docked iPad loses sync

[WK154]

Maybe a series resistor on the power line would work just as well?

Already there as a 166 ohm on ground return, unfortunately it affects all signals that reference that ground.
Don't ask!
 

[RoadRanger]

Maybe a series resistor on the power line would work just as well?

Already there as a 166 ohm on ground return, unfortunately it affects all signals that reference that ground.
Don't ask!

So, when it is charging at 2 amps there's a 332 volt drop across it ? I'm pretty sure that resistor isn't in the charging ground .

[WK154]

Maybe a series resistor on the power line would work just as well?

Already there as a 166 ohm on ground return, unfortunately it affects all signals that reference that ground.
Don't ask!

So, when it is charging at 2 amps there's a 332 volt drop across it ? I'm pretty sure that resistor isn't in the charging ground .

Where do you get 2 amps from?

[RoadRanger]

Apple's iPad charger is rated at 2.1 amps.

[WK154]

Apple's iPad charger is rated at 2.1 amps.

That's Apple but Mackie's puny little 5V supply probably only puts that out for everything 5V in the DL. I've yet to see a spec on the charging current for the DL/iPad. Something I was planning to measure if I get to the cable experiment.
Corrected see report below.

[RoadRanger]

Dunno what you saw but IME the DL1608 does fastcharge an iPad - it takes about the right amount of time to go from half to full. I've slowcharged an iPad off a 1/2 amp laptop port and it takes FOREVER.

[WK154]

When I get a chance I'll revisit that. It may have only been the trickle charge. It's been a while.

[WK154]

OK RR so you made me drag that POJ out of it's peaceful resting place in the closet and put it on the rack again. My memory cells must have had one too many numbers to deal with  but that's what I have you for RR. Here is the last measurements based on the lab supply at 12V. At 100% charged iPad and no other inputs/outputs or Phantom power on the DL draws 1.87 amps @12V. Without the iPad it draws 1.33 amps @ 12V. A charge current of ~ 1.3 amps @ 5V if the converters are 100% efficient which they're not. At 90% efficiency it's 1.17 amps at 5V. Haven't measured the DL charge voltage, left to a future cable test measurement. The ground I was using for the old above numbers was the Firewire ground (29 & 30) which has the 166 ohm resistor and is clearly not the return on the charging circuit.  RR. Who'd a thought that you put the power line on one connector and the return on another? Way to go Mackie, more stellar engineering.  This lead to my confusion. Pin 15 & 16 are the return 23 is the +5V. Need to check again if there's anything on the paddle board to affect this. I let the DL run with only the MF app and nothing else. It looses charge and makes it's way to 80% at which time it fails. The current draw goes from the 1.87 to 1.97 amps (1.53amps @5v) in the process over several hours. A word of caution. All this is using an iPad 3 retina display. iPad 2's have a battery half this capacity and off course will have differing results as will the iPad 4, Mini and Air. If there is any change you can count on from Apple it's another battery and charger, $$. When it fails the charge current drops to 1.55 amps @ 12V. That minus the DL's draw makes it the power available (~500 Milli-amps) from a USB v1.1/2.0 powered port. Apple unfortunately declares this as not charging in the DL's case and Mackie immediately declares it not sync'd. A bad idea at best. Not sure if that's an Apple status generated that Mackie is acting on or their own. In any case Mackie does make the final determination about sync. Battery level at 80% is a far cry from being a problem.
In conclusion there is no fast charge from the DL only one range around 1.2 - 1.5 amps until it looses sync. It then drops to 0.5 amps and not connected. That's it for now.

[Kevin]

Hmm, so sounds like this problem might be related to 5v charging issue that makes using a cable for docking a PIA?

In that case, the workaround should be the same: keep a full charge on the ipad and dont dock it at less than 90%. That has been working for me, knock on wood.

Having to fully charge the iPad before each gig is pretty lame though.

Most likey the noise is ripple on the Mackie's 5v power supply for the iPad. Hopefully the Ipad charging circuit has its own supply. Its probably a standard buck converter, in which case increasing the output capacitor and a better inductor will usualy improve ripple. These parts are generally large enough to replace with a regular soldering iron. The trick is figuring out what the existing values are.

If someone knows a service center tech who can look up the part number for the converter chip, the data sheet probably has some info on chosing componants for minimal ripple.  It would be interesting to see if it is really rated to put out 2A. Anyone know if Mackie has a service manual for the DL1608 or how to get a set of schematics?

Its possible that it might be able to make out the part number from the markings on the chip, but that is pretty hit or miss.

WK154 - Do you have a higher rez version of the 1608 motherboard picture that you posted in your Autopsy FAQ a while back? Could you post that in the FAQ?

[WK154]

Kevin unfortunately the 5V buck converter and all others except for the 3.3V (on main board) were stuffed into the analog section that I never opened. Hence no pics.

[Wynnd]

Congrats to having probably found the sync problem source.  I only had that happen once, but mine is an ipad 2 and I didn't notice the charge level.  Been keeping it fully charged on days I need the mixer.  (I keep an ipad mini wireless at my keyboard.)  One thing about keeping an ipad fully charged.  They will go for days of no use with no noticeable drop in charge level.  I could probably leave my ipad docked.  It would get used at least on rehearsal every week and that would probably be enough to keep it charged. 

Since I've had the ipads, I don't take my laptop out much.  Used the laptop last night to play a DVD for German Movie Night.  (I'm a minor officer in the Edelweiss Preservation Foundation.  We do German Language movies once a month.)  I ran the macbook on battery power with an external DVD drive for about 2.5 hours and still had 32% left on a 4 year old computer.  Apple products continue to amaze me.  I didn't know about the battery life until after I bought this macbook.  I just wanted good firewire for my recording hardware.  Could never get my windows laptop to daisy chain the firewire.  The mac has always worked fine that way.  I haven't tried it yet, but I upgraded my internal drive to a Momentus XT hybrid drive.  (Nice improvement and I needed the extra space.)  I've only done a few channels recording since then.  It might be fast enough to handle 16 channels, but I need to test it.  I'm rambling again.

[Kevin]

Kevin unfortunately the 5V buck converter and all others except for the 3.3V (on main board) were stuffed into the analog section that I never opened. Hence no pics.

How much extra work do you think it would take to get into the analog section? I'm guessing it probably looked like a PIA, or else you would have done it when you did the autopsy.

If the problem turns out to be undersized component(s) in the 5V charging circuit causing excessive ripple at full load, that may be fixable with a few bucks in parts and a few minutes of soldering - but that's once you have the circuit in front of you, which may or may not be a bit of work. If its the switcher itself that is undersized, that might be tough because at the 2A range, a higher rating usually increases the footprint. You can usually get better caps and inductors in the same footprint by spending a little more.

Until someone comes up with a workable mod for the power supply, keeping the iPad charged every gig is probably the way to go. I very rarely go below 92-93 percent by the end of 7 hours max total ontime with Bluetooth off and mostly undocked.  I can see how this would be more of a PIA for someone with lots of back to back gigs and it is lame to be restricted to only using 10 percent of effective battery capacity. I can also see how I might need do this down the road if the battery on my iPad starts to fade.

BTW - off topic - but reading your autopsy post again, I recall there was an unpopulated circuit that you speculated looked like an unused firewire circuit. That could be an indication that Mackie contemplated firewire recording when they designed the DL, and then left it out because they either didnt have the firmware worked out, and/or they figured most people arent going to want to bring another computer to the gig for recording via firewire anyway. The fact that they left the traces on the PCB and just didnt populate the circuit may be a good sign that they at least wanted to keep this as an option going forward. Whether anyone at Mackie is pursuing this option is TBD.

[Kevin]

My iPad is a 3 and I don't know how Mackie implemented Sync determination other than poorly. If data is coming over the pseudo USB lines and surviving the sync disconnect then there is an obvious problem. I can make it fail with<90% charge on the ipad. Large charging current disrupts the frail USB lines or whatever other way they have of determining Sync. This problem existed since day 1.

Always has a full charge, doesn't get undocked, so that's probably not it. Airplane mode makes no difference, for whatever that's worth.

You say you can make it fail? How? I would love to reproduce it at will, give us a chance to solve it.

Is anyone else seeing this? It's complete loss of mixer control. I would think people would be screaming if it was at all common.

For people who are seeing this happen on a fully charged iPad, how old is your battery and how quickly does it drop to 90 percent when you run wireless? If its an older iPad and just started happening, it may be that the battery is worn to the point where it cant stay in trickle charge mode even if its always docked. Once Mackie's iPad charger comes out of trickle charge mode, it seems to get noisy, and this causes comm errors. Each release of MF and IOS seems to add to the likelihood that a comm error gets fumbled and you can blame programmers but to fix it you need to get rid of the noise, which may mean new battery.

edit - if anyone has 2 iPads and is having this problem, you might want to try docking the newer iPad and using the older as the remote and see how that works.

[WK154]

Kevin as I look back on the autopsy and the subsequent problems that arose with the DL I would have approached it quite differently. At the time I was only interested with what was under the hood in terms of CPU's, memory and A/D-D/A's. I had no interest in the rest. Yes it is a PIA to get at the analog section. I have now put my Systems Engineering hat on again which has as a cardinal rule: "Remove any substandard units". I have since installed more XAP800's that could easily have been DL's except for it's potential failure modes. DL's are off my list. I bought the DL to have first hand knowledge before recommending or using it for customers. I occasionally get drawn in to deal with individual unit problems but I have found over the years it doesn't pay. It turns out to be a big waste of time and resources, as in this case with a non-responsive manufacturer. "Live and Learn" most of the time.
In my notes I found two part # which are voltage controllers possibly one on the main board. A LP 3853 and a LTC 3853 but I don't remember which is in use. My guess would be the LP 3853. Others not listed are Ram's ISSI IS 42S16 160D 16Meg x 16bit and a 4 Meg x 16 IS42S16 16400. The ROM is a Spansion S29GL256P90DTFCR2  16Meg x 16 bit (32 MB). Metacom indicated ~3.9 MB usage out of 18MB, again from his mods to the DL. Then there's the 10/100 Ethernet chip a SMCS 8710A. This leads me to believe that the engineer that designed the DL was light on the abilities of the chips he used. There are far more chip capabilities than are being used.

[Kevin]

In my notes I found two part # which are voltage controllers possibly one on the main board. A LP 3853 and a LTC 3853 but I don't remember which is in use. My guess would be the LP 3853.

The LP3853 is probably for the micro, DSP and any other 3.3v parts. The LTC3853 looks like overkill for a charger circuit and is probably for all the 5v analog circuits. Its possible that the iPad charger circuit also uses this, but I think thats unlikely because (a) if this rail gets noisy it would affect the sound quality, and no one has reported that and (b) that would be a rookie move and so far this unit actually looks fairly well thought out to me.

If anyone is bothered enough by this noise issue that they are willing to follow up on WK154's autopsy and figure out what it takes to get the iPad charger circuit, I suspect that will reveal a cheap 2A buck switcher like the AOZ1050PI for the iPad charging circuit. With a little luck the part number will be legible from the package. You can spot the chip because it will probably be next to a bobbin inductor and largish SMD capacitor. Once the switcher part number is known, one can check rated output and see if there might be a beefier drop in replacement.

Metacom indicated ~3.9 MB usage out of 18MB, again from his mods to the DL. Then there's the 10/100 Ethernet chip a SMCS 8710A. This leads me to believe that the engineer that designed the DL was light on the abilities of the chips he used. There are far more chip capabilities than are being used.

That is not so unusual and I would say is more of a sign of an engineer who is planning ahead. Firmware almost always lags behind the hardware, but getting certifications for hardware is expensive. Mackie has CE, C-Tick and DENAN for the DL1608. A smart HW engineer will load up the PCB with every circuit that marketing might need for the next 5 years, certify the whole thing,  depopulate whatever you cant / dont want to support going out of the gate and hopefully have capacity to add hardware and firmware features for the next several years without spinning a new PCB or trigger re-certification requirements. Depopulated Firewire circuit and lots of extra Flash capacity is a sign to me that someone was thinking ahead for ability to do future revs on same HW platform