Hi,
First post so be gentle
I'm in the process of repairing a faulty digital audio board, and the thermal fault finally points hands down to a PCM1700u. Now I know this chip is obsolete, and if this has been covered before, I apologise in advance, but remember 'it's my first posting on here'.
Ok so I can go to Digik*y and buy one today probably, but what is the usual 'ongoing maintenance' replacement for this chip? Fortunately there's some room around the circuit to patch in a daughterboard (the board has several PCM1700u's on it).
I'm very new to ADC and DACs, everything has only ever needed logic levels for me until now, just need some advice to make the right choice to move on this and get some decent sound back out of the box, instead of crackles and pops after it warms up.
TIA
diy-er
First post so be gentle
I'm in the process of repairing a faulty digital audio board, and the thermal fault finally points hands down to a PCM1700u. Now I know this chip is obsolete, and if this has been covered before, I apologise in advance, but remember 'it's my first posting on here'.
Ok so I can go to Digik*y and buy one today probably, but what is the usual 'ongoing maintenance' replacement for this chip? Fortunately there's some room around the circuit to patch in a daughterboard (the board has several PCM1700u's on it).
I'm very new to ADC and DACs, everything has only ever needed logic levels for me until now, just need some advice to make the right choice to move on this and get some decent sound back out of the box, instead of crackles and pops after it warms up.
TIA
diy-er
I wouldn't think there's any particular fail rate problem with
this DAC. May just be an odd failure. I'd replace it and
see if it happens again before doing anything rash. Also,
be sure to do things like check for bad power supply voltages,
wrong components around the DAC, etc.
this DAC. May just be an odd failure. I'd replace it and
see if it happens again before doing anything rash. Also,
be sure to do things like check for bad power supply voltages,
wrong components around the DAC, etc.
Hi Brian,
Good point. The board is virtually a carbon copy of the BB application notes, which is handy.
The problem was one of bad design (manufacturer withheld) a hot linear supply sits in an enclosed metal box, no ventilation, precious little heatsinking, and that PCM1700 was the closest to the heat source. Thermally cycling everything on the board tracked me right down to this one (of four on the board) PCM1700's. Freezer spray on top of the IC alone cures it. Supply voltages are well within tolerance. The thing is I don't think the insides ever got near BBs upper temperature limit, the chips are warm to the touch when operational, but not hot (unlike the 10 x OPA2604AU's nearby, they all run very warm, which doesn't seem normal to me - your thoughts please - supply is +- 13.5V on those, 0v on all inputs and outputs w.r.t analog gnd).
I gather that it wouldn't be so straightforward to replace with one of the AD1865 (?) DACs as they have a seperate LE for left and right data, or is the LE capable of being commoned and the DAC capable of being clocked with left and right channel data simultaneously? Suppose I'm learning something new here, I've always been put right off DACs as being awkward....
I'll probably get a replacement 1700u, but was thinking in terms of when all the obsolete spares dry up.
Thanks for your answer good to know there's someone out there
Adam
Good point. The board is virtually a carbon copy of the BB application notes, which is handy.
The problem was one of bad design (manufacturer withheld) a hot linear supply sits in an enclosed metal box, no ventilation, precious little heatsinking, and that PCM1700 was the closest to the heat source. Thermally cycling everything on the board tracked me right down to this one (of four on the board) PCM1700's. Freezer spray on top of the IC alone cures it. Supply voltages are well within tolerance. The thing is I don't think the insides ever got near BBs upper temperature limit, the chips are warm to the touch when operational, but not hot (unlike the 10 x OPA2604AU's nearby, they all run very warm, which doesn't seem normal to me - your thoughts please - supply is +- 13.5V on those, 0v on all inputs and outputs w.r.t analog gnd).
I gather that it wouldn't be so straightforward to replace with one of the AD1865 (?) DACs as they have a seperate LE for left and right data, or is the LE capable of being commoned and the DAC capable of being clocked with left and right channel data simultaneously? Suppose I'm learning something new here, I've always been put right off DACs as being awkward....
I'll probably get a replacement 1700u, but was thinking in terms of when all the obsolete spares dry up.
Thanks for your answer good to know there's someone out there
Adam
The other thing I forgot to mention was that it's okay when switched on from cold, but when it warms up the crackle is on audio peaks, scoping the DAC output shows the noise on the o/p.
TIA
Adam
TIA
Adam
Curious. If the dac is not hot to the finger, then my best guess
is that it's a marginal unit (for shame!). But it's possible that
you can fix the problem by improving the cooling. Conversely,
if it really is warmer than you think, a new dac chip may have
the same problem.
I'd suggest adding a heat sink, but you really need to keep
external heat away from the dac.
Warm running SMT op-amps don't surprise me. The quiescent
current of an OPA2604 is typically just over 10mA and at +/-13.5V
supplies, you're dissipating over a quarter watt just sitting
there doing nothing. That will give it close to a 25C rise above
ambient per the datasheet. One reason I don't like to use
most audio op-amp in SOIC packages -- they run too warm.
Unfortunately, market forces are pushing us that way.
I don't think you have a really big temperature issue with the
op-amps, though you could consider little heat sinks. I think
places like Digi-Key might have some (super glue
application?) or you can make your own.
BTW, Digi-Key does not show any in stock.
is that it's a marginal unit (for shame!). But it's possible that
you can fix the problem by improving the cooling. Conversely,
if it really is warmer than you think, a new dac chip may have
the same problem.
I'd suggest adding a heat sink, but you really need to keep
external heat away from the dac.
Warm running SMT op-amps don't surprise me. The quiescent
current of an OPA2604 is typically just over 10mA and at +/-13.5V
supplies, you're dissipating over a quarter watt just sitting
there doing nothing. That will give it close to a 25C rise above
ambient per the datasheet. One reason I don't like to use
most audio op-amp in SOIC packages -- they run too warm.
Unfortunately, market forces are pushing us that way.
I don't think you have a really big temperature issue with the
op-amps, though you could consider little heat sinks. I think
places like Digi-Key might have some (super glue
application?) or you can make your own.
BTW, Digi-Key does not show any in stock.
Thanks Brian,
I'm looking into the PSU area as well, one thing I did notice is that the supply requirement is quite tight. The regulator on the PS is an adjustable type. I may have overlooked that the supply may have been ever so slightly out of tolerance, so an update to the PSU (maybe even a fine adjust might cure that problem) - I'll know a bit more later when I check the thing over accurately.
Expecting a metal chassis to conduct heat in a closed-no-ventilation system seems to be a major shortcoming with some audio gear. Even when a substantial aluminium bracket is used as a 'transducer' to the case there's not a lot a chassis can substitute for proper heatsinking. It's a shame that so much effort goes into a sound electronic design, to be let down by a poor physical approach on packaging. Ostrich syndrome slightly in this case.
As an aside I have a small audio amp on the bench using two TDA2050's in for repair. The first one which has a heatsink attached, is fine; the other which missed the heatsink and relies on the casing conducting the heat into the nearby heatsink ran away with itself and catastrophically (blew to bits) failed. 20mm more heatsink would have removed that problem entirely. The worrying thing was that this equipment was so called 'broadcast' equipment and even the heatsink isn't bolted to chassis, but glued
Perhaps some manufacturers assume we work in the Arctic, and don't need any seriously-considered cooling
Ok, I can expect the SMD Opamps to get warm, that's good enough. Sweating equipment isn't a good thing in my book. Overworked PS's is a recipe for disaster. Maybe I've caught this one from an early demise. I'll have to implement some measures to get the heat out of the box.
I spoke with d*gikey last night, as you say they're out, they recommend the AD1865 as a replacement, not pin for pin, but with some work it may be possible. That goes 'on hold' until I work out whether the PS is out of tolerance.
Thanks Brian, good to have someone to mull this over with
All the best
Adam
I'm looking into the PSU area as well, one thing I did notice is that the supply requirement is quite tight. The regulator on the PS is an adjustable type. I may have overlooked that the supply may have been ever so slightly out of tolerance, so an update to the PSU (maybe even a fine adjust might cure that problem) - I'll know a bit more later when I check the thing over accurately.
Expecting a metal chassis to conduct heat in a closed-no-ventilation system seems to be a major shortcoming with some audio gear. Even when a substantial aluminium bracket is used as a 'transducer' to the case there's not a lot a chassis can substitute for proper heatsinking. It's a shame that so much effort goes into a sound electronic design, to be let down by a poor physical approach on packaging. Ostrich syndrome slightly in this case.
As an aside I have a small audio amp on the bench using two TDA2050's in for repair. The first one which has a heatsink attached, is fine; the other which missed the heatsink and relies on the casing conducting the heat into the nearby heatsink ran away with itself and catastrophically (blew to bits) failed. 20mm more heatsink would have removed that problem entirely. The worrying thing was that this equipment was so called 'broadcast' equipment and even the heatsink isn't bolted to chassis, but glued
Perhaps some manufacturers assume we work in the Arctic, and don't need any seriously-considered cooling Ok, I can expect the SMD Opamps to get warm, that's good enough. Sweating equipment isn't a good thing in my book. Overworked PS's is a recipe for disaster. Maybe I've caught this one from an early demise. I'll have to implement some measures to get the heat out of the box.
I spoke with d*gikey last night, as you say they're out, they recommend the AD1865 as a replacement, not pin for pin, but with some work it may be possible. That goes 'on hold' until I work out whether the PS is out of tolerance.
Thanks Brian, good to have someone to mull this over with
All the best
Adam
No it wasn't the PSU, it's definitely the IC 🙁
Oh well, now to locate a source for these. If anyone has a couple of these lurking, drop me a direct mail or message.
Thanks Brian, back to the drawing board I guess
Cheers
Adam
Oh well, now to locate a source for these. If anyone has a couple of these lurking, drop me a direct mail or message.
Thanks Brian, back to the drawing board I guess
Cheers
Adam
If heat is the suspect you should use “freeze spray”.
Allow internal temperature to rise to the point where it affects the sound (crackling noises). Spray A BIT of this freeze spray on to ONE IC AT THE TIME. If heat is the problem, you’ll find the faulty IC immediately using this method.
Nick
Allow internal temperature to rise to the point where it affects the sound (crackling noises). Spray A BIT of this freeze spray on to ONE IC AT THE TIME. If heat is the problem, you’ll find the faulty IC immediately using this method.
Nick
Thanks Nick, this is what I did, thermal cycling usually flushes them out.
There's four on the board, all four show an extent of thermal sensitivity, some worse than others. The four run a three-channel-and-main stereo output and send-return loop. You can select which PCM1700 does the work from the front panel (all outputs are buffered and summed at the main output with ten OPA2604's. It's a textbook design straight off the datasheets.
The good thing is that I can miss all that out and make a new audio stage, as the Digital Bus is all nicely labelled, LEDAC0..3, CL, SERDL, SERDR, if it's absolutely necessary, even has the print on the PCB for a IDC Header. If I can't get the originals (and I'm still looking without any success) I'll probably have to dead bug the replacements and patch them in.
Cheers
Adam
There's four on the board, all four show an extent of thermal sensitivity, some worse than others. The four run a three-channel-and-main stereo output and send-return loop. You can select which PCM1700 does the work from the front panel (all outputs are buffered and summed at the main output with ten OPA2604's. It's a textbook design straight off the datasheets.
The good thing is that I can miss all that out and make a new audio stage, as the Digital Bus is all nicely labelled, LEDAC0..3, CL, SERDL, SERDR, if it's absolutely necessary, even has the print on the PCB for a IDC Header. If I can't get the originals (and I'm still looking without any success) I'll probably have to dead bug the replacements and patch them in.
Cheers
Adam
- Status
- Not open for further replies.