One assumes the rectifier diodes on the left hand side of the board are correctly oriented. Also what use does rhe empty space for another blue connector block at top right of pcb serve. The middle screw on the block that is present where does this and the other two connect.
It would be useful if you would post a photo of the underside of the pcb with the same orientation as the component side.
It would be useful if you would post a photo of the underside of the pcb with the same orientation as the component side.
Excellent 👍 it will be interesting to see what is going here.
If you have removed the caps then I would suggest you first of all:
1/ Measure the voltage across the empty pads with the DAC 'freestanding' and not connected to the following equipment. If there is any DC voltage present then note the polarity i.e. which way around the meter leads need to be to show the voltage as positive and also of course the value of voltage.
The expected result without seeing the circuit would be a minimal DC voltage of perhaps just a few millivolts.
2/ With the caps still removed now couple up to the equipment you have been using and repeat the test. The voltage readings should really be the same in this configuration i.e. no significant DC voltage present.
One further thought. You mention in post #1 using different opamps.
Have you confirmed the problem is still present with the original parts fitted?
Are you sure the opamps are suitable and that you have not accidently fitted a 'single' for a 'dual' or vice versa as that could well cause something like this. If something like that had occurred the DAC would not function at all though (no audio output from the opamps).
May I ask; why is the middle of the board obscured with a black blob? No problems if you don't want to say.
I tried but they have communication issues 🙂 They are sales people who are not interested in anything other than selling so when I ask anything technical they would either ignore me or say something completely unrelated like "sorry my friend it is out of stock".
I am cheap so I made a mistake of ordering a multimeter through a Chinese site so it will take a couple of weeks but I am also very curious about this and I'm eager to learn electronics (in a forced way I never imagined - but I guess it's a great way to learn heh). I'll report back for sure!
This was my bad - I was trying to draw circles around the capacitors, and as I was struggling to do this, I made those dark spots by mistake and saved the pic without realizing...
BTW if anyone is interested here is the flip side - the top two red pairs are for the capacitors and the bottom pair of 8 dots are for op-amps. The center green dots are for the terminal block which I want to remove (to make space) as it does not connect to anything - hope it is OK !
p.s. Contrary my original impression - it seems like the left capacitor is driving both op-amps and the right capacitor path is leading to the R/L RCA outputs? (3rd port is for COAX input)
p.s. Contrary my original impression - it seems like the left capacitor is driving both op-amps and the right capacitor path is leading to the R/L RCA outputs? (3rd port is for COAX input)
Last edited:
You're very lucky, they normally go bang and emit foul smelling mix of steam and dirty Al foil, its a good idea to wear eye-protection commissioning an unknown board (and if you have access to a thermal camera, use it, can catch issues like this immediately).
It is a great way to learn.
Not always easy just working from pictures but it looks pretty standard, the cap seems to go from pins 6 and 7 of the opamp (non inverting input and the output pin) via a series resistor and a cap to ground. The resistor and small ceramic type cap provide some measure of filtering high frequency noise.
So it seems to look like this, a classic 'voltage follower or buffer with a voltage gain of '1'
pin 4 -15V pin 8 +15V presumably. There has to be some IC output dc of incorrect polarity charging C1 for this to fail not immediately but by erosion over time.
I did see a couple of IC voltage regulators in post 1 image. Is one or other of these faulty or not correctly set up - e.g. resistors with odd value band colors.
I did see a couple of IC voltage regulators in post 1 image. Is one or other of these faulty or not correctly set up - e.g. resistors with odd value band colors.
That would not explain correct operation and good sound for two weeks, before capacitor bulge became apparent. It would also not explain why doubling the WVDC rating of the capacitors (25V ---> 50V), did nothing to prevent bulging. Same bad result with 50V as with 25V capacitors. Explain that.
That's correct. The sound is beautiful that's why I insist on Aliexpress DACs with 90's multibit chips no matter what others say... I just added an additional picture in my original post to reveal accidentally hidden spots. Once again nevermind the incorrectly attached capacitor please .... 🙂
n
If C1 is an aluminium electro type then according internet sources, for all examples it should be possible for them to withstand a reverse polarity of up to minus 0.5 volts dc - which is well within the operating voltage output range of a 5532 op amp.
With ac wave forms this could be for very brief time periods. Conceivably this may have sustained performance within limits for a week or so as you have observed.
If the reverse voltage reaches minus 1 or 1.5 Volts then it appears things come unstuck as has been witnessed.
When Matts19 receives his multimeter he will be in a position to measure the dc offset of the op.amps
There are pads on the pcb topside which allow components to be soldered so the leads projecting to the bottom side of the pcb are mechanically restrained. These should be soldered before moving to the underside of the pcb.
This looks Ok from a distance but closer scrutiny reveals a less glossy situation and I think this needs a going over.
With this in mind I think the purchase of a temperature controlled iron or work station with a selection of end tips, and a solder pump would be valuable tool investments for this and future projects.
Op.amp negative dc output on the positive connection to C1 as depicted in post 29 - this being due to bad joint in the positive supply connection to pin 8 of the IC reducing the voltage at than point and changing the output dc level.
If C1 is an aluminium electro type then according internet sources, for all examples it should be possible for them to withstand a reverse polarity of up to minus 0.5 volts dc - which is well within the operating voltage output range of a 5532 op amp.
With ac wave forms this could be for very brief time periods. Conceivably this may have sustained performance within limits for a week or so as you have observed.
If the reverse voltage reaches minus 1 or 1.5 Volts then it appears things come unstuck as has been witnessed.
When Matts19 receives his multimeter he will be in a position to measure the dc offset of the op.amps
There are pads on the pcb topside which allow components to be soldered so the leads projecting to the bottom side of the pcb are mechanically restrained. These should be soldered before moving to the underside of the pcb.
This looks Ok from a distance but closer scrutiny reveals a less glossy situation and I think this needs a going over.
With this in mind I think the purchase of a temperature controlled iron or work station with a selection of end tips, and a solder pump would be valuable tool investments for this and future projects.
The multimeter arrived..!! And the reading on BOTH capacitors were about -3.1V with DAC powered on ! I did not measure while the music was playing though, and I could not ground the board to chassis while measuring. Note: I measured the 470uf/50V capacitors which I had soldered in, not the empty pads.
I made sure I got the red/black probes right, and even tested with the 1.5V battery to make sure - with the reading coming out at +1.2 or +1.3V.
So, the PCB must have the misprint..!? I am surprised that the music had been playing beautifully and the op-amps, DAC chip etc. seemed to be fully functional without burning out, etc.
I guess this means I don't really need to go out there and buy bipolar caps, and just go back to the original spec of 470uf/25V and I should be set, correct...? I just need to solder it "backwards"..... ?
I made sure I got the red/black probes right, and even tested with the 1.5V battery to make sure - with the reading coming out at +1.2 or +1.3V.
So, the PCB must have the misprint..!? I am surprised that the music had been playing beautifully and the op-amps, DAC chip etc. seemed to be fully functional without burning out, etc.
I guess this means I don't really need to go out there and buy bipolar caps, and just go back to the original spec of 470uf/25V and I should be set, correct...? I just need to solder it "backwards"..... ?
Nothing will burn out in this instance. The capacitors are acting as coupling caps in this application. The DAC and I/V stage look to produce a DC offset of -3.1V and this requires a coupling capacitor on the output to block it. If installed with the wrong polarity the coupling cap will not be happy but for the cap to explode it needs to get hot and for that to happen current needs to flow.
First and foremost there's a hard current limit on how much can flow and this is dictated by the 5532s output and this will be short circuit limited to a few tens of mA. This places a hard limit on how much power could be fed into the capacitor and thus any internal heating caused. Also after the capacitor are series connected resistors that will also act to current limit. I am not familiar with how electrolytic capacitors fail, when connected with reverse polarity, but I'm assuming some sort of leakage current starts to flow through them and their DC resistance falls. My guess is the amplifier you're using is AC coupled, as most are, and this will protect the amplifier from any DC present on its inputs but will also prevent any DC current flowing through the input silicon and input impedance setting resistor. Therefore very little current can actually flow through the reverse biased cap and any failure would be inherently self limiting.
Looking at the PCB again it would appear that there are resistors installed already from the RCA signal pins to ground. This is going to be where a small amount of current can flow if the reverse bias coupling caps are leaky. I'm assuming that these resistors are there to keep the outputs of the DAC very close to ground potential, as they could otherwise float up given the DC biased outputs with nothing connected to the output of the DAC.
It's terrible that the DAC has been produced with these caps around the wrong way 😭
First and foremost there's a hard current limit on how much can flow and this is dictated by the 5532s output and this will be short circuit limited to a few tens of mA. This places a hard limit on how much power could be fed into the capacitor and thus any internal heating caused. Also after the capacitor are series connected resistors that will also act to current limit. I am not familiar with how electrolytic capacitors fail, when connected with reverse polarity, but I'm assuming some sort of leakage current starts to flow through them and their DC resistance falls. My guess is the amplifier you're using is AC coupled, as most are, and this will protect the amplifier from any DC present on its inputs but will also prevent any DC current flowing through the input silicon and input impedance setting resistor. Therefore very little current can actually flow through the reverse biased cap and any failure would be inherently self limiting.
Looking at the PCB again it would appear that there are resistors installed already from the RCA signal pins to ground. This is going to be where a small amount of current can flow if the reverse bias coupling caps are leaky. I'm assuming that these resistors are there to keep the outputs of the DAC very close to ground potential, as they could otherwise float up given the DC biased outputs with nothing connected to the output of the DAC.
It's terrible that the DAC has been produced with these caps around the wrong way 😭
Well there you go 🙂
I think all has been said above. New caps and fit them to suit the actual measured voltage and you should be good to go.