Thanks RS for your help. I've narrowed my problem down to D1. It was reversed. I started tracking the voltage from D3 and downwards. D3 & D4 measured 18. C2, C3, C4, C5 also measured 18. Across D1 it's zero even after I've de-soldered and corrected the polarity. So I might have overcooked it while I'm trying to de-solder it.
Once thing that I don't quite understand is U6. Between in-out it's 5.8v and ground-out is 12v. Is that because the circuit before it is open?
Once thing that I don't quite understand is U6. Between in-out it's 5.8v and ground-out is 12v. Is that because the circuit before it is open?
Last edited:
Matthaios, if you have a DMM you can run the voltage checks I'm working on publishing on my blog. If nothing is getting hot it sounds like there's a DC problem of some sort in one channel. Hopefully you're not using valuable headphones unless you're sure they're not seeing any DC. But the fuzzy sound might be a serious DC offset and when it stops completely it might be pure DC. If you don't have a DMM, turn off the amp immpediately after it stops playing in one channel. If you hear a big pop in the headphone when you turn it off, it's likely there's a large amount of DC on the output.
@RocketScientist: I double checked the temps of everything and from startup until a few minutes after it was turned on, all components were cool to touch. I then went ahead and checked the voltages with my dmm as follows.
AC - 12.8-13. I'm a bit confused on this, I thought I selected the wall wart from the BOM which is rated as a 12V, not the 13.5v+ that the testing states.
Battery - 11.9v on each, 23.3v across both
C8/C9 - 11.7v onv c8 and 11.6v on c9, 233v across both
U6 - 11.9v
U5 - 4.6v
C2/C - 16.6v
P2 - with the black pin in the square hole and red in the last hole it starts at 1.1v and slowly climbed up to 20v before I shut it off. Every other way I tried to measure P2 read 0.0v.
BTW, thanks for creating this project and for the help trying to figure this out.
Sent from my Transformer TF101 using Tapatalk
AC - 12.8-13. I'm a bit confused on this, I thought I selected the wall wart from the BOM which is rated as a 12V, not the 13.5v+ that the testing states.
Battery - 11.9v on each, 23.3v across both
C8/C9 - 11.7v onv c8 and 11.6v on c9, 233v across both
U6 - 11.9v
U5 - 4.6v
C2/C - 16.6v
P2 - with the black pin in the square hole and red in the last hole it starts at 1.1v and slowly climbed up to 20v before I shut it off. Every other way I tried to measure P2 read 0.0v.
BTW, thanks for creating this project and for the help trying to figure this out.
Sent from my Transformer TF101 using Tapatalk
Matthaios,
The AC power is a bit low but that's OK. The wall wart is rated at 12 volts, and depending on your home line voltage, should put out about 13.5 volts with no load.
I assume the battery voltages are with no batteries installed.
The pin outs are different on U5 and U6 and you're measuring is the voltage across U5--i.e. it's dropping 4.6 volts which is correct (16.6 - 12).
It's impossible for the DC offset at P2 to go past + or - 12 volts with respect to ground (the square hole). Do you mean 1.1 millivolts to 20 millivolts? (i.e. 0.001 - 0.02 volts)?
If the DC offset is really 1.1 volts or higher that's bad and would explain the distortion. Please don't use any headphones and wait for my revisions to the article.
The AC power is a bit low but that's OK. The wall wart is rated at 12 volts, and depending on your home line voltage, should put out about 13.5 volts with no load.
I assume the battery voltages are with no batteries installed.
The pin outs are different on U5 and U6 and you're measuring is the voltage across U5--i.e. it's dropping 4.6 volts which is correct (16.6 - 12).
It's impossible for the DC offset at P2 to go past + or - 12 volts with respect to ground (the square hole). Do you mean 1.1 millivolts to 20 millivolts? (i.e. 0.001 - 0.02 volts)?
If the DC offset is really 1.1 volts or higher that's bad and would explain the distortion. Please don't use any headphones and wait for my revisions to the article.
RS, sorry I accidently measured p2 as ac not dc. It is actually starting at around 0.7v and rapidly climbs up to 11v I assume, and stopped measuring at around 9v.
No rush or anything, and but when should we expect the update to the article? And again, thanks for the help.
Sent from my Transformer TF101 using Tapatalk
No rush or anything, and but when should we expect the update to the article? And again, thanks for the help.
Sent from my Transformer TF101 using Tapatalk
@Matthaios, @hbarradah, @kmc52087,
Please have a look at the following newly written section on my Blog. It's likely full of typos, formatting/editing problems, etc, as it's past my bedtime and I've been at it too long. But it should help quite a bit in many circumstances:
O2 Troubleshooting
Important! Please note I've also substantially revised the DIY Testing section as well and you need to make sure you follow what's there first:
O2 Initial DIY Testing
@Matthaios, I really can't think of what would cause a climbing DC offset like that but swap U3 and U4 and see if the problem swaps channels. If so you have a bad 4556. If it stays in the same channel, measure from ground to pin 5 of U3 and U4 and see if there's also an offset there? If so, check across C13 and C14. There's just not much to go wrong with the output stage that would cause a symptom like that besides the op amps themselves but you never know I guess.
Please have a look at the following newly written section on my Blog. It's likely full of typos, formatting/editing problems, etc, as it's past my bedtime and I've been at it too long. But it should help quite a bit in many circumstances:
O2 Troubleshooting
Important! Please note I've also substantially revised the DIY Testing section as well and you need to make sure you follow what's there first:
O2 Initial DIY Testing
@Matthaios, I really can't think of what would cause a climbing DC offset like that but swap U3 and U4 and see if the problem swaps channels. If so you have a bad 4556. If it stays in the same channel, measure from ground to pin 5 of U3 and U4 and see if there's also an offset there? If so, check across C13 and C14. There's just not much to go wrong with the output stage that would cause a symptom like that besides the op amps themselves but you never know I guess.
Make sure your power switch is on. You won't get a voltage drop across D1 unless there is some current going through it. Does your LED still light when the power switch is turned on? Since the voltage drop across D1 will be that 0.16V - or - so make sure your DMM is on a low range, if manual, like 2V and on DC volts.
0.16V is OK for D1 on the diode test mode since it is a Schottky. Try measuring between ground and D5's cathode, see if you are getting around -12V. Then measure the voltage drop across D5.
@RS, I wasn't expecting the updates to the article so quickly. I was expecting a few days at least. Thanks for your continual support on this project. I'm in middle of work and trying to test what I can, but it seems we're narrowing down the problem. I swapped u3 and u4. Same channel still. I checked the voltages also. It looks like the issue is at u3. I wasn't sure which was pin 5 so I checked 5 and 6. One of them was 8.5v and the other started to climb just like it was doing on p2. I checked p2 and it seemed the voltages were climbing at the same rate. U4 had the same pin at 8.5v and the one that was climbing stayed at 0.0v. Fyi, this was tested with batteries this time. I'll test with wall wart later.
Once I get a moment I'm going to go through the article updates as well.
Sent from my Motorola Atrix using Tapatalk
Once I get a moment I'm going to go through the article updates as well.
Sent from my Motorola Atrix using Tapatalk
You're welcome. Glad you got it working!
@mathaios, Please ground your negative DMM lead to one of the two center battery terminals (i.e. the negative terminal of BT1 closest to the gain switch) and use the positive probe to check all eight pins of U3 (the bad channel) after the DC offset has developed. Please report each of the eight voltage readings when you've done that such as:
Pin 1 = x volts
Pin 2 = y volts
.
.
Pin 8 = 11.8 volts
Pins 1/2, 3/4, and 6/7 are supposed to be connected together. But please measure each of the pins anyway at the IC in the socket from the top side of the board. That will help show any soldering problems, a problem with the socket, etc.
Please also report the voltage at each side of R12 and R13 when the amp has the offset (four more readings).
matthaios - one more thing to check in addition to all that RocketScientist posted. Check your connections on R12. Try re-heating them with the soldering iron. Also - with the power switch off - try a resistance reading across R12 and see if it is at least 40.2K.
In theory at least if R12 were missing, or a much higher value, the input bias current of U3 could start charging up C13. That would cause the non-inverting input voltage of U3 to ramp up over a few seconds and the output voltage would follow it in this configuration.
RocketScientist - those new troubleshooting and testing additions look great!
In theory at least if R12 were missing, or a much higher value, the input bias current of U3 could start charging up C13. That would cause the non-inverting input voltage of U3 to ramp up over a few seconds and the output voltage would follow it in this configuration.
RocketScientist - those new troubleshooting and testing additions look great!
I think this is the best place to make this post; hopefully I'm not derailing the troubleshooting. Sorry if I am.
Would anyone be kind enough to look over the front panel I designed? It's a play off of what's offered by JDS Labs (don't really want their logo) and the front panel GB (I was too late).
By my eye, everything looks OK, but I've never done vector drawing before and never designed a front panel so it might be rubbish. I've included the .fpd along with the .plt files I used in a zip below. Thanks.
Would anyone be kind enough to look over the front panel I designed? It's a play off of what's offered by JDS Labs (don't really want their logo) and the front panel GB (I was too late).
By my eye, everything looks OK, but I've never done vector drawing before and never designed a front panel so it might be rubbish. I've included the .fpd along with the .plt files I used in a zip below. Thanks.
OK, so inhaled it plugged back in to the wall wart now and found something interesting when testing u3.
Pin1 - slowly goes up in voltage
Pin2 - same as pin 1
Pin3 - read below
Pin4 - 11.7v
Pin5 - read below
Pin6 - same as pin 1
Pin7 - same as pin 1
Pin8 - same as pin 1
Pin 1,2, 6, and 7 slowly go up in voltage like it's doing in p2. The interesting part is when I measure pin 3 or 5 it reads those same voltages and then instantly goes down to 0.8v. Once I measure one of the other pins again, the voltages start at 0.8v and start to climb again.
R12 and r13 both read 0.003v.
Sent from my Transformer TF101 using Tapatalk
Pin1 - slowly goes up in voltage
Pin2 - same as pin 1
Pin3 - read below
Pin4 - 11.7v
Pin5 - read below
Pin6 - same as pin 1
Pin7 - same as pin 1
Pin8 - same as pin 1
Pin 1,2, 6, and 7 slowly go up in voltage like it's doing in p2. The interesting part is when I measure pin 3 or 5 it reads those same voltages and then instantly goes down to 0.8v. Once I measure one of the other pins again, the voltages start at 0.8v and start to climb again.
R12 and r13 both read 0.003v.
Sent from my Transformer TF101 using Tapatalk
@agdr, I went ahead and checked the stuff you told me. The resistance on both R12 and r13 are 39.9k. I ended up resoldering in R12 and it looks to have solved my issue. The voltages in p2 were both 0.003v and u3 pin voltages match u4 now. Pin 4 and 7 are 11.7v and the rest are 0.003v. Testing it now to make sure.
Sent from my Transformer TF101 using Tapatalk
Sent from my Transformer TF101 using Tapatalk
Hey that is good news!
3mV is right on the mark for the output offset voltages. That is what I get on my O2s.
@matthaios, agdr and I were both on the same track. That's why I asked you to measure both ends of R12 as I suspected it might be an open circuit. For the geeks in the audience, the 4556 was "sourcing" its bias current through the coupling capacitor from the input stage until. That only worked for a while and had the side effect of charging the capacitor to a DC value which was passed on to the output.
This is a good example of why it's important to follow the Initial DIY Testing. Those who skip it might end up with massive DC offset from a soldering error.
This is a good example of why it's important to follow the Initial DIY Testing. Those who skip it might end up with massive DC offset from a soldering error.