Philips CD650 mods

Jump to Latest
#121
Thomo said:
The i/v stage i was looking at was fully discrete🙁
Click to expand...

Hi Lee,

At the bottom of the Pedja's page headed "Pedja Rogic Audio Pages are no more," there is a small box labelled "files." If you click on that it will take you to some sort of archive or "index of files." There is a sub heading "tda1541a." Again click on that and it will take you to some schematics, including one labelled "I-V-discrete.pdf"

I'm not sure if this is the i/v stage you were referring to, because I am having trouble reading the text of the pdf file. Seems to be incompatible with my Mac. However, the schematics show up.

The direct link to the "Index of /files/tda1541a" is http://www.pedjarogic.com/files/tda1541a/

Cheers,

Joe
 
#130
SimontY said:
I think the superglued heatsink was not well enough attached (thermally). I will put a 7805 on there and then move up to an S Power if the chip wants one. I'll see how it sounds on a 7805 before making that move.
Click to expand...

I used a LM317T based "RayReg" (from the Marantz thread) on the SAA7220 in my Rotel RCD965BX. I used a PCB vertical mount TOP-3 heatsink with largish fins (4.5 degree C/W) and it still got very hot. I then glued on smaller heatsinks using "Arctic Silver" heatsink epoxy (for computer CPU coolers) until the thing cooled down enough. It now runs at 35 degrees C and seems to be happy at that, although it looks a bit of a mess with all those heatsink bits stuck together.

Joe
 
#132
In a CD85 I upgraded I used a copper 13.2 C/W sink and a ally sink bolted together with compound when using an SPower, this ran @ 38 deg c with the lid on on the same chip, iirc the voltage headroom is greater in the cd85. The SPower ran very hot without the sink and long term I did not fancy leaving it like that.

I was tempted to bolt the reg to the ally chassis but wanted the reg as close to the chip as possible.

Brent

P.S I don't recommend bolting the reg to the CD650 chassis :hot: lol
 
#133
Originally posted by philpoole Its fairly easy to knock up a simple op amp IV - just make sure you don't forget the capacitor in parallel with the resistor in the feedback path - or you might get oscillation and heat and distortion. (but don't worry, it can be simple to do). Also, a servo might be pushing it, but instead, a current source into the IV stage can be used to bias the output so it has zero DC offset, and thus no capacitor required. The DAC output swings from 0A to -4mA (so the output swings from 0v to +6.4v, so if you add 2ma to the input, then the centre point will be output at 0V (so -3.2 to +3.2v). I guess it can drift with temperature (hence a servo would be better), but its probably negligible.
Click to expand...
I think the DAC swings 2mA and draws a constant current of 2mA with can be nulled with a 2mA Constant Current Source like in the Pedja Rogic DAC.
Click to expand...

Hang on, I think that's what I said ?!
 
#135
rowemeister said:
In a CD85 I upgraded I used a copper 13.2 C/W sink and a ally sink bolted together with compound when using an SPower, this ran @ 38 deg c with the lid on
Click to expand...
I have to do a bit more work on getting my SAA7220 to run cooler and there is not much room left in my Rotel for more heatsinking. The ambient temp in Perth today is 30 deg C, and its only just starting to warm up for summer here. 38 - 40 deg C is common.

I am wondering whether a lower supply voltage into a low drop out 3 pin reg will run cooler. The 317 based "Rayreg" I am using is currently supplied with a 12V. Perhaps I can exchange the LM317 for something like LT/LM1086 and use a 7-9V supply?

I was tempted to bolt the reg to the ally chassis but wanted the reg as close to the chip as possible.
Click to expand...
Another alternative is mount a large heatsink on the back panel, but the SAA7220 is near the front panel on my Rotel and so the supply lines from reg to chip would be about 250mm long. Like you, I want to mount the reg close to the chip, so fiddling with the heatsink or the regulator type and supply voltage seems to be the deal, bearing mind ithat I have a tight budget.

Bearing mind ithat I have a tight budget, any other suggestions are welcome.

Joe
 
#136
You should be able to get enough drop from 9v for a 5v regulator. 7v might be a bit low, but 9v is more than enough. I have a 7805 powering my SAA7220, from 9v, and it doesn't have a heatsink. That doesn't necesarilly mean it doesn't need one! But, at least it isn't too hot to touch, and it hasn't cut out.

30 degrees, going up to 38-40! Don't rub it in! 🙂
 
#137
philpoole said:
You should be able to get enough drop from 9v for a 5v regulator. 7v might be a bit low, but 9v is more than enough.
Click to expand...

Hi Phil,
OK. I can try 9V. I already have a spare unregulated 9V rail from the original Rotel transformer (I have installed another 5 to feed the 15 regulators), so I can try that and see if the 317 based reg runs cooler, although it will have to wait until after Christmas now.

philpoole said:
30 degrees, going up to 38-40! Don't rub it in! 🙂 [/B]
Click to expand...
I suppose it sounds good when you are in the snow, but it starts to be too much of a good thing after a whole month of it. Would you believe that our Super 14 Rugby season starts in a few weeks (teams from Australia, South Africa and New Zealand) – right in the early to middle part of summer. They have to put the players in an ice bath at half time to cool down an some need fluid replacement by IV drip. Rugby used to be a winter sport where we froze at half time and spent most of the game sliding around in the mud.

Cheers,

joe
 
#138
9v in, 5v out, 200mA for the saa7220 = 0.8w dissipated. That's OK for a 317 alone (max is about 1W continuous but check the datasheet for any temperature derating) and a small clip-on heatsink will help a lot.

The point when you can only just bear to touch for 3seconds is about 55-60degreesC, that's a good maximum for long term reliability (remember the junctions inside the package run hotter!)
 
#140
That works well, as can simple brute-force RC using a resistor. Try, say 10ohms / 10uF before the reg. TIt makes for a nice filter at 1.6Khz, chopping HF noise down before the reg, and will drop about 2v / disspate 0.4W. Use a 2W metal film resistor.


Oh, you can use more capcitance here too; it's not remotely critical. Using say 100uF pushes the 'knee' down to 160hz, which gives the reg slightly less line regulation to do of course.
 
You must log in or register to reply here.
Top Bottom