JP1 Remotes

[Vynce]

I would like to be able to control my projector (ECP 4100) with my 6131. There is a pronto CCF with all the codes here: http://www.remotecentral.com/ftp/ccf_templates/projectors/el-proj-marquee.zip

I have tried to convert the CCF with DecodeCCF, but it is unable to determine the protocol. Here is some sample output:

[jon_armstrong]

I know there was an attempt made to use the device. I have an email from John Fine who wrote a protocol for this and here are the salient features:

From: John Fine
Sent: Thursday, January 09, 2003 10:20 PM
To: 'jon armstrong'
Cc: 'Jason Anniballi'; 'Rob Crowe'
Subject: Electrohome protocol

I just failed in my first attempt to test this (because I couldn't get
IR to understand the result of learning it from one remote to another).
I'll try more tomorrow.

Upgrade protocol 0 = 01 E1 (S3C8)
37 82 01 8B 13 CF 84 50 08 07 00 BE 00 AA 00 BE
00 AA 59 42 00 BE 02 E6 58 03 E6 03 3E 6C 1A A6
C6 02 6B 0A A6 C6 18 EB 09 37 50 02 6C 06 59 03
5C 3E F6 01 46 6A E8 AF
End

Upgrade code 0 = 0C 57 (Cable/1111)
E1 1C 41 0A 0B 0C
End

(The protocol 01E1 and device 1111 were just numbers I selected at
random.)
It's set up for any newer S3C80 remote (though it won't be "Cable" for
some of those).

It takes an 8 bit OBC. The top 7 bits are transmitted. The bottom bit
tells whether to transmit that 4 or 22 times. It also transmits the
start/stop command 2 times before and after that main command. (All
that is as designed. I failed to test).

I just emailed Jason to see if he can help. If nothing else the protocol has probably been written or is close. IIRC, the low and high frequency units used the same protocol just a different frequency, and yes an S3C80 JP1 can put out an IR carrier frequency at least up to 1MHz. The one above is 42.3 KHz.

I have the same ccf files that you referenced and can probably figure out the commands.
_________________
-Jon

[Vynce]

Cool! Thanks for your response. So I basically need a working protocol upgrade and then I can use the info from DecodeCCF to create a device upgrade in RM.

[jon_armstrong]

AFAIK, DecodeCCF/DecodeIR.dll won't decode that Protocol, but the command set is in the notes of the ccf files.

Let's try to reason through whethwer it is the high frequency version or the low frequency:

Some questions:
1. How do you control the device now?
2. Do you have the OEM remote?
3. Do you have any OFA learning remotes with the JP1 cable.
4. Have you ever gotten learned commands to work using any learning remote?
5. If you have the OEM remote try leaning a few commands with any learning remote that you have and see if it works.

There is probably some exception to this statement, but it is very unlikely that a learning remote can learn the high frequency signals.

I will still create a high frequency version of this but, I would be nice to test knowing which is more likely to work.
_________________
-Jon

[Vynce]

I currently control the projector using the built-in wired remote. I do not have the OEM remote or any learning remotes, but I do have a modified 6131 with a JP1 cable. A digital storage oscilloscope would be useful, but I don't have one of those either (yet ). I have a UIRT2 IR transceiver connected to my PC as well as a simple LIRC-type receiver.

Apparently high frequency units should have a yellow sticker/band on the actual IR receiver. Mine does not have this yellow sticker, which means that it is probably the low frequency type. The sticker could just be missing though.

I would love to know what frequency my projector uses, but I can't think of any simple way of testing it

[jon_armstrong]

Vynce,

We can find the frequency "indirectly".

I never heard back form anyone so I tested the protocol made a minor tweak and I think it now works. There are two upgrades one for the low freqency and the other for the high frequency. Ionly assigned a fe commands in the buttons tab. There are numerous power commands and I'm not sure what they do. The Power command set on the button requires 22 repeats, that John Fine put into his upgrade. And they are there because I counted them

Most of the rest require only four repeats. Keep in mind that the reset command may reset convergence and the rest of the commands came from testing by the Pronto guys. Some of the commands match up with a very poor copy of three pages of the factory manual for an ECP projrector of some sort.

So you might try them with the Power On and then test the Power. Please refine the device upgrade and we can repost. Here is the link.
_________________
-Jon

[Vynce]

Cool! I didn't think there would be anything to test so soon!

I just tried both the 42kHz and the 336kHz variations. They both work. The 336kHz version seems to work quite a bit better than the 42kHz version. Neither of them work very well though.

I tested power toggle (POWER), discrete power on (Power On), discrete power off (Power Off), contrast (CONT), brightness (BRITE), and the arrows (Up, Down, Left, Right).

With the 42kHz version, only the power toggle (POWER) worked. It took several tries, angles, and distances for it to do anything.

With the 336kHz version, the power toggle took less fiddling than the 42kHz version, but it still didn't work right away. Discrete power on and off didn't work. Contrast and brightness worked, but not consistently. The arrow buttons didn't work.

Here is a pdf of the two pages of the service manual that contain IR info. Perhaps this is what you already have. I have the rest of the service manual if you need it (it is 42MB though).
http://vynce.home.mchsi.com/ecp4100/ecp4100_ir.pdf

Thanks for your work on this. Hopefully we can get it working reliably.

[jon_armstrong]

I'm going to guess it is really a high frequency model. When I changed the frequency I forgot to bump up the duty cycle to 50%, that you need for higher frequencies. I just modified the file and you can download and try again -- same link as before. The only thing that you need to do is delete the previous protocol upgrade and add the new one.

Or you can just copy it from here -- the two bytes in bold are what has changed:

Upgrade protocol 0 = 01 E2 (S3C8)
0A 0A 01 8B 13 C4 85 10 08 07 00 BC 00 A8 00 BC
00 A8 59 EF 00 BC 03 98 58 03 E6 03 BE
F6 01 46 59 03 56 C5 03 EB 02 5C 0B DF C0 03 02
55 F6 01 46 5A FB E6 03 BE F6 01 46 8D 01 46
End

If I'm right, that should now work much better.

BTW, that pdf is a much better copy, of what we saw before. If you convert those hex commands to decimal, multiply by 4 and add 1 and put those values into KM as an OBC, you get the 4 repeat version. If you just multiply by 4 you get the main command repeating 22 times.

I think all those commands are already in there.
_________________
-Jon

[Vynce]

Ok, tried the new 50% duty cycle protocol. It works, but it is still unreliable. If I repeatedly press the brightness button, for example, it will work every 2 to 12 presses - 5 presses seems to be the most common.

[jon_armstrong]

Just as an experiment, first try sending the command with just a tap. Don't hold the button down. Then if that doesn't improve anything, change all the "OBC's" by subtracting 1 except for Power. That makes them repeat 22 times. Just tap those too, since a total of 26 segments will be sent without holding. Then, if that still doesn't change things hold it down.

In the pdf you sent, there is a diagram of the ideal signal. From what I saw in testing the Protocol, it sends exactly that command with all the requisite repeats. The timings were also very close.

The Pronto files send only the two first segments and the second segment repeats as long as the button is held. The final segment (end code) repeated twice is never sent. So trying the long signal will be a close approximation of the Pronto. On some up and down type commands it will probably incement or decrement too much. At least it will give us some indication of what is going on. I don't have time now but I'll look at things like total frame length and tell you what paramenter to tweak.

When I tested, I didn't test long button presses because I want to see all segments. So it's possible that the entire sequence repeats, although I believe that just the center data segment should repeats. The spec is unclear as to whether the 4 and 2 repeats are minimum for some commands.

It will be important to narrow down specifically which commands work more reliably in what case. If would expect the right and left arrows to not work right with 22 repeats, but some others would work better.

Finally, try a different set of batteries. Even "new" ones out of the pack can be defective.
_________________
-Jon

[Vynce]

With the buttons set to repeat 4 times, tapping requires several taps before it works. Holding down the button seems to make it work faster, but there is still a delay - the command still has to be repeated several times before it is recognised.

With the buttons set to repeat 22 times they always work, but there is still a delay before the command is recognised.

The arrow buttons never work - tapping, holding, 4 repeats, 22 repeats. The discrete power buttons also never work, probably because they seem to repeat 4 times and actually need to repeat 22 times.

[jon_armstrong]

Lets try reducing the frame length slighty.

The Spec calls for 46,046 uSec. In the sample I did, the measured frames all are 46,456. The Pronto hex files have 45,700 (theoretical from decoding the Pronto hex).

To change the protocol to that frame length here is an updated protocol upgrade for the low frequency Protocol.

Upgrade protocol 0 = 01 E1 (S3C8)
37 82 01 8B 13 C4 85 10 08 07 00 BC 00 A8 00 BC
00 A8 58 78 00 BC 03 98 58 03 E6 03 BE
F6 01 46 59 03 56 C5 03 EB 02 5C 0B DF C0 03 02
55 F6 01 46 5A FB E6 03 BE F6 01 46 8D 01 46
End

You can just edit those numbers in bold in IR, for the high frequency protocol. Try both the high and low frequency versions. Did you try new batteries? Failing/weak batteries have range and recognition problems. Don't focus too much on the arrows, until we get the rest of the commands working properly.
_________________
-Jon

[Vynce]

I tried new batteries and it didn't help.

I just tried the new low frequency frame-adjusted protocol. It works, but much more unreliably than the duty-cycle-adjusted high frequency version. It only works when the remote is about a foot away from the sensor after several presses. I'm fairly sure now that my projector is a high frequency model.

Does the frame length need to be adjusted on the high frequency protocol?

[jon_armstrong]

[Vynce]

The high frequency version works much better than the low frequency version. The power toggle generally works first time, but the other buttons still take several presses before they work.