Re: Upgraded to Windows 7 - now I can't print anything!

Ohioguy <> wrote:
> I have a Samsung ML-1740 Laser Printer. It hooks up using a parallel
> printer port. When I upgraded my motherboard about 15 months ago, I
> bought one that didn't have a parallel port built in. I had to decide
> whether to buy a new USB printer, an adapter cable, or just get a PCI
> Parallel port card. The last solution was also the cheapest, so I got a
> Sunix 4008a card for about $14 shipped. It worked fine.

> Problem is, evidently I didn't save the driver install file like I
> normally do, so I had to track it down again and download:

> http://www.sunix.com.tw/cc/en/detail...34&prod_id=330

> Notice that the driver only goes up to Vista - no Windows 7

> It LOOKS like the driver for the PCI Parallel card installed ok, but
> I have no idea how to test the card properly. For example, if it was a
> network card, I could ping it. What do you do to test a parallel port?

You an get test equioment, but expect to pay fore than 100 EUR/USD for
it. Also, the paralell port is standardized and should not require
an extra driver. Unless those sorry excuses for software engineers
in Redmont dropped it from their "OS"?

> Thankfully, the Samsung website had a driver specifically for Windows
> 7 for the ML-1740. They are really on the ball. I got that installed.

> When I first tried to print a test page, it came up and asked me what
> to call the file. That was confusing.

> In the control panel, I've tried setting the printer to LPT1, LPT2,
> and LPT3. None of these seems to work - I just get an error.

> When I look at the Parallel port in device manager, it shows
> "communications port Com1" - this device is working properly. Device
> Type = Ports (COM & LPT) Port settings + 9600 Bits per second, 8 Data
> bits, Parity=None, Stop bits 1, Flow Control None. Under Port Settings
> > Advanced, I can set the com port for it from Com1 to something like
> Com200.

That wold be the serial interface.

> Anyone have any ideas on what I might be able to try so I can print
> again? As it is right now, I feel like I've exhausted the things I can
> think of to try, and I'm unable to print anything at all.

Well, get a driver for that card. It may of course be a case of
"if you buy cheap, you buy three times". In that case you can try to
get an LPT card that is certified to work in Win 7 or get an USB to
LPT adapter that is again certified to work in Win 7.

Arno

--
Arno Wagner, Dr. sc. techn., Dipl. Inform., CISSP -- Email:
GnuPG: ID: 1E25338F FP: 0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
----
Cuddly UI's are the manifestation of wishful thinking. -- Dylan Evans

On Wed, 28 Oct 2009 18:14:51 +0000, Arno wrote:

>> It LOOKS like the driver for the PCI Parallel card installed ok, but
>> I have no idea how to test the card properly. For example, if it was a
>> network card, I could ping it. What do you do to test a parallel port?
>
> You an get test equioment, but expect to pay fore than 100 EUR/USD for
> it.

Most test software for PCs will test a parallel port. You may have to
make/buy a wrap plug, but you won't pay *that* much.

> Also, the paralell port is standardized and should not require an
> extra driver. Unless those sorry excuses for software engineers in
> Redmont dropped it from their "OS"?

Good excuse for M$-knocking, but in this case there's a bit more to it.
Some of these cards use chips that really do need their own driver. I
know the Sunix serial cards need a subtle driver change; I had to do it
once.



--
Use the BIG mirror service in the UK:
http://www.mirrorservice.org

Bob Eager <> wrote:
> On Wed, 28 Oct 2009 18:14:51 +0000, Arno wrote:

>>> It LOOKS like the driver for the PCI Parallel card installed ok, but
>>> I have no idea how to test the card properly. For example, if it was a
>>> network card, I could ping it. What do you do to test a parallel port?
>>
>> You an get test equioment, but expect to pay fore than 100 EUR/USD for
>> it.

> Most test software for PCs will test a parallel port. You may have to
> make/buy a wrap plug, but you won't pay *that* much.

>> Also, the paralell port is standardized and should not require an
>> extra driver. Unless those sorry excuses for software engineers in
>> Redmont dropped it from their "OS"?

> Good excuse for M$-knocking, but in this case there's a bit more to it.
> Some of these cards use chips that really do need their own driver. I
> know the Sunix serial cards need a subtle driver change; I had to do it
> once.

Serial is actually more difficult, especially if you need things
like multi-port, because serial needs interrupts to work reluably.
You also may have serial parameters out of the ordinary spec,
again requiring drivers. Standard serial ports should also
not need external drivers.

Paralell needs interrupts only for advances modes like EPP/ECP,
but not for plain LPT, which also does not need parameters.
So LPT should not require any external drivers, unless the OS
designers either desided to be lazy or that you do not need them.
Both is unacceptable IMO, at least if you have a single port at
a standard MMIO location.

Arno
--
Arno Wagner, Dr. sc. techn., Dipl. Inform., CISSP -- Email:
GnuPG: ID: 1E25338F FP: 0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
----
Cuddly UI's are the manifestation of wishful thinking. -- Dylan Evans

Arno schrieb:

> Paralell needs interrupts only for advances modes like EPP/ECP,
> but not for plain LPT, which also does not need parameters.

What's the alternative for interrupts? A polling line printer
software/driver could work, of course, but is an anachronism in an fully
interrupt driven environment.

DoDi

Hans-Peter Diettrich <> wrote:
> Arno schrieb:

>> Paralell needs interrupts only for advances modes like EPP/ECP,
>> but not for plain LPT, which also does not need parameters.

> What's the alternative for interrupts? A polling line printer
> software/driver could work, of course, but is an anachronism in an fully
> interrupt driven environment.

So what? Just because a concept is old does not make it bad.
And in an environment with fast task-swiching, polling is not
that much of an issue, especially because there are no
incomming messages/events, unlike a serial driver, hence
polling is only done while actually transferring data to the
parallel device.

But the real difference to serial is this: Especially
with industrial serial cards, you quickly exhaust the maximum
2 (or was it 4?) interface limit a standard PC and have to
get creative with the interrupts, hence requiring a driver.
As serial interfaces are used a lot in industrial automation
the case where more than two are needed in one process control
PC is not that rare.

While there may be cases where people need more than two
LPT interfaces, these are very rare. And with a maximum of
two LPTs, it is completely clear how to operate the
hardware and which interrupts are used, hence no need for
an external driver at all.

Side notes: One reason parallel interfaces are usually not used
in automation is cable length limits and the simple need for
more wires. Also, a serial interface is bi-directional, while
for parallel this is either a more recent development (EPP/ECP)
or requires electrically problematic output sensing. Nether
of this approaches is robust against long cables, contrary to
the far better noise immunity of serial lines. And if you want
to insulate a serial line, you just need 2 opto-couplers, while
with parallel, you need 11 or more and you loose bi-directionality
in the bargain.

Arno
--
Arno Wagner, Dr. sc. techn., Dipl. Inform., CISSP -- Email:
GnuPG: ID: 1E25338F FP: 0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
----
Cuddly UI's are the manifestation of wishful thinking. -- Dylan Evans