An article about this programmer also appeared in Silicon Chip (Australia), March, 1999.
YES! NOPPP programs the PIC 16F84A, which is 100% compatible with the 16F84. But you must use .HEX files that were assembled for the right one (16F84 or 16F84A); the files are not identical.
No e-mail please...Thank you for your interest in NOPPP. After corresponding with over 500 people who have built NOPPP, I'm confident that it works reliably.
I am getting more e-mail about NOPPP than I can answer. I do not get any continuing income from the NOPPP project, so as far as I'm concerned, it's a finished project and I can't devote several hours a week to unpaid user support. Please get help locally if you can't get your NOPPP to work, since almost all inoperative NOPPPs are due to wiring errors. (Mixing up the leads of Q1 is the most common error; cheap assorted transistors often do not have the pins in the arrangement described on the box!) I can't see, or fix, a wiring error remotely. If you're sure there's not a wiring error, then either your PC parallel port is out of specification or you have a defective PIC chip.
If you must contact me, please read this entire document carefully before doing so. Your questions are already answered here.
If you are having problems with the Ramsey or Oatley kit, you should contact the kit manufacturer for help, and have them contact me if there is a problem they cannot resolve.
If you have a substantial technical question, not answered here, about the Electronics Now article, you can submit it to that magazine at q&a;@gernsback.com. They may answer it themselves or refer it to me.
I cannot design circuits free of charge, nor can I give free programming assistance to anyone other than my own students at the University of Georgia. If needed, I can do custom design work for US $75 to $100 per hour.
-- Michael Covington
Rob King has modified the NOPPP software to skip the read check, and is working on porting it to Windows.
Although not precisely "no parts," NOPPP is unusually simple and uses no hard-to-find parts. You can probably build it using parts you already have on hand. Here's the circuit (revised somewhat from the original design, for greater reliability):
Pin numbers are those of the 25-pin connector on the PC.
Capacitors are in microfarads.
Circuit description: On the PIC, pin MCLR is +5V for normal operation (not used here), +12V for writing, and 0V for resetting. Crucially, the +12V supply does not actually "burn an EPROM" -- the higher voltage is merely a signal to activate the internal flash memory programming circuit. It must be greater than 12.0 volts. The D0 output of from the PC controls this signal. No harm results from applying it at inopportune times.
The PIC communicates by means of a two-wire (plus ground) synchronous serial protocol. Pin B6 is the strobe signal; pulses on it tell the PIC when to accept or transmit each next bit of data. Pin B7 is both an input and an output. When the PIC is receiving from the PC, SLCTIN is held low and D2 does not conduct; D1 and R1 are effectively out of the circuit, and the PIC receives data from AUTOFD.
When the PIC is sending, SLCTIN and AUTOFD are high, D1 does not conduct, and D2 and R1 provide pull-up. Some additional pull-up is provided by R2 plus the pull-up resistor on AUTOFD inside the PC's parallel port (nominally 4.7k, sometimes much lower in newer CMOS parallel ports). The PC reads the data on the BUSY pin, which is 0.6 volt higher than the output of the PIC because of D2. The PC parallel port has (or should have) CMOS or Schmitt inputs and should not require true TTL logic levels.
R2 and R3 help to reduce cable crosstalk by isolating the input capacitance of the PIC so that less current flows during sudden transitions. The PIC has Schmitt inputs and does not object to the resulting reduction in rise time. R4 protects the base of Q1.
Revised circuit: The circuit originally published in Electronics Now and software are quite reliable. However, for greater compatibility with a few PC parallel ports that had trouble with the original, I made some slight changes and additions:
Yes!
A simple kit based on NOPPP can be purchased from Oatley Electronics. The kit consists of a circuit board and parts. (Located in Australia, Oatley Electronics takes American credit cards and ships worldwide. Note that an Australian dollar is smaller than a U.S. dollar.)
A more elaborate kit, with power supply, enclosure, software disk, demonstration board, and manual, is now available from:
It's called "Picpro," Part No. PIC-1, and sells for $59.95 + $6.95 shipping/insurance (in the U.S.A.; slightly more overseas). Credit cards are accepted.
The NOPPP software does not support this. Use the TOPIC software that is also included in NOPPP.ZIP.
The answer is unfortunately no, because I'm quite busy just helping people with the original NOPPP supporting only these three chips. If you succeed in modifying NOPPP to program other PICs, please let me know!
It works with all the PCs I have tried, which comprises over 15 parallel ports, as well as CPUs ranging from a 4.77-MHz 8088 to a 300-MHz Pentium II. The revised circuit shown above takes care of problems with a few nonstandard parallel ports.
You need +5 volts and +12 to +14 volts, both at very little current (under 20 mA). One solution is to use a 12-volt battery (or a combination of 6-volt or 1.5-volt batteries in series) plus a 7805 or 78L05 regulator chip to derive the +5V supply.
You can also borrow +5V and +12V from a disk drive power connector inside your PC, but make sure the 12-volt line is actually 12.0 volts or higher.
Here is a suggested power supply circuit for development work:
The extra diode raises the output voltage of the 78L12 by 0.6 volt, ensuring that it is within the 12.0-to-14.0-volt range specified for the PIC.
From Microchip, Inc., the manufacturer of PICs. You can download it; in the past, Microchip has also distributed it on CD-ROM upon request, but these are not always available. Contact your local Microchip sales office for assistance.
The complete package for Windows, including assembler and simulator/debugger, is called MPLAB and is about 7 megabytes in size. That's less than an hour of downloading at 28.8 kbaud.
It doesn't matter which version of MPLAB you get; all reasonably recent versions work fine. Recent revisions fix bugs affecting other PICs, not the 16C84/F84/F83.
If MPLAB is too big for you, it is apparently sufficient to download just MPASM (the assembler).
NOPPP will program any PIC16C84, PIC16F84, or PIC16F83 chip that you can plug into it. The cheapest and most readily available of these is the PIC16F84-04P. Here 04 means the maximum clock speed is 4 MHz (as opposed to 10) and P means it's in a plastic DIP package (not surface mount).
NOPPP is also believed to be suitable for the low-voltage PICs, PIC16LC84 and PIC16LF84.
You can buy PICs and PIC-related items from:
The last of these has a particularly interesting selection of resources for experimenters and hobbyists.
You can download PIC manuals in PDF form, or request printed manuals, from Microchip, Inc.
Beyond that, I particularly recommend the book Easy PIC'n, by David Benson, published by Square One (squareone@zapcom.net).
Advanced PIC programming is covered in Design with PIC Microcontrollers, by John B. Peatman, published by Prentice-Hall.
Intermediate between these is Myke Predko's Programming and Customizing the PIC Microcontroller, published by TAB.
You can buy books online from Amazon.
TRIS is a fully supported instruction in the PIC16C84 and related chips. Microchip discourages people from using it for only one reason: some of the higher-end PICs lack it, and they want you to be able to move up.
I find the TRIS instruction much more convenient than the alternative, and I see no reason not to use it. If I ever port my programs to a higher-end PIC, I'll be able to remove the TRIS instruction, or change it to a macro, at the same time that I make other changes.
After careful thought, I decided it wasn't. Here's why: The usual hazard from inserting an IC into a socket with power already on is that CMOS chips will go into "SCR latchup." This happens when an input is high but the V+ pin is not yet powered up.
In NOPPP, that doesn't happen. When you're told to insert the chip, V+ is powered up (+5V) but all the other pins are near 0 volts.
The NOPPP software is unable to detect the NOPPP hardware. With the original circuit, this is not necessarily a problem; the original circuit will sometimes not be detected if powered off. In that case, power up the NOPPP hardware (without a PIC in the socket, of course) before starting the software on your PC.
Your .HEX file was assembled for a different PIC (such as 16F84 rather than 16C84). It may not work if programmed into the PIC that you have currently selected.
Note that DEMO.HEX, provided with NOPPP, was assembled for the 16F84. To change it, go into DEMO.ASM, change both references to "16F84" to "16C84", and reassemble it (generating a new .HEX file) using MPASM or MPLAB. Reportedly, however, it works on the 16C84 without modification, despite error messages.
You are trying to verify a blank PIC without having programmed it, or...
This is a common symptom of a data communication problem; it indicates that the PIC is not receiving commands correctly. Try a shorter cable; check the circuit carefully; run the voltage checks.
Also make sure that your .HEX file was generated for the right kind of PIC (16F84 or 16F84A; they are not the same). The options when running the assembler, the directives in the PIC program itself, the choices you make when running NOPPP, and the actual processor that you are using must all agree. (Thanks to Evan Robinson for this tip.)
Note that this is an experimental circuit and may or may not work with your PC. Most people have found it less reliable than the circuit with the 1N34 diodes shown above. However, Yves Oesch has written to tell me that it works just fine after adding an 0.1-microfarad capacitor across the +5V and ground pins of each IC. He also suggests that instead of tying together the two inputs of the AND gates, we should use one input and tie the other one to V+. That way there is less of a capacitive load on the incoming signal.
Please note that the voltage checks in the NOPPP software do not apply to this circuit. I do not support this circuit; if you need troubleshooting assistance, please build the circuit that I support, the one with the 1N34's.
PIC consulting services
I provide a consulting service to design PIC-related software and
hardware for clients. See my main web page
for rates and other information.
Covington Innovations home page