'Tait/Maplin Programmer (was Re: basic report ...)'
|> Hey, Andy, I've been hinting - send me a web reference to or
> e-mail of the Maplin programmer circuit and let's try and get
> it going.
The Maplin programmer is more or less the same as the ASCII sketch in
the file pic84pgm.zip (all programs mentioned are in my PIC archive).
It should work with the software in pic84v05.zip and numerous other
programs available on the web. It is a kit and I'm sure some people
have difficulties because they have not built the thing correctly.
Other problems are for the most part fixed by using a short connection
from PC to programmer, adding a 4.7k pullup from ACK to +5V and, like
a lot of these parallel port designs, adding a couple of 100-470pF
caps from RB6/7 to ground. From what Andy says Maplin don't give
users a lot of support. They have bundled some of my old software
with their kit and so I get a few moans sent to me, but to be fair I
also hear from people who didn't have any problems. PC speed is often
cited as a problem and that may be true for some PCs. I tried hard to
make the pic84v05.zip software as processor speed independent as I
could but I rely on the PC timer chip working the way I think it
should (you can check that by running tchk.exe from topic03.zip).
> > otherwise I recall that people have mantioned problems with incorrect
> > diode sourcing on the D.T. programmer.
I have at least 5 programmer designs on my web site and I'm not sure
which is the most likely culprit here - perhaps my 16C5X programmer as
I specify a Schottky diode for one component.
> The design does appear to be a bit dodgy and substitution of the wrong
> *brand* of part could be the problem.
Sorry about that. Apart from some head scratching when I was putting
together my very first prototype I've never had any problems with my
designs. Because of that I was happy to make them available for free
so that others could have some fun too. If my designs are really
"dodgy" and causing people lots of frustration it would pain me a lot.
Nowadays I can only deal with a fraction of the mail that my
programmer stuff and web pages attract (which gives me a permanent
guilty feeling) but at one time I spent a lot of time trying to help
people get their programmers working. I wasn't 100% successful as
trying to fix software/hardware by e-mail is not easy. On balance I
think my stuff is still useful but perhaps I'm wrong.
|part 0 3747 bytes
Sorry David I didn't mean to imply in any way that you are to blame for the
problems I have encountered, as I have said before ALL blame goes to Maplin
for not checking compatibility with different equipment, as they took on
your design and sold it is there responsibility.
I apologise if I have made you feel guilty about this, you should not. It is
totally feasible that I have made a total mess of things as I have a knack
for this (I once shut down the whole of dvlc time and attendance system for
two days and cooked the terminal server and erased several thousand
records ) and IM supposed to be an engineer!
One thing that has cropped up is that I had a message about RB7 being bad
but this has now disappeared, I have dug the programer out and will have a
mess around and get back to this space later tonight I hope, although I
havent slept for a couple of days so may keel over very soon!.
I have attached a diagram supplied by someone on this list and was wondered
what you might make of it.
Attachment converted: wonderland:com84.gif (GIFf/JVWR) (00029CEC)
Your efforts have been a source of inspiration for many.
The work you have done is/was excellent.
I have not tried the Maplin version, but I do use on a daily basis the DIY
programmer - it's great.
You get my vote.
At 17:59 10/02/99 +0000, you wrote:
David Tait wrote:
> On balance I
> think my stuff is still useful but perhaps I'm wrong.
Your not wrong !!!
Your WWW-Site was one of the first I found (on PIC4s) and helped me a
get into PICing !!!
Sorry, I didn't mean my last mail to be taken as a plea for sympathy,
plaudits or even more brickbats. If my stuff proved useful that's
great and there is no need to tell me. Critiques should be sent to
man.ac.uk but I must apologize in advance if I don't respond d.tait
quickly or at all.
|David Tait wrote:
> The Maplin programmer is more or less the same as the ASCII sketch in
> the file pic84pgm.zip (all programs mentioned are in my PIC archive).
OK, got that. Diagram is fine.
> It is a kit and I'm sure some people have difficulties because they
> have not built the thing correctly.
I'm sure that would be right! That's why they don't warranty kits!
> If my designs are really "dodgy" and causing people lots of
> frustration it would pain me a lot.
Hard to tell, isn't it? Especially given the point above.
My point about part substitution is that firstly, I'm uncomfortable
with 4066 gates being used to provide programming current; in the case
of the 16F84, this relates not to Vpp but to Vdd which is specc'ed at
50mA maximum which the 4066 is most certainly not designed to supply. I
realise this is by no means usual, but ... that's the spec.
I'm not even sure a 74HC4066 is intended to supply this current when
HCMOS gates in general are not. I'm much more comfortable with Don's
style of circuit using PNP transistors (which are probably cheaper
as well!). To be quite honest, I can't figure why David picked the 4066
in the first place, but then I have had some "fun" using a 4051 to drive
LEDs many, many years ago which put me off using analog gates in this
Secondly, the description of the gate used as "TTL O/C buffers,
74xx06, or 74xx07" worries me. Only the 7406/ 7407/ 7416/ 7417 and the
Texas Instruments brand of 74LS06/ 74LS07 are specc'ed to operate at
12 volts; other brands of LS gates, all HC gates and *all* 74xx05 are
specifically *not* designed for this application. They may perhaps
work, but the design sheets specifically say they *won't*!
> Nowadays I can only deal with a fraction of the mail that my
> programmer stuff and web pages attract (which gives me a permanent
> guilty feeling) but at one time I spent a lot of time trying to help
> people get their programmers working.
No surprises there. But it's a bit of a challenge for others to do
> On balance I think my stuff is still useful but perhaps I'm wrong.
Quite so, and by no means.
I still want to know what ICs Andy has in his kit, what results he
had on the "commissioning" test (setting the port to the appropriate
values and testing the voltages at the programming socket, using the
appropriate 100 ohm "dummy load" across pins 5 and 14), and those tests
passed, exactly what could and could not be done with the device.
In fact, http://www.dontronics.com/dt001cha.html#testing gives a good
outline of the steps, even though this is a different version.
|A few lines on Paul Webster's critique. They may be of some
interest to a few PICLIST members too. If that doesn't include
you then please accept my apologies.
> > If my designs are really "dodgy" and causing people lots of
> > frustration it would pain me a lot.
> Hard to tell, isn't it? Especially given the point above.
I agree it is hard to tell, especially for the Maplin kit. Maplin
just took my early stuff from the web, published an article and
produced a kit. The kit has nothing to do with me apart from that.
The story is here:
I have no control over the Maplin programmer but I can remove my
designs from the web if they really are causing more heat than light.
I guess my software is past its sell-by date anyway as most people
eschew DOS these days.
> My point about part substitution is that firstly, I'm uncomfortable
> with 4066 gates being used to provide programming current; in the case
> of the 16F84, this relates not to Vpp but to Vdd which is specc'ed at
> 50mA maximum which the 4066 is most certainly not designed to supply. I
> realise this is by no means usual, but ... that's the spec.
I cover this point in a rather elderly FAQ which was written before
I know all the things I know now:
You have to remember that despite the fact that many people may have
built 16C84 programmers in early 1994 there were NO designs in the
public domain. This was several months before the appearance of AN589
and Russ Reiss' Circuit Cellar article and before Henk Schaer's design
appeared on the web. I had a very early copy of the 16C84 data sheet
and the preliminary programming spec. There was NO spec on VDD
current during programming (that appeared much later) but there was a
spec on the risetime of /MCLR. I had some initial problems with
getting the PIC to go into programming mode despite all waveforms
looking perfect so I thought the risetime spec must be critical (I now
know the real reason for that spec). Anyway, the risetime requirement
was easily met using a 4066 switch. A few experiments showed that the
VDD current during programming was small enough to be supplied using
another switch with a 100nF reservoir cap (paralleling all three
remaining switches didn't seem necessary but is possible). That's the
history of the 4066 selection. I can understand your criticism given
the very high current that appears in the spec nowadays. I have
actually come to believe this spec is NOT a spec on the current DEMAND
but on the current SUPPLY. That is the current into VDD should be
limited to less than 50mA - which a 4066 does admirably :-) (I'm
probably wrong about that but based on my own tests 50mA is more than
an order of magnitude higher than typical). The fact that the thing
has worked flawlessly for me in all the years I've been using it has
perhaps coloured my treatment of criticism on this decision. Despite
this, in the later versions of this particular programmer package I
recommend PNP transistors instead.
> Secondly, the description of the gate used as "TTL O/C buffers,
> 74xx06, or 74xx07" worries me. Only the 7406/ 7407/ 7416/ 7417 and the
> Texas Instruments brand of 74LS06/ 74LS07 are specc'ed to operate at
> 12 volts; other brands of LS gates, all HC gates and *all* 74xx05 are
> specifically *not* designed for this application. They may perhaps
> work, but the design sheets specifically say they *won't*!
Again this was covered (though not so comprehensively) in my old FAQ.
> But it's a bit of a challenge for others to do
> so! [support my programmer stuff].
I agree Maplin should provide some support for their kit. I don't ask
anyone to support my own stuff - it is free and like anything that is
free comes with no warranty. I don't think I'm deceiving anyone. I
give enough information for a technically competent hobbyist to build
the programmer hardware and I provide all the source code for people
who simply take the stuff as a starting point for further
experimentation or even a product.
> I still want to know what ICs Andy has in his kit, what results he
> had on the "commissioning" test (setting the port to the appropriate
> values and testing the voltages at the programming socket, using the
> appropriate 100 ohm "dummy load" across pins 5 and 14), and those tests
> passed, exactly what could and could not be done with the device.
I have never seen a Maplin kit so don't really know what is shipped
with it. If you are very curious I will try to dig out the Maplin
article and send you a copy. For the reason's cited above I think
that "dummy load" test is unnecessary. The software in pic84v05.zip
has a built-in debugging mode.
If someone wants to program a 16F84 and has +5V and +12V available
then they can emulate my students (who can't be bothered with more
sophistication) and simply use a couple of resistors and the serial
port of a PC:
Thanks, Paul, for taking the time to put forward your criticisms
I will certainly bear them in mind next time I update my web
Re the picture of programmer you sent. This is the COM84
design I mention briefly on my PIC page. It is very neat but
can be tempramental. Jens Madsen has an improved version
which you could check out:
a more conservative version (one that needs power
supplies) is the serial port version of my quick-and-dirty
programmer. It is described along with a couple of other
P.S. I know you've just joined the PICLIST so you might think
I'm always writing long defences of my programmer stuff there.
You would be wrong - the recent messages I sent were the first
for well over a year!
From: David Tait <MAN.AC.UK> david.tait
To: MITVMA.MIT.EDU < PICLISTMITVMA.MIT.EDU> PICLIST
Date: 12 February 1999 19:21
Subject: Re: Tait/Maplin Programmer (was Re: basic report ...)
I will take a look at those sites, the only thing I
wondered about the pic was why it took so much to program when the levels at
the com port should be pretty compatible and (so im told) there are 12v, 5v
and -5v avalable there for powering things (within current restirctions of
As for the cp being set, I read the pic with a borrowed picstart programer
and this was infact not the case.
would I be right in thinking the maplin programer cant read? I have read so
much lately I forget which relates to what!!
|At 18:36 12/02/99 -0000, you wrote:
David Tait wrote:
>P.S. I know you've just joined the PICLIST so you might think
>I'm always writing long defences of my programmer stuff there.
>You would be wrong - the recent messages I sent were the first
>for well over a year!
This reminds me of the time I had something to say about the
tait style programmers some years ago.
Below is a letter I sent to David Tait by way of explaination.
Be warned though, It seems I was so nice and contrite back then
it is sickening to read.
This may want to make you puke just as it did me.
>You asked about a circuit diagram for the COM84 cct:
>I'm surprised it works but it seems that a lot of people
>are using it now.
Thanks David, I take a look if I get the chance. I know that you were a
pioneer with these direct driven programmers. Of late I have been critical
of some aspects of them. I certainly don't want to be critical of anything
in as much as it proves to be genuinely useful. I do think the promotion of
them has became over blown. In the last two weeks I have replied to four
requests for help on the piclist.
I do not know who has done what, but some of the designs, and there are
many of them, seem under done. If people are going to use them they may as
well work. I would like to see some moves among their suppliers to ensure
REPEATABLE quality. Maybe they need a few extra components and a PCB
instead of trying to be "as cheap as."
There is absolutely nothing wrong with a 16C84 programmer along these lines
as the 16C84 is self timed. Other devices need an accurate timing source.
This can be achieved, in ball park terms, in DOS with good software but in
windows, out of the question.
It would be good to see a minimum standard that defined voltage regulation,
and DOS and/or windows compatability. Windows compatible direct drive
programmers would require an onboard timing source such as a 555 timer.
This could be calibrated in by the PC in purely DOS mode.
Documentation should also meet a minimum standard so that people have
details how to test and trouble shoot their programmer. The problem is
beginners are using them and these people are the ones who can least afford
to have problems.
A current trend seems to be:
1) Save dollars on a cheap programmer
2) Throw dollars in the bin with damaged chips
3) Ask someone on the piclist to help
What started out as a good idea now seems to be running of the rails. I am
dishearten that people don't seem to appreciate quality either in their own
work or someone elses.
Anyway, these are just my thoughts on the matter and I don't want to be
offending anyone but I am a sucker for robust, quality design. My point is
that maybe the time has come for a "cheap programmer benchmark" to protect
those just joining the PIC fraternity. I don't want to be seen as rubbishing
someone else's efforts and I hope you don't feel I am attacking you or even
acting out of my own commercial interests. In fact I'm thinking of giving
it away. It has not been a good last week.
End of sickofantic email.
P.S. I'm off those pills now.
MPLAB compatible PIC programmers and firmware
upgrades for many programmers.
More... (looser matching)
- Last day of these posts
- In 1999
, 2000 only
- New search...