I know it's very common for people to replace the XO in CD/DVD players with fancy low jitter clocks, and I'm just wondering if there're DVD players (which tend to be more complicated) whose startup test would fail if the XO is replaced by an external clock.
I ask because I'm working on a pioneer dvd player (dv-48av). The new clock seems to output the right voltage and frequency, but the player just refuses to start up. It will, however, if I reconnect the old XO back using the two wires I use for the new clock board's output, so there's nothing wrong with the player or the connections. Has anyone had experiences like that?
Thank you!
I ask because I'm working on a pioneer dvd player (dv-48av). The new clock seems to output the right voltage and frequency, but the player just refuses to start up. It will, however, if I reconnect the old XO back using the two wires I use for the new clock board's output, so there's nothing wrong with the player or the connections. Has anyone had experiences like that?
Thank you!
Thanks for your replies!
wakibaki, that's a good question, I don't really know
(I had some reasons but they were bad) I'm corresponding with the manufacturer of the clock, though I'm pretty sure he tested the boards carefully before sending them out.
stephensank, that's a very interesting point. My DV-48AV typically takes quite a few seconds even to just light up (and it won't start up if I remove the cover), so I'm sure there were some processes that start after others. I've attached the schematic of my power supply board, and I'm using pin 3 of CP502 (AT +12AV for MUTE) to power my clock board. Does anyone have any comments? Is there a way to make the power go on sooner?
wakibaki, that's a good question, I don't really know
(I had some reasons but they were bad) I'm corresponding with the manufacturer of the clock, though I'm pretty sure he tested the boards carefully before sending them out. stephensank, that's a very interesting point. My DV-48AV typically takes quite a few seconds even to just light up (and it won't start up if I remove the cover), so I'm sure there were some processes that start after others. I've attached the schematic of my power supply board, and I'm using pin 3 of CP502 (AT +12AV for MUTE) to power my clock board. Does anyone have any comments? Is there a way to make the power go on sooner?
Attachments
The first thing to try is to move the power feed to the other side of that regulator transistor, to the PCON+12V output. This still may not time the start up right.
I have not yet worked on this model(DV46AV is most recent I've experience with), but I do not think it's normal for the unit not to power up with the cover off, so I would look into that, for sure. Unless there is strangely an interlock switch that senses the cover, it can't be normal. Don't off hand have any suggestion as to possible cause of that weirdness, though.
Once you can get it powering up with cover off, assuming the power feed change doesn't make it work, you'll need to reconnect the stock clock, then get at it with an oscilloscope(10x probe) to see exactly when that clock starts up after power up. Then, see if either a)you find a corresponding suitable supply feed starting at same moment, or b)add a resistor & capacitor to the incoming power of the clock to cause it to start up later.
BTW, I don't know why they made the DV46 & later units take so damn long to power up, but I find it incredibly annoying. But I'm most annoyed by the fact that the auto-power-off idle timeout is built into the syscon processor & can't be defeated. Really hate that. Like my stuff to stay powered up.
I have not yet worked on this model(DV46AV is most recent I've experience with), but I do not think it's normal for the unit not to power up with the cover off, so I would look into that, for sure. Unless there is strangely an interlock switch that senses the cover, it can't be normal. Don't off hand have any suggestion as to possible cause of that weirdness, though.
Once you can get it powering up with cover off, assuming the power feed change doesn't make it work, you'll need to reconnect the stock clock, then get at it with an oscilloscope(10x probe) to see exactly when that clock starts up after power up. Then, see if either a)you find a corresponding suitable supply feed starting at same moment, or b)add a resistor & capacitor to the incoming power of the clock to cause it to start up later.
BTW, I don't know why they made the DV46 & later units take so damn long to power up, but I find it incredibly annoying. But I'm most annoyed by the fact that the auto-power-off idle timeout is built into the syscon processor & can't be defeated. Really hate that. Like my stuff to stay powered up.
Thanks stephen, I actually just tried again and now I can power up without the cover, no idea what happened last time...
I've also tried the PCON +12V output to no avail - in fact that's what I used first before moving on to pin 3. The reason I thought pin 3 could be better is because it comes out directly from the transformer T501, just as the AT+3.3V does, and I believe the AT+3.3V powers the factory clock. It sounds like the problem is the clock starts too early rather than too late? That's really weird - what do you think the processor is looking for? Does that mean an externally supplied power would be even more difficult?
Sadly I don't have a scope and can't really afford spend a few hundred on one if I only get to use it once... wish I still have access to academic labs. Perhaps I should try the resistor+capacitor approach. What values do you think I should try? Have you had a similar experience with DV46?
I've also tried the PCON +12V output to no avail - in fact that's what I used first before moving on to pin 3. The reason I thought pin 3 could be better is because it comes out directly from the transformer T501, just as the AT+3.3V does, and I believe the AT+3.3V powers the factory clock. It sounds like the problem is the clock starts too early rather than too late? That's really weird - what do you think the processor is looking for? Does that mean an externally supplied power would be even more difficult?
Sadly I don't have a scope and can't really afford spend a few hundred on one if I only get to use it once... wish I still have access to academic labs. Perhaps I should try the resistor+capacitor approach. What values do you think I should try? Have you had a similar experience with DV46?
Hmm. That makes me think that perhaps the stock clock starts instantly & is always on. Without a scope, the only way to check this is to monitor the supply voltage at the stock crystal, and see if it gets voltage at that instant, and if it stays or cuts out after a moment and then comes back up during power on. Either way, I think you may have to run the clock from an always on supply line, then design in an enable/disable switching fet to turn the clock instantly on and off from the stock clock's enable or supply line.
Indeed, if I simply disconnect the stock crystal (either leaving the circuit open or shorting it) and plug the player into a wall, the hdmi indicator lights up. The machine is then unresponsive. So the problem is not just with syncing the start up, but to make sure that when the power is plugged in, the machine doesn't just enter into some error mode because the stock crystal is absent. Quite unbelievable isn't it?
Sure you are actually connecting the clock generator output to the clock crystal input and not the output as I have done at least once? (And is that input 3.3V or 5V tolerant depending on the supply used? Attenuation needed?) The power supply that runs the IC that contains the original clock oscillator is the one you want to use. (I think) I've reclocked an SACD player, but not a DVD player, FWIW proved to be mostly a waste of time, but works fine to this day which is amazing considering how much I changed the clock architecture. (The particular player had other design issues that resulted in little audible improvement)
Yes, I believe I connected the new clock to the input (on the schematic there's a XTALI, XTALO and a GND). I couldn't use the power supply for the stock crystal because the supply is 3.3V, and my new board requires 10-25V as input in order to output 3.3V. I have to tap into the 12V supply for the opamp (I'm not using the analogue out btw, this is only a transport).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.