Troubleshooting Hardware Implementation: No Sound from Speaker

[tnecniv]

I have done the hardware implementation using the planned schematic below. but somehow no sound seem to be able to play from the speaker except some noises. Anyone can help me check if something is wrong with DAC shematic?

https://lh5.googleusercontent.com/-D4ko8ebTROY/TrexzepM2iI/AAAAAAAADRI/L76UwiuzJoo/s800/DAC.jpg

thank you

 

[MATLABdude]

Please do not cross-post to multiple sub-forums (especially when both are usually frequented by the same people!)

Unfortunately, your schematic is at or below the limits of readability--can you post a bigger version? If you're on a Windows machine, you can take a screenshot of a single frame using ALT+PrintScreen. There's also usually a CAD processor that can spit out a PDF or an image or some such. If you have something like PDF Creator or Adobe Acrobat, you can usually print directly to a PDF.

That said, I'd still list what IC you're using, as well as what input you're giving it, and how close you are to the reference schematic.

 

[tnecniv]

Please do not cross-post to multiple sub-forums (especially when both are usually frequented by the same people!)

Unfortunately, your schematic is at or below the limits of readability--can you post a bigger version? If you're on a Windows machine, you can take a screenshot of a single frame using ALT+PrintScreen. There's also usually a CAD processor that can spit out a PDF or an image or some such. If you have something like PDF Creator or Adobe Acrobat, you can usually print directly to a PDF.

That said, I'd still list what IC you're using, as well as what input you're giving it, and how close you are to the reference schematic.

thanks for the reply and sorry for double post. i have attached the dac pdf file below.
components i am using are:

DAC - DAC0832LCN
2Mbit EEPROM - M27C2001-12F1
amplifier - MIK3860825/2
8bit Latch - SN74HCT573N
AND Gate - MM74HCT08N
speaker is 8ohm speaker

please advise. At a lost bcos not sure is hardware having problem or software.
Hardware is exactly the same as the schematic

 

[yungman]

I don't have time to read into the detail of the schematic. One thing already caught my eyes. According to the schematic, you left the EPROM unused address bits floating. This will not work as the open input can float and become a random pattern. You won't be able to read the correct location. Ground all the unused addresses, check your program to make sure it match the address location and try again before doing anything else.

 

[tnecniv]

my apologies for not indicating earlier, all the address bits of the eeprom are wired to the microprocessor data/address lines, they're not shown in the diagram so that it isn't so messy

 

[Jiggy-Ninja]

Before we can help you with what the problem is, you need to find out where the problem is. If you are able to reprogram the microcontroller to perform tests that will make it a lot easier.

First, check that you have ground and Vcc properly connected on EVERY chip, and that you are receiving power to every chip from your power source. Use a continuity checker to check for a solid connection between the power pins and their appropriate bus.

Numerous troubleshootings, by myself and my classmates, have been resolved by "check your power".

Notably, I see on your schematic that V_s of the audio amplifier is connected to ground, not Vcc. Double check that.

If you did not make such an annoyingly frustrating error in your circuit wiring, here's some other things to consider:

Double check your address and data lines, and make sure they are going to the right places, if you can, use different wires for address, data, power, ground, and whatever else. There's nothing I hate more than seeing a spaghetti mess of wires all in just two colors.

Once that's done, isolate each individual piece and make sure it works. Make sure that the uC is reading the EEPROM properly by having it sample certain addresses and displaying them somehow, such as on an LCD screen or a serial terminal on a computer.

Have the uC send fixed values to the DAC, and measure the output to see if it makes sense.

If those things work, then you need to step through your program and debug that somehow.

 

[tnecniv]

Before we can help you with what the problem is, you need to find out where the problem is. If you are able to reprogram the microcontroller to perform tests that will make it a lot easier.

First, check that you have ground and Vcc properly connected on EVERY chip, and that you are receiving power to every chip from your power source. Use a continuity checker to check for a solid connection between the power pins and their appropriate bus.

Numerous troubleshootings, by myself and my classmates, have been resolved by "check your power".

Notably, I see on your schematic that V_s of the audio amplifier is connected to ground, not Vcc. Double check that.

If you did not make such an annoyingly frustrating error in your circuit wiring, here's some other things to consider:

Double check your address and data lines, and make sure they are going to the right places, if you can, use different wires for address, data, power, ground, and whatever else. There's nothing I hate more than seeing a spaghetti mess of wires all in just two colors.

Once that's done, isolate each individual piece and make sure it works. Make sure that the uC is reading the EEPROM properly by having it sample certain addresses and displaying them somehow, such as on an LCD screen or a serial terminal on a computer.

Have the uC send fixed values to the DAC, and measure the output to see if it makes sense.

If those things work, then you need to step through your program and debug that somehow.

Thank you. i will give it a try now

 

[tnecniv]

i just tried sending certain values to the DAC. I gave a value that increases from 0-255 and loops , which gives a saw tooth waveform after it goes thru the dac. Am i right to say this?
If i am, i do not see any saw-tooth on the oscilloscope. is there anything i need to take care of (eg delay between each value i send)?

 

[yungman]

No, the instruction is so slow, it should take. Can you post a better schematic? Not only I cannot see the name of the pins, you did not even post the pin number. I looked at the data sheet of the DAC0832, it is /CS, /WR and other control signal, if you did not set up correctly, nothing will go into the DAC!

I can't really see the schematic. I see you have the ALE to strobe the data into the 373. I can't see any clock strobbing the DAC. So does it mean the 373 set the address to the EPROM and the data of the EPROM drive the DAC?

If so, look at whether you have the ALE strobe first, no stroble no nothing. Next, you should see the address lines of the EPROM toggling. This is to verify you do sent in data. Then look at the data lines output to the DAC. If you see lines toggling. You have data feed into the DAC.

But as I said, without a complete schematic, we can't talk. Look into the control signal and post back.

 

[berkeman]

I don't see it mentioned yet (I could have missed it), but in addition to other issues, you probably need to put a DC blocking capacitor between the DAC output and the LM386 input. Your DAC output is all above ground, so that will introduce an offset in the LM386 output that could saturate it.

 

[yungman]

Basically I can't see much on the schematic, no pin numbers, all bets are off.

 

[tnecniv]

Basically I can't see much on the schematic, no pin numbers, all bets are off.

I apologize for lack of clarity. The pins of the ICs are located beside the name.( i tried to put on the wire but i couldn't figure out how to do it.)

we are able to access the eeprom. yes the D-latch does set the address to the eeprom. The ALE works fine as i am able to access data from the eeprom.

i have placed a dc blocking cap between the DAC and the var. resistor.

 

[yungman]

Much better. I see you grounded /XFER and /WR2 to make the DAC Register straight pass through. You pull Ile high and use /CS and /WR1 to control the latching data into the first latch inside the DAC0832. You need to look at the timing to make sure you conform to the setup and hold time of both /CS and /WR1.

Second and the most important. This is an R/2R current DAC. The way you are doing is not good because first you cannot have C3 in between the output and the amp. This is a DC signal. But still without the C3, it will not very well because it needs a vertual ground. Did you look at the data sheet, it show how you hook up an opamp using the -ve input. basically pin 12 to ground. pin 11 straight to the -ve input of the opamp. Then the output of the opamp connect back to pin 9 to the internal resistor back to pin 12. This is the standard way of hooking up the current DAC. Go to this site and download the data sheet:

http://www.national.com/mpf/DA/DAC0832.html#Overview

You need to do the following.

1) Read page 12 to 14. It show you exactly what I tried to described to you.

2)Read page 11 on the timing required to strobe the data in and verify you have the signal.


If you do these two, you are going to have output signal.

 

[tnecniv]

Much better. I see you grounded /XFER and /WR2 to make the DAC Register straight pass through. You pull Ile high and use /CS and /WR1 to control the latching data into the first latch inside the DAC0832. You need to look at the timing to make sure you conform to the setup and hold time of both /CS and /WR1.

Second and the most important. This is an R/2R current DAC. The way you are doing is not good because first you cannot have C3 in between the output and the amp. This is a DC signal. But still without the C3, it will not very well because it needs a vertual ground. Did you look at the data sheet, it show how you hook up an opamp using the -ve input. basically pin 12 to ground. pin 11 straight to the -ve input of the opamp. Then the output of the opamp connect back to pin 9 to the internal resistor back to pin 12. This is the standard way of hooking up the current DAC. Go to this site and download the data sheet:

http://www.national.com/mpf/DA/DAC0832.html#Overview

You need to do the following.

1) Read page 12 to 14. It show you exactly what I tried to described to you.

2)Read page 11 on the timing required to strobe the data in and verify you have the signal.


If you do these two, you are going to have output signal.

i was told by another person from another forum that the lm386 has an internal feedback so there was no need to connect the Rfb from the Dac to the power amp.

i found another schematic from one of seniors(AZPDAC) and most ppl are using it can u tell me what's the difference between mine and this one n whether it would be better to use it.
Thanks

 

[yungman]

i was told by another person from another forum that the lm386 has an internal feedback so there was no need to connect the Rfb from the Dac to the power amp.

i found another schematic from one of seniors(AZPDAC) and most ppl are using it can u tell me what's the difference between mine and this one n whether it would be better to use it.
Thanks

I designed so many of these kind of circuits. Don't trust me, trust the data sheet. The reason is because it is R/2R, current is set by voltage drop from the reference voltage to the Iout pin. You can only set the current if you have a fix voltage at the Iout pin. If the Iout is driving a resistor and start swinging, how do you expect it to be accurate?

Also resistor inside the IC has 30% error compare to external resistors. BUT they can match very accurately between resistor INSIDE the same die. You can never mix voltage divider using internal and external resistors. I designed IC and I know. There is no if and buts about it. that is the reason why they always provide the Rfb the feedback resistor internally so that resistor will transform the current accurately when using an opamp shown in the data sheet.

there are "voltage DAC" that you can hook up like what you have. But this is a current DAC.

For you, the Iout has to be ground reference like you have to ground the /Iout.