Hi a few Q's for empeg,

Myself and others had problems with the power off thuds on the MK1. Has this been rectified on the MK2? I was wondering if I need to remove my de-thump circuit when I install this weekend?

I was also curious, what subtle changes are there between the MK1 and MK2, I've not done a comparison of the specs. (apart from the obvious - VR and Ethernet), i.e. why does it boot faster etc. Also, I'm sure this is in the spec., is the Ethernet 10 or 10/100? When also do I need to start sourcing a in-car microphone and what type (i.e. when is the VR being released)?

Another Q.. I assume it's faster to download from HD to USB, rather than cloning (USB-USB), I assume the bandwidth is shared using two isochronous transfers doing a clone, rather that allocating all bandwidth for a standard upload?.


Regards,


_________________________________________
John, (S/No 0114, 20G).

I installed my Mk2 in the car yesterday, and when you turn the ignition off, there is a second or so before the empeg clicks off.

I assume this does what the de-thump circuit does. (I never had a problem with it, as I use a second head unit to control the amps)

The ethernet is 10 meg only.

I'll pass on the other Q's (Although on my machine, a 4 gig clone took 11 hours)

Jazz
(List 112, Mk2 12 gig #40. Mk1 for sale 4 gig #30, apply within)

Hi,

In reply to:

---

I installed my Mk2 in the car yesterday, and when you turn the ignition off, there is a second or so before the empeg clicks off.

I assume this does what the de-thump circuit does. (I never had a problem with it, as I use a second head unit to control the amps)

The ethernet is 10 meg only.

I'll pass on the other Q's (Although on my machine, a 4 gig clone took 11 hours)

---

Sounds like that'a stop the thump on my installation. I thought it might have been 10, as I remember somewhere that the empeg's StrongARM doesn't use DMA which I guess you'd need for 100?

Regards,


_________________________________________
John, (S/No 0114, 20G).

If you want to clone the structure, the clone application is the least hassle way of doing it as it's just fire & forget. The USB is bulk, not isochronous - iso is used only for things like softmodems, speakers, etc.

The permanent feed means that there is no need for a de-thump circuit. One thing you may come across is that there's no ground on the serial connector due to a slight oversight (oops) - and as the cables are custom moulded, we can't change this. Just make sure your attached device uses car ground and there will be no problems.

Ethernet is only 10mbit as the SA1100 doesn't have external DMA and even though there are 100mbit chips which will run DMA-less, it's really not worth the hassle.

Hugo

Hi,

The permanent feed means that there is no need for a de-thump circuit. One thing you may come across is that there's no ground on the serial connector due to a slight oversight (oops) - and as the cables are custom moulded, we can't change this. Just make sure your attached device uses car ground and there will be no problems.

Are you saying that the ground pin (5 from memory) on the DB'9 isn't terminated at the connector? If that's so, that's going to kill my RM-XS2 interface box which only connects to this cable and expects TX, RX, 12V & GND to be there! I have made no provision for externally flying ground lead. I'm off to get the AVO and check before I go off on a tangent. If so, what are your plans to recitfy the error?

Regards,



_________________________________________
John, (MK1 114-20G, MK2 15-36G).

Currently, the only plan is to make up some 9w male-female connectors with a flying ground wire for those people who are affected.

The mk2 also has no power on the 9 way connector, but we did add flow control (CTS/DCD) which are shared with the cellphone mute & lights on sense.

This means that external devices need to have their own power feed/fusing.

Hugo

You wrote: "... we did add flow control (CTS/DCD) which are shared with the cellphone mute & lights on sense."

I hope you mean RTS/CTS? DCD (Data Carrier Detect) doesn't sound too useful for a non-modem device like the empeg. Or am I wrong here?

And that "is shared by" confuses me a little. What exactly do you mean? On my car, the 'lights on' sense is a very low impedance line (simply because it's directly connected to the headlights), so you cannot use it as an in- or output for anything else. And the telephone mute is directly connected to two other telephone mute lines (from my head unit and from a separate tuner/amplifier), so if you use that for anything else, these units will get really confused...

Daniel

No, this is CTS/DCD. Both are inputs, and so will not affect anything else wired to them. It's "shared by" as if you're going to use it for flow control, you don't get it for lights on/cellphone (& vice versa).

We didn't have the pins on the docking connector (and 24 way is the biggest one that (a) we could find and (b) could be worked with by the harness manufacturers) to have full serial docked - the first priority would have been for docking ethernet, really.

The CTS line is there to allow a remote device to throttle the empeg's data flow. There's no RTS, but this shouldn't be a problem as the empeg generally doesn't get to the point of not being able to buffer data :)

Hugo

Sorry to bother you again, but I still don't really get it. Are you saying that I can *physically* connect only either a serial device or lights on/cellphone sensor lines?
My understanding now is: if 'lights on' is connected, and a serial device is trying to drive that line to the opposite voltage of what is fed by the 'lights on' function, the serial device will actually drive against 0 ohm impedance. Serial devices with unprotected output will get toasted by this - and you may be in for trouble...
Talking about the other line - if I connect the cellphone sensor line, and a serial device drives this line low, all my other units will go mute. The empeg may know the difference between an actual cellphone call and a serial transmission, but certainly my head unit display and my tuner/amplifier will not, and will start stuttering to the beat of the serial transmission's handshake...

The second scenario will not cause any permanent damage and probably affect only very few people (those with more than one unit that has a cellphone sensor line), but the first looks really dangerous to me! So unless I'm totally wrong (but I can't see where) you should print a warning somewhere saying that connecting both a serial device and the lights on sensor may damage the serial device. To avoid this you could solder a resistor (1k or something like that) into the lights on line before it actually merges with the serial line, so that the serial device will not try to drive against 0 ohm.

Daniel

The serial port has a female 9-way (ie, the opposite of a PC) on purpose: this is to stop people plugging standard serial devices into it.

Think of it less as a generic serial port and more as an expansion connector.

Hugo

Then may I respectfully submit that http://www.empeg.com/products/product_car_specs.html has errorneous data on it. (claiming that it has 9-pin PC serial port, data rate to 230Kbits/s, also docked in-car, since what you get, while docked is NOT a 9-pin PC serial port)

Heck, I don't think you can even APPROACH 230kpbs on the docked serial port, can you? I thought I read somewhere that it was severely limited (4800?)

D

It's not limited to 4800, it _Defaults_ to 4800 when in car...

One entry in the config ini file, and it's back to 115200 or whatever...

Jazz
(List 112, Mk2 12 gig #40. Mk1 for sale 4 gig #30, apply within)

I dunno... if it doesn't have bi-directional flow-control, it can't do FIFO on its input, and I just have a hard time accepting that it can take 230Kbps input with NO flow control whatsoever.

Besides, my criticism goes to the statement that it's a PC-standard 9-pin serial port when docked, which it most certainly is NOT (as it is missing some of the lines). altman (I think) just admitted "this is not a serial port", so my contention is that it shouldn't be advertised as one. I suspect they'll agree and remove the "also when docked" portion which brings it right back into line with reality.

It's a pefectly functional serial port when docked - a serial port doesn't need to have any control lines to still be a serial port, it just needs transmit, receive, and ground (which is missing unfortunately - see above!).

The port does have a hardware FIFO and we've had luck with 230k operation. IRQ latency on the ARM is generally better than that on PCs.

As with PCs and 16550 FIFOs, the flow control lines are software driven - you have to respond to IRQs and twiddle the flow control manually (by which point you could have emptied the fifo anyway). Hardware flow control on a PC with a 16550 is useless, really - all it can do is throttle the remote end if the software receive buffer is getting full.

UARTs with *actual* hardware flow control (eg, the 6850) are rare - the Startech 16650 does it though, as does the TI 16750 (though in a different way, just to be annoying).

Hugo

Hi,

It's a pefectly functional serial port when docked - a serial port doesn't need to have any control lines to still be a serial port, it just needs transmit, receive, and ground (which is missing unfortunately - see above!).

What get's up my goat, is that fact that empeg have changed the specification to the serial port (which is their right I guess) from the MK1 to the MK2, when they know people were developing applications for the empeg. I personally have 150 UKP invested in PCB's for a Sony RM-X2S interface board, I've got 50 made up and was waiting to test on the MK2 before releasing the 1st batch. This will now require an additonal cable to support the missing Ground & 12V lines. I know I haven't made a big thing of the interface, it's been something I've been playing around with when I get the odd moment, but it was neat that it just needed to be screwed onto the empeg's serial lead and away it went!

For information: -
Completed a 25G download session onto the MK2, it took 26 hours without a hitch using USB.

Regards,



_________________________________________
John, (MK1 114-20G, MK2 15-36G).

Hi,

Currently, the only plan is to make up some 9w male-female connectors with a flying ground wire for those people who are affected.

The mk2 also has no power on the 9 way connector, but we did add flow control (CTS/DCD) which are shared with the cellphone mute & lights on sense.

This means that external devices need to have their own power feed/fusing.

So, is this required for all MK2 sleds or is this only for the 1st batch? In other words, are you going to ammend the pinout of this connector through it's production life? As a possible developer, it's quite important to me...

Regards,



_________________________________________
John, (MK1 114-20G, MK2 15-36G).

Unfortunately, we can't amend it for the first 4000 units as the harnesses have already been made.

The easiest way to fix it would be power/ground wires coming out of your box/the cable used between your box and the empeg: I know it's not ideal, but the amount it would cost to split the harnesses, get new moulded cables, and re-do it all is well into 5 figures :(

Hugo

I'm sorry your board relied upon the physical connections - removal of power was intended, removal of ground was a mistake. As the people making the harnesses are car audio specialists, they didn't see anything wrong with this funny 9-way connector which didn't have a ground on it :(

The only good point is that a 9w M-F lead can easily supply the power to the sony interface as well as ground. I'm sure you'll have no problems selling the units even with the extra wires, though - as I remember, we were going to link your site from the empeg one when you were ready?

Hugo

As far as I remember, there never was a voltage feed on the PC 9-pin DTE serial connector anyway!

John, you could supply your kit with three of those blue IDC splice connectors from 3M (Scotchlok) and by dint of a bit of careful instruction, get buyers to splice into the feed and ground lines of the sled harness, then also splice pin 7 wire of the serial connector into a suitable signal ground. That way you'd only need two flying wires and three Scotchloks extra per kit. Not as neat, I grant you, but I have had just this sort of effort from one of the largest Mini parts suppliers in the UK, so you would be in good company

Hugo, surely, if the pin 7 wire in the serial connector is actually still in there, then you could pre-splice this to a ground line with the same Scotchlok splicers before despatch? This would mean someone sitting down with a pair of pliers and 4,000 harnesses and Scotchloks, and just plodding through them. If you pay for my lunch I'll pop over from NL and do it for you this weekend if you like - OK?

One of the few remaining Mk1 owners... #00015

Unfortunately there is no pin 5 (it's pin 5 on a 9 way, 7 on a 25 way) in the cable. It only has 4 cores, according to our suppliers (though to be honest I've not tried ripping one apart to check... good point)

Hugo

Hi,

The only good point is that a 9w M-F lead can easily supply the power to the sony interface as well as ground. I'm sure you'll have no problems selling the units even with the extra wires, though - as I remember, we were going to link your site from the empeg one when you were ready?

If only we lived in an ideal world... As you say, not a major set-back, the question still stands; will the wiring of the harness be corrected sometime in the future or indeed changed again? You've supplied some details about the additional changes that's been made to the serial connector, any chance of detailed information sheet I and others could refer to?

Regards,

_________________________________________
John, (MK1 114-20G, MK2 15-36G).

It'll be corrected after the first 4000 units.

The corrections will be to fit power at pin 4 (same as mk1) and ground at pin 5.

DCD (cellphone mute) has been fitted to pin 1 and CTS (lights on sense) to pin 8.

Obviously, if you add power/ground to the connector in the same places, then when it is attached to the 4000+ batch connectors it'll still work just fine.

Hugo

Would Empeg be willing to offer a 'swap' of the harness (it's only two screws on the back of the sled), to those pre-4000 buyers who find it a requirement for their peripherals?

(hoping)



Murray 06000047
____________________

No, but we will make up a converter lead if necessary (male-female) which fits onto the end of the harness and has flying ground/power leads and a couple of scotchblok connectors to attach them to the main harness.

Hugo

Hi,

It'll be corrected after the first 4000 units.

The corrections will be to fit power at pin 4 (same as mk1) and ground at pin 5.

DCD (cellphone mute) has been fitted to pin 1 and CTS (lights on sense) to pin 8.

Obviously, if you add power/ground to the connector in the same places, then when it is attached to the 4000+ batch connectors it'll still work just fine.

Sounds fair to me, Okay - I'll ammend my design details to reflect the two new cables on the D'Type: -

Pin 1 - Input - (DCD) Cellphone mute, Active Low, MK2 Only
Pin 2 - Input - RX
Pin 3 - Output - TX
Pin 4 - Power - (DTR) VBat +12V
Pin 5 - Ground - GND
Pin 8 - Input - (CTS) Lights Sence, Active High, MK2 Only

Next question, will the correction cable that you've offered to send come with the carry case, ISO cable and coloured panel that Rob said will be dispatched this week?

Regards,

_________________________________________
John, (MK1 114-20G, MK2 15-36G).

No, we don't have any correction cables. These will get sorted at some point in the future, we simply havn't had a chance to do anything yet. I suspect they'll be hand-made for people asking for them at first (ie, I'll be soldering them...)

Hugo