IR link for HP 28s

[Frank Rottgardt]

Found this on Hpcalc.org

HP Read 1.3

"DOS program which reads data sent by infrared pulses from an HP calculator (such as an HP-28S) to a Radio Shack GP1U52X Infrared Receiver"

How well does this program work with a 28s and can it be used to save programs on a PC? Have not yet had a closer look on this Radio Shack IR receiver and if it still is available. What PC-hardware is required to be able to run this type of receiver (drivers, adapters etc.)?

Is there anybody out there having tried this program together with an 28s?

[sylvandb]

I don't recall anything special about the GP1U52X. It is a simple three-terminal (+5v, data, common) IR demodulator. I'm skeptical how well it would work with the 28s (was the 28s IR modulated or simple data burst?), but it sounds like it works sufficiently.

There are many modern equivalent IR demodulators and they probably would work about the same. However all of them are designed to work with consumer IR (remote control) and/or Serial Infrared (SIR) modulated signals, not necessarily the calculator IR.

The "DOS program" was the driver.

Hardware required is a power supply and a data input, both with the required voltage level (the old devices were all 5v, modern stuff is usually lower voltage).

The LIRC project has many plans for connecting IR demodulators to a serial port, using the handshaking outputs for power and data input and software to decode the binary pulses. However LIRC is intended for consumer IR (remote control) signals, not the calculator IR and not SIR.

sdb

[Frank Rottgardt]

Thanks everybody for the technical input.

However, it seems to be easier to establish the link to the 28s with help of a 48GX I also own and software provided from Christoph G. To get the data later from the 48 to my laptop should not be a problem. However, I am not talking about a ROM image, rather User RPL programs I would like to "save" in digital form.

So I wonder in what format the 28s is sending data to the standard HP printers resp. 48 (plain ASCII code??)

I think there also exists a HP printer emulator that can be used together with EMU48? Maybe the raw data transfered from the HP28 to the 48 can be loaded into EMU48 and send to the printer emulator in order to make User-RPL "visible" again? In the end I only want to make a print out of my RPL programs which is much more comfortable that writing them down.

[Eric Smith]

I've used it in the past with the 28 and 48. It worked fine on MS-DOS 6.22. I don't know if it will work properly under any version of Windows.

The GP1U52X had the advantage of being available from Radio Shack (is it still?), but an IR receiver/demodulator with a 32.768 kHz center frequency will work better than the 42 kHz of the Sharp GP1U52X.

The GP1U52X requires that its metal shield be grounded well, or it will not work well. The IR receivers from Vishay don't have this issue. You might try the Vishay TSOP4838, which is available from Mouser. It has a center frequency of 38 kHz, reasonably close to the calculator's 32.768 kHz. The TSOP4833 would be even better, with a 33 kHz center frequency, but Mouser does not stock it.

[Eric Smith]

Newark Electronics has the Vishay TSOP4833 in stock. The Newark web site incorrectly refers to it as an IR emitter, but the Vishay data sheet confirms that it is a receiver.

I've used the Vishay parts for other IR receive applications. I have not yet used one to receive from HP calculators, though I expect it should work quite well.

If you use the Vishay part rather than the Sharp, make sure you check the pinout and wire it correctly. Both parts have three pins, power, ground, and logic output, but they might not be in the same order.

Edited: 24 Oct 2009, 5:54 p.m.

[Christoph Giesselink]

Hello Eric,

from own tests the Vishay TSOP4833 isn't working in connection with the HP Redeye protocol. The earlier, but sold out, TSOP1833 does.

In every article you see that the TSOP4833 is the successor of the TSOP1833, but different TSOP4833 don't work where TSOP1833 work without problems.

Where's the difference in the data sheet?

After several hours I found a difference:

TSOP1833: Suitable burst length >=6 cycles/burst

TSOP4833: Suitable burst length >=10 cycles/burst

The HP-Redeye protocol defines 6-8 cylces (7 at the HP28 and HP42) for one IR burst and that seem too less for the TSOP4833.

Cheers

Christoph

[Eric Smith]

Good detective work! I hadn't noticed the minimum burst spec. It's a shame that the new part won't work. It appears that they phased out all of the old parts that were specified for 5.0V +/- 10% operation, in favor of new parts that work over a wider supply range.

The Vishay web site lists some parts designed for "fast" IR protocols. Perhaps the TSOP4133 might work. The burst and gap specs seem similar to those of the old TSOP1833.

[Christoph Giesselink]

Hello Frank,

using such receivers is quite a problem today. Solutions like HP Read need a native DOS operating system. At end of the 80'ies there where a commercial product called PrintHP sold in Germany. This product used a control line of the serial port for input. But a 486DX2 66 was too fast for this solution. I personally wrote a DOS receiver program for the PrintHP hardware in the early 90'ies (latest change in 1993) which worked also on faster PC's.

My latest try with my own written receiver was some years ago was on a P3/850MHz on MSDOS 6.2. It worked, but was very critical to get a proper result. With this program I captured all my UserRPL programs on the HP28S in 1993. After removing the checksum bits, the HP28S output is ASCII in Roman8 character set. About 10 year later I used the captured data to transfer them to an emulated HP28S. :)

My personal solution today for capturing data from the Redeye port is a self made hardware with a TSOP1833 receiver, a 8052 compatible MC decoding the data and sending them over RS232C to the PC.

Drop an email if you want to have a sample of a HP28S printer output.

Cheers

Christoph