3mm vs 4mm) and costs less $1.33 in singles from Digikey )"
Part codes for this are
CP2102N-A01-GQFN20
CP2102N-A01-GQFN24
CP2102N-A01-GQFN28
So new, it looks like stocks are due Digikey 4 Oct ...
Thanks Bob.
No new version. I would have to see if I have any pcbs left although I can get more. I have a number of real FT232RL remaining.
However, I now use the 6pin USB to Serial from Alicennnnn (forget what the numeric number is) on eBay. Last I looked they were $1.45ea posted. Of course you need the transistor reset circuit which all my latest boards have as a linked option. Plus you can power from the USB too.
They are the CP2102 QFN dated 1507. Works great as I use them all the time now, even on W10. Don't even recall needing to load a driver as Windows did it automatically I think.
You've got me there. The price and the capabilities of the CP2102 family and the new 'N version especially look to be great.
I do have an aversion to QFN packages, due to routing issues and also due to headaches in manufacturing. Mainly though I have the "familiarity" syndrome. I've used the FTDI products (in tssop) for years now. I've perused their data sheets and app notes backwards and forwards. I know their FT-Prog configuration utility, and I have my layouts and my configuration templates all set, and I have hundreds of units deployed out in the field. If I start using Silicon Labs products, I'll have to start over with data sheets, app notes, layouts, templates, configuration utilities and drivers, along with the uncertainty (good or bad) that comes with applying any new chip. And I'd have to keep the wherewithall to support both FTDI and Silicon Labs. I have a small business, and I do understand design inertia! The FT231XS is $1.98 @ 100 vs the CP2102N-24 at $1.25 @ 100.