Do you think it's possible that after you update or flash the player it is ok for a little while, then one or more files gets corrupted or overwritten or something? Mine was okay for about a week after my ridiculously long post where I documented what I did to "fix" mine. A week later, same crap. I am very angry at my player right now...

Nah, I've flashed mine up and down, and the problem almost seems random. I wonder if this temperature thing is worth looking into? Perhaps a component on the display (or display power) is sensitive to change? I am very angry at 2 players right now...

I just brought mine in from outside. I didn't flash the firmware or anything. I was playing around with the dimmer in "forced dc mode". I was having the problem with the dimmer not working correctly, and my friend started pulling on the AC cord. Now, for some reason, the dimmer is working again. I guess we can hopefully narrow this down to EITHER the temperature causing the problem, or maybe the AC plug, or the sensor that realized the unit is plugged into AC. I heard there is a switch that can tell if the unit has the AC power supply plugged into it. Maybe if this switch is in an intermediate position instead of fully off or on, it is causing problems intermitantly with the dimmer setting when the player if is DC mode...

I don't know...

Temperature could be causing it too...

I JUST DON'T KNOW!!!!!!!!!!!

I'm gonna go on record here officially guessing that the main cause of the dimmer problem is due to the AC adapter plug insertion. I think it keeps reporting either a one or zero to the player, but I'm not sure how often the player polls the switch. If I had to guess, from the way my dimmer screws up all the time, I would guess it polls the switch about every second or so!?!?!? Does this sound like it could cause the problem!?!?!?!

I have an MK2a. Never installed Hijack. Only used in the home on AC. I have seen the problem in 1.03, beta 7, and beta 11. Sometimes it boots up dim, sometimes bright. The only pattern that I think I've seen is that if the player has been running for quite a while, it seems that it is more likely to boot up dim, which seems to support the temperature theory.

Once again, I put it into my truck this morning and it having the same problem again. I guess that blows the temperature theory out of the water. It was in warm house all night and it was warm out today. I thought maybe it was doing it because it was so cold outside. I didn't have time to mess with it today. I really hope someone finds out what is causing it so I know if I should send it in to get it fixed or not...

----------
The dimmer is very eratic. It will start out at 100 and then get dimmer at 90, 80, turn OFF at 70 (black screen), come on a little brighter at 60, full bright at 50, etc. Also, if I set it to 100 and keep turning the knob to the right, the display will change brightness while all the while saying 100%.
------------------------------

Ruffles gave a very good description of this problem...

Am I correct in assuming only about 7 or 8 people have this problem, or is it more widespread then that...???

Does anyone notice that it doesn't happen all the time? My player was fine for about a week, then it happened again. Then I messed around wiggling the ac connector and plugged it back into my truck and it was fine for a day, now it's messed up again. Everytime I think I figured out what it is, reproducing those same steps again to fix it doesn't work.

Add me to the body count. My unit has the same problem. I also believe that it is present on all software revisions back to and including 1.03

i know i've posted about this problem before, but i have the same problem with my units. in my case there is no correlation between temperature and the dimmer problem.

--dan.

I tried the same thing and it did NOT fix my player. I sent it in for service today.

For what it's worth, the player that I just sent back to empeg has a working temp sensor where my original player with out the dimmer problems temp sensor has stopped working.

I have exactly that problem right down to the temperature thing as well as what was described by Rezolution. David (Moss) seems to think it could be software (the dimming) but I am fairly convinced it's hardware.

I have also found that the knob control is very erratic when adjusting.

This sounds stupid, but every day I take it inside to try to figure out what the hell is wrong with it. After I leave it pluged in for like 5 minutes on AC power (and adjust the dimmer about 100 times to see if its working right or not )with it forced to DC mode, the dimmer clears up and starts working, then the next morning it just gets screwed up again. Maybe my car stereo only likes the AC adapter or something.... I tire of this!

You gentlemen will all be very happy to know that I have found the CAUSE OF THIS PROBLEM. I read through the entire thread AGAIN and noticed a few people kept mentioning temperature, so I figured what the hell, and hooked up my empeg inside of one of our thermonics units at work...

Guess what happened...

From 5 degrees C till about 24 degrees C (thanks to mark lords kernal hack - using both his force dc power mode to get the dimmer adjustment menu and his vital signs screen to get the temp) my dimmer does not function correctly. From about 25 degrees C to about 35 degrees C it works just fine. I didn't want to go above 35 for fear of wrecking something...

I will be emailing these results to David Moss tomorrow asking him what might be causing this and what he can do to fix this problem...

I cycled the temperature four times, and each time it reproduced the same results with a variation of about 3 degrees. 23 Degrees C to 26 Degrees C is about the range the dimmer goes from nonfunctional to functional...(on my player anyway.)

I don't know if anyone else (who is having the same dimmer problem I am) could verify this by maybe leaving it in their car overnight in the cold and trying it in the morning, then leaving the heat on full blast for a few minutes and watching mark lords "vital signs" with his kernal hack and see if your dimmer is fixed around 28 degrees C or so...Please don't wreck your EMPEGS though!!!!!

This would explain in part why every time I brought the empeg inside to play with it, I thought that it was fixed...

Right now I have my empeg sitting on top of my computer monitor to heat it up Ill let you know if it works out.

You can also put a piece of paper over the little air vent holes on top to help warm it up, that will recirculate the heat inside. Don't cook your EMPEG okay??? I can't have the death of an EMPEG on my conscience...

Well I wasnt paying attention but when I looked it was at 29 deg and guess what you are right no more dim screen. You rock this is definately hardware sadly.

I'm just sorry I posted sooooooooooo many things that I thought were a fix before I actually found what the real problem was. I feel like I junked up the thread with a bunch of crap. I wonder if someone like Tony emailed my results to David Moss, if it would "carry a little more weight", and get looked at sooner by the tech guys....

Test number 2 in progress

How are you cooling it? Dont tell me in the refrigerator!!!

Also please remember that sometimes you have to adjust the dimmer before it will glitch out, just another thing I noticed...

Ill tell you though, once you get the temperature up around 28, the display really looks beautiful at 100% brightness! Around 22 or below, it just doesnt glow nicely.

Haha no not the refridgerator my roommate likes to have the AC on at night so I just put it in front of the vent. Since Im in AC mode I had to reboot this time to get the brightness to change but it did looks like we have a cause now we need a solution.

I don't have the schematics, or the resources, or the patience to figure out the rest of the problem, thats what the techs at customer service are for. I can tell them, hey look my dimmer doesnt work from 0-24 deg c, but it works from 24-35 deg c. Then they can get their digital multimeters out and their oscilliscopes and stuff and tell me that the tri-molecular discombobulater is broken and its going to cost 8000 dollars to fix. Oh wait, thats my local dodge dealer. The empeg guys will tell me its just a temp sensative capacitor and that they are gonna fix it for free cause its under warrantee and even if it wasnt they would fix it for free cause they are cool like that. I am going to bed!

Night Acura, and thanks for further reinforcing my test results. Please, I beg all of you, don't preheat your EMPEGS so that your dimmers will work, just wait for the tech fix!!!!

*Don't* use the rotary control when in the dimmer menu. The same PIC on the display board deals with buttons, rotary control *and* receiving dimmer commands from the host. It doesn't have any interrupts, and some of this stuff is in tight loops.

Using the buttons to knock the brightness up/down a notch is more reliable.

The rotary control/dimmer interaction is just a design issue, there's nothing we can do about it.

Also, players can boot up dim but then have correct brightness restored as the player runs - this is a hardware issue that I've never been able to track down. It appears to maybe be noise on the dimmer control line before the CPU boots; I have tried sending new dimmer commands earlier in the kernel bootup but I couldn't get any conclusive results.

As far as I can see here, there are 3 categories of problem:

- Boots up dim, but fine when the player runs: known issue with the hardware, not classified as a serious problem as it doesn't affect player operation.

- Rotary control / dimmer interactions: known issue with the hardware. Use the buttons instead.

- Player thinking lights are on/off randomly and setting dimmer to suit. This is probably over-sensitive detection of the glitches on the headlight sense line and we can probably do something about it in the player. Can people suffering from this please describe their setups?

Hugo

I think this is a common problem. Hugo mentioned to me that communication between the player software and the display PIC is 1-wire serial with software timing (i.e. asynchronous), and sometimes the data get slightly out of time during sending. I wonder if instruction cache issues affect the timing loop?

I don't think the player ought to be sending any dimmer commands in home mode, but I do sometimes see reduced brightness after boot-up, so I guess it does.

In reply to:

---

*Don't* use the rotary control when in the dimmer menu. The same PIC on the display board deals with buttons, rotary control *and* receiving dimmer commands from the host. It doesn't have any interrupts, and some of this stuff is in tight loops.

---

Is the remote handled in the same PIC, or is that the best bet for making adjustments?

I don't believe it's temperature related; I think you're just seeing an interaction between the rate of commands being sent to the display board (accurate timing from CPU crystal) and the rate the display board PIC is running at (circa 20MHz, but uses a ceramic oscillator and so is pretty vague).

There may be something we can do about this in the kernel, possibly the bit send times is display_sendcontrol can be tweaked (shorter/longer low time - see line 1015 in empeg_display.c - if anyone has any conclusive results, please let us know and we'll roll the change into the next release).

However, sending it back won't help - there is no fault as such with the display board. Anything we replace it with is just as likely to behave in the same way, as it will have identical circuitry on it.

There have been *genuine* display faults in the past (on mk2s, not mk2a's) where the 60v display PSU becomes unstable due to capacitor breakdown on the feedback line; the symptom of this is that the display is just *very* dim, even at 100% brightness

To check for this if you're electrically savvy, measure the voltage between the case and TP1 on the back of the display board - this should be between 54 and 60v. Note that this involves opening your player and poking around a high voltage line, so don't do it unless you suspect a fault (ie, very dim at 100%) and are confident about measuring such things.

Hugo

The remote is handled by the same PIC, yes, but only on a pass-through basis. The remote is the safest option for dimmer adjustments as it doesn't tie the PIC in a busy loop to send info back to the host.

When we signal to the display PIC, we pull the comms line high for 100ms to let it finish what it's currently doing and sync up with the main board - then we clock in 8 bits. When the PIC is "synced up" (ie, displaycontrol line high) it is doing nothing else - not processing IR, buttons, or rotary events. It's in a busy loop looking for a short (<2us) low or a long (>5us) low of the first bit.

The short low has to be long enough for the PIC to notice it; this shouldn't be an issue really, as the PIC runs at 20MHz (hence 5 mips) and the loop is only a few instructions. Maybe the delay is currently too short in kernel and it misses a bit. If the line stays low for too long (like milliseconds), the PICs watchdog kicks in; this way we can force a pic reboot if we need to.

Ah, this might be the problem actually - looking at the pic code, there should have been a watchdog reset in the code that waits for the first low. Maybe setting the kernel /* Wait 100ms */ delay a bit shorter will help, as I suspect the pic is rebooting several times whilst waiting for the first bit.

Hugo

Hows this for an issue Im in the car. Lights off. Display Dim(may or may not have a bright boot sequence). Go to dimmer control. Shows display is at 100%. Using remote go to 90% gets brighter. Go back to 100%(it may take a couple times back and forth) and its bright. Of course if you are in AC mode you cannot adjust the brightness if it is dim, but I will admit as soon as the temp reaches 25 C the display is fine.

Count me in on the list.....

Hugo,

I've had a bit of play with these numbers but given I don't really know exactly what they are doing, I haven't been real successful.

I have tried the 100ms delay as both shorter (50ms) and longer (166 and 200ms) and the problem only seems to get worse either way. I tried changing the length of the for (a=0;a<50;a++) loop to 25 on line 1016 for a worse result. I also replaced the for loop with the original udelay(1), which seemed to really slow things down.

You say you think it's not temperature related - however if you are only using a ceramic resonator for the PIC then surely that could/would affect the frequency that it's operating at. Mine definitely gets much much worse when it's "warmed up" (40C according to hijack) as compared with cold (about 25C) here in Oz.

If there's anything you think I should try let me know. As I said, I'm guessing big time from what you have recently said so I'm probably wasting my time.