I just built a mini tda1543 from hifidiy.net. First digital project. Have had good luck with chipamps and crossovers and such.
Anyways, it doesn't work. Power LED comes on. I get 5 volts on the middle pin of the lm317's. About 3.6 volts from pin 3.
On the output I get zero voltage and no sound. tracing back from the output I get zero volts until I get to the "input" side of the last cap before the output. Then voltage goes to 2 volts measured to output ground. Those capacitors are in line, not to ground. I'm afraid I don't have a very good understanding of the circuit. I thought I was pretty careful about getting the polarity right on all the caps.
The TDA1543 chip gets slightly warm. I understand this is normal. The LM317's get a little warmer. By no means hot, probably just over 100 degrees F after it's been on for 10 minutes. The CS814 is ever so slightly above room temp.
I was careful about ESD. I've looked very closely at the board and don't see any shorted solder joints. Already went through and re soldered all the joints anyways. I tried it on 2 known good working optical outputs. Haven't tried the coax yet.
If I had to make a guess I'd say I probably reversed the caps at the end of the output stage. My understanding is the TDA1543 is constant voltage out. Well I get 2 volts out just before the cap.
Any ideas would be appreciated.
Eric
Anyways, it doesn't work. Power LED comes on. I get 5 volts on the middle pin of the lm317's. About 3.6 volts from pin 3.
On the output I get zero voltage and no sound. tracing back from the output I get zero volts until I get to the "input" side of the last cap before the output. Then voltage goes to 2 volts measured to output ground. Those capacitors are in line, not to ground. I'm afraid I don't have a very good understanding of the circuit. I thought I was pretty careful about getting the polarity right on all the caps.
The TDA1543 chip gets slightly warm. I understand this is normal. The LM317's get a little warmer. By no means hot, probably just over 100 degrees F after it's been on for 10 minutes. The CS814 is ever so slightly above room temp.
I was careful about ESD. I've looked very closely at the board and don't see any shorted solder joints. Already went through and re soldered all the joints anyways. I tried it on 2 known good working optical outputs. Haven't tried the coax yet.
If I had to make a guess I'd say I probably reversed the caps at the end of the output stage. My understanding is the TDA1543 is constant voltage out. Well I get 2 volts out just before the cap.
Any ideas would be appreciated.
Eric
I have a suspicion, and it might be just your description, but your power supply seems a little low.
The TDA1543 needs a 5v supply. Are you feeding 5V to the LM317, and getting 3.6V out? If so, that might be the problem.
Cheers,
Phil
The TDA1543 needs a 5v supply. Are you feeding 5V to the LM317, and getting 3.6V out? If so, that might be the problem.
Cheers,
Phil
I'm feeding 12 volts to the 317, which would be pin 1 on the schematic. Pin 2 has 5 volts and feeds the tda1543. pin 3 is the one with less voltage. The kit came with a schematic on the back of the instructions, but it's pretty fuzzy. I'll see if I can get a good scan or pic. Here's the online assembly instructions.
http://us.hifidiy.net/Article.asp?ArticleId=152
More pics of the board.
http://us.hifidiy.net/Article.asp?ArticleId=174
http://us.hifidiy.net/Article.asp?ArticleId=152
More pics of the board.
http://us.hifidiy.net/Article.asp?ArticleId=174
Got my pins mixed up. Pin 3 on the schematic is the middle pin on the 317. Anyways both the cs8414 and tda1543 are getting 5 volts.
I wish I had some better measurement tools. A scope would probably be handy right about now.
I wish I had some better measurement tools. A scope would probably be handy right about now.
Ah right, okay then.
If it were me, I'd probably double check the I2S connections from the CS chip to the 1543, then the input circuitry. Its quite likely its all powered up, and ready for a signal, but it either isn't getting anything, or is failing to sync.
If it were me, I'd probably double check the I2S connections from the CS chip to the 1543, then the input circuitry. Its quite likely its all powered up, and ready for a signal, but it either isn't getting anything, or is failing to sync.
I'm going to pull the TDA5413 so I can measure the serial data connections for continuity. I'm glad I put the 5413 in a socket. I was planning on sampling a few different 5413's just as an experiment. Everything looks good on both ends but who knows.
Anyways I'm home today with my 3.5 and 1.5 year old boys so it's unlikely I'll get much of a chance to do much but stare at it until this evening.
Thanks for your suggestions.
The good news is I've read the spec sheets for both IC's and I've studied other portions of the circuit so I have a much better understanding of what is going on.
Thanks again.
Eric
Anyways I'm home today with my 3.5 and 1.5 year old boys so it's unlikely I'll get much of a chance to do much but stare at it until this evening.
Thanks for your suggestions.
The good news is I've read the spec sheets for both IC's and I've studied other portions of the circuit so I have a much better understanding of what is going on.
Thanks again.
Eric
Well, the connections between the 1543 and 8414 are good. At this point I'm thinking I must of somehow killed either the 8414 or the 1543. The 317's are much warmer today, to the point where it is uncomfortable to hold after 10-15 seconds. The only thing I did was re-solder c10 and c11 because solder didn't wick up to the ground side of the board. Shouldn't have mattered because the holes are through plated anyways.
Found a better schematic on the website.
http://us.hifidiy.net/Article.asp?ArticleId=176#
For a capacitor, the short leg is negative right? Most of them were not labeled other than having a short leg, which is now snipped of course. When looking at the top of an electrolytic, if one side has a colored edge would that be negative? I'm second guessing myself now.
I don't think I'm going to buy a replacement 8414, it would be $21. I think it'll go to the junk drawer to be revisited if I ever get a scope.
Found a better schematic on the website.
http://us.hifidiy.net/Article.asp?ArticleId=176#
For a capacitor, the short leg is negative right? Most of them were not labeled other than having a short leg, which is now snipped of course. When looking at the top of an electrolytic, if one side has a colored edge would that be negative? I'm second guessing myself now.
I don't think I'm going to buy a replacement 8414, it would be $21. I think it'll go to the junk drawer to be revisited if I ever get a scope.
I soldered the power supply leads direct. I don't think that could have happened.
Thanks for the ideas though.
Eric
Thanks for the ideas though.
Eric
I think you'd be unlucky to have cooked it (especially if your Vcc on the chip legs read the correct voltage).
I suspect you've just made a simple error somewhere, perhaps one you'll kick yourself for, but that doesn't matter when you spot it and fix it.
I would double check your circuit against the schematic. You've made sure the power is getting to the chips, now double check ground pins buzz through to ground. Check the discrete parts of the circuit too.
Also, check you don't have a short somewhere. You might have something like DATA and WCLK shorted together, or something completely different, that won't do any real harm, but would screw up the signal between the chips. Perhaps a splash of solder somewhere?
Hopefully, you'll spot it without the scope.
If you have an electrolytic the wrong way round, if probably won't do any harm to a circuit, but after a few minutes or so (depending on voltage, compared to max voltage) it will go bang - so you might have seen them fail by now if they were wrong. Typically the stripe down the side of the cap indicates the negative pin.
Caps that aren't electrolytic don't need to be polarised. They can be connected either way.
Cheers,
Phil
I suspect you've just made a simple error somewhere, perhaps one you'll kick yourself for, but that doesn't matter when you spot it and fix it.
I would double check your circuit against the schematic. You've made sure the power is getting to the chips, now double check ground pins buzz through to ground. Check the discrete parts of the circuit too.
Also, check you don't have a short somewhere. You might have something like DATA and WCLK shorted together, or something completely different, that won't do any real harm, but would screw up the signal between the chips. Perhaps a splash of solder somewhere?
Hopefully, you'll spot it without the scope.
If you have an electrolytic the wrong way round, if probably won't do any harm to a circuit, but after a few minutes or so (depending on voltage, compared to max voltage) it will go bang - so you might have seen them fail by now if they were wrong. Typically the stripe down the side of the cap indicates the negative pin.
Caps that aren't electrolytic don't need to be polarised. They can be connected either way.
Cheers,
Phil
I emailed Lei at hifidiy.net and he gave me a list of what the voltages should be for the chips. Here's the list with my measurements. OK if it measured correct.
Check LM317 output voltage
and
The CS8414 voltage:
PIN Voltage
1 0 V ok
2 5.02V ok
3 5.02V ok
4 5.02V ok
5 5.02V ok
6 5.02V ok
7 5.02V ok
8 0V ok
9 1.678V 0v
10 1.68V 0v
11 2.528V 1.33v
12 2.517V 1.38v
13 0V ok
14 0.1V -.1? definetely negative, but don't have the resolution to measure exactly
15 0.842V 2.2v
16 5.02V ok
17 0V ok
18 0V ok
19 2.433V 5v
20 0.5V 1.5
21 0V ok
22 5.02V ok
23 0V ok
24 5.02V ok
25 5.02V ok
26 0V ok
27 0V 5v
28 5.02V ok
TDA1543 voltage
PIN Voltage
1 2.516V 1.38
2 2.528V 1.33
3 0.5mV - analog meter not enough resolution but it's definetely negative
4 0V ok
5 4.99V ok
6 3.015V 2.5
7 2.152V ok
8 3.017V ok
Not sure what to look for. Looks like possibly a short on the 8414 but I don't see any, even with a magnifying glass. Maybe I'll desolder it and try again.
Eric
Check LM317 output voltage
and
The CS8414 voltage:
PIN Voltage
1 0 V ok
2 5.02V ok
3 5.02V ok
4 5.02V ok
5 5.02V ok
6 5.02V ok
7 5.02V ok
8 0V ok
9 1.678V 0v
10 1.68V 0v
11 2.528V 1.33v
12 2.517V 1.38v
13 0V ok
14 0.1V -.1? definetely negative, but don't have the resolution to measure exactly
15 0.842V 2.2v
16 5.02V ok
17 0V ok
18 0V ok
19 2.433V 5v
20 0.5V 1.5
21 0V ok
22 5.02V ok
23 0V ok
24 5.02V ok
25 5.02V ok
26 0V ok
27 0V 5v
28 5.02V ok
TDA1543 voltage
PIN Voltage
1 2.516V 1.38
2 2.528V 1.33
3 0.5mV - analog meter not enough resolution but it's definetely negative
4 0V ok
5 4.99V ok
6 3.015V 2.5
7 2.152V ok
8 3.017V ok
Not sure what to look for. Looks like possibly a short on the 8414 but I don't see any, even with a magnifying glass. Maybe I'll desolder it and try again.
Eric
It doesn't look that bad.
I assume you have no input.
The input to the CS8414 is 0v, so no signal I guess (I confess I don't have any real experience with the CS chips - my DAC uses I2S input). The digital signals to the TD1543 will be a value between 0 and 5 v each, and slowly variable, because they are high frequency digital signals (imagine a square wave, train of zeroes and ones with exact mark/space ratio giving a voltage of 2.5v), so its hard to tell if they are at fault.
The output is a bit odd. If the CS8414 was sending correct signalling with zero level for both channels, then I'd expect to see 6 and 8 at the same voltage (after IV conversion - iow across the 1k resistor). So, either the output resistors are wrong or imbalanced, or the CS8414 is pumping out bogus data, or its connected up wrongly, or something like that.
I'm not sure its important, it might be a red herring, but the schematic states 8412, and you state 8414. I unfamiliar with them, but it might be possible the pinout between the two chips are different. Perhaps the configuration of the other chip is such that the output format is different to that required by the TD1543? Someone who knows about CS84* chips might be better qualified to comment on that.
I assume you have no input.
The input to the CS8414 is 0v, so no signal I guess (I confess I don't have any real experience with the CS chips - my DAC uses I2S input). The digital signals to the TD1543 will be a value between 0 and 5 v each, and slowly variable, because they are high frequency digital signals (imagine a square wave, train of zeroes and ones with exact mark/space ratio giving a voltage of 2.5v), so its hard to tell if they are at fault.
The output is a bit odd. If the CS8414 was sending correct signalling with zero level for both channels, then I'd expect to see 6 and 8 at the same voltage (after IV conversion - iow across the 1k resistor). So, either the output resistors are wrong or imbalanced, or the CS8414 is pumping out bogus data, or its connected up wrongly, or something like that.
I'm not sure its important, it might be a red herring, but the schematic states 8412, and you state 8414. I unfamiliar with them, but it might be possible the pinout between the two chips are different. Perhaps the configuration of the other chip is such that the output format is different to that required by the TD1543? Someone who knows about CS84* chips might be better qualified to comment on that.
Might be getting somewhere.
First, the schematic does not match the pcb 100%. Pins 9 and 10 are RX+ and RX-. The schematic shows one going to the coax in, and the other going to the optical. What happens on the PCB is both the coax and optical go to a jumper where you can select which input to use. Unfortunately I can't actually see which pin is hooked to the jumper. Probably pin 9.
Second, between the jumper which selects the digital in and pin 9 is a capacitor, C8 on the flawed schematic. I get voltage on the jumper side, but 0 volts on the chip side. That capacitor is 0.iuF Polyester MKT. No polarity on those right?
The chip is in fact 8414, and you are right, the schematic looks like 8412, but the pinout of the 8414 matches the schematic, so it's probably a typo.
Eric
First, the schematic does not match the pcb 100%. Pins 9 and 10 are RX+ and RX-. The schematic shows one going to the coax in, and the other going to the optical. What happens on the PCB is both the coax and optical go to a jumper where you can select which input to use. Unfortunately I can't actually see which pin is hooked to the jumper. Probably pin 9.
Second, between the jumper which selects the digital in and pin 9 is a capacitor, C8 on the flawed schematic. I get voltage on the jumper side, but 0 volts on the chip side. That capacitor is 0.iuF Polyester MKT. No polarity on those right?
The chip is in fact 8414, and you are right, the schematic looks like 8412, but the pinout of the 8414 matches the schematic, so it's probably a typo.
Eric
can't you get a scope and measure the clock and datasignals? That normally solves things in 10 minutes (at least diagnose it 🙄 )
doede
doede
Totally agree on the scope making it easier. I've been keeping an eye out for a good used one.
Eric
Eric
I had a little trouble with mine - I went over the tiny connections on the pcb and found a receiver leg unattached - resoldered all connections and fired right up - are you jumpering the Coax and optical digital inputs correctly?
No progress to report, but I did buy a used Fluke 8050A on ebay. Might be able to pick it up tonight. Also bidding on a couple scopes. Probably end up getting a Tek 2225, but we'll see. I've narrowed my options to local auctions, hate to have stuff like that shipped.
Eric
Eric
- Status
- Not open for further replies.