Hi,
My p2p lm3886 chipamp was working very well for few days until I meessed up with the cable of one speaker and I installed a dodgy ebay dact type attenuator. The fact is that I must have shorted either the speaker cables or the signal cables and one channel is gone.
I just switched on the channel and it sounded a very very quiet, constant tic tic tic tic tic and caput!.
Furtunately the speaker was not damaged and I do not think there was any shorting involving the DC cables. And there was no smoke or any nasty smells or weird sounds and the chip was not hot.
So if there was a problem on the signal path I do not know were to start to correct it. Should I get a nother chip? Could the caps or resistors also be damaged- which ones? Some guideance on where to start would be very appreciated.
This are the schematics http://www.vikash.info/audio/p2plm3886/
Thanks in advance.
Antonio
My p2p lm3886 chipamp was working very well for few days until I meessed up with the cable of one speaker and I installed a dodgy ebay dact type attenuator. The fact is that I must have shorted either the speaker cables or the signal cables and one channel is gone.
I just switched on the channel and it sounded a very very quiet, constant tic tic tic tic tic and caput!.
Furtunately the speaker was not damaged and I do not think there was any shorting involving the DC cables. And there was no smoke or any nasty smells or weird sounds and the chip was not hot.
So if there was a problem on the signal path I do not know were to start to correct it. Should I get a nother chip? Could the caps or resistors also be damaged- which ones? Some guideance on where to start would be very appreciated.
This are the schematics http://www.vikash.info/audio/p2plm3886/
Thanks in advance.
Antonio
That ticking sound could be the Spike protection. If you are lucky it is protecting the IC, because the fault is still present. But may as well be the only thing left alive.
Here are some questions that you can try to answer yourself, and that may help you to find the mistake.
If you swap the channels, does the ticking swap as well? If not, the amp is okay and you have to check elsewhere, e. g. preamp, input buffer, attenuator, and so on.
Are both rail voltages present and at their correct values? If not check the fuses.
Fuses okay or not present? Disconnect the amp from the power supply and measure again.
Are the voltages too low with the amp connected and correct with the amp disconected? There is a short somewhere on the amp board, the LM3886 might be broken. Take it out, and check the voltages again.
Voltages okay? Replace LM3886.
Are the rail voltages still low with the amp disconnected? Fault is in the power supply.
Is that attenuator still connected? Disconnect it, and check again. Fault gone? Check the attenuator connections.
Check for shorts everywhere and correct them.
Is the ticking present only with something connected to the inputs? Search the fault outside of the amp.
Ticking always present? Fault is probably in the amp.
Measure the output voltage of the ticking amp. About the same value as the rail voltage? Replace the LM3886.
More than half the rail voltage? Replacing may or may not help, but worth a try.
Here are some questions that you can try to answer yourself, and that may help you to find the mistake.
If you swap the channels, does the ticking swap as well? If not, the amp is okay and you have to check elsewhere, e. g. preamp, input buffer, attenuator, and so on.
Are both rail voltages present and at their correct values? If not check the fuses.
Fuses okay or not present? Disconnect the amp from the power supply and measure again.
Are the voltages too low with the amp connected and correct with the amp disconected? There is a short somewhere on the amp board, the LM3886 might be broken. Take it out, and check the voltages again.
Voltages okay? Replace LM3886.
Are the rail voltages still low with the amp disconnected? Fault is in the power supply.
Is that attenuator still connected? Disconnect it, and check again. Fault gone? Check the attenuator connections.
Check for shorts everywhere and correct them.
Is the ticking present only with something connected to the inputs? Search the fault outside of the amp.
Ticking always present? Fault is probably in the amp.
Measure the output voltage of the ticking amp. About the same value as the rail voltage? Replace the LM3886.
More than half the rail voltage? Replacing may or may not help, but worth a try.
pacificblue said:
Thanks a lot Pacific. The problem seems solved but I have to keep checking to find the precise cause.
Voltages were ok and I removed the attenuator. Then I tried with a different pot and nothing. Not even ticking.
Then I checked all the connections and found that the power ground to pot and speaker and also the signal ground to input were either disconnected or making a false contact (my set up is provisional that is why they were not soldered but just joined with metal sockets).
Also I must have been making the wrong connections to the attenuator and pot because I removed the pot and connected the source directly without volume control and Voila! the amp worked well. I believe the IC did indeed trigger its protection circuits due to the false contact and possible connection errors.
I learned a lesson: always double check connections and make them proper even if temporary.
I am glad I did not kill my main speakers. I was told that a false contact can easily kill an amp and speakers.
I am sure your troubleshooting procedure will be of much help to other newbies as well.
Thanks again.
Antonio
A Sanchez
Get yourself some cheap test speakers, or even just a couple of test drivers. Its better to fry Richersounds cheapies than your pride and joy.
I take it you are aware of the "in series lightbulb tester" as often recomended about these parts? It saved me when I accidentally shorted a new playstation power supply I was measuring. There was a spark and the bulb lit but no damage was done. Tape up your multimeter probes leaving only the tips exposed next time you have it out, it saves shorts if you are poking about in tight spaces!
John
Get yourself some cheap test speakers, or even just a couple of test drivers. Its better to fry Richersounds cheapies than your pride and joy.
I take it you are aware of the "in series lightbulb tester" as often recomended about these parts? It saved me when I accidentally shorted a new playstation power supply I was measuring. There was a spark and the bulb lit but no damage was done. Tape up your multimeter probes leaving only the tips exposed next time you have it out, it saves shorts if you are poking about in tight spaces!
John
Actually I will take the opportunity to ask another question in this thread rather than starting a new one.
Its about volume/gain:
Currently the amp has 20db gain because it is supposed to be used with an active pre, which it is not the case. So at the moment at full volume the amp is loud but not much and I feel the speakers could easily handle more power. Also I feel a bit weird turning the volume more than half turn to obtain a moderate volume.
So the questions are: will an increase of the gain to around 30db make a noticeable difference? If yes, should I go for a bit more than 30db?
Also, is it the same to alter either 2k RF (R3 in schematics see below) or the ground 100R resistor R2 to achieve the desired gain. Basically I could go for 3k for RF and leave R2 untouched or change R2 to 65R and leave RF untouched. So, is it the same to change one or the other? Or should I change both and go for 20k and 680R . The designer did suggest to keep the resistors low but I would like to understand the logic and benefits keeping values low or high if in the end the result (ie 30db gain) is the same.
If needed these are the amp and PSU schematics: http://www.vikash.info/audio/p2plm3886/ I am using 2x22 300v toroids (one per channel) and a 10k pot. My speakers are Jordan mltl full range driver : 6 Ohms and below 88 sensitivity
Many thanks
Antonio
Its about volume/gain:
Currently the amp has 20db gain because it is supposed to be used with an active pre, which it is not the case. So at the moment at full volume the amp is loud but not much and I feel the speakers could easily handle more power. Also I feel a bit weird turning the volume more than half turn to obtain a moderate volume.
So the questions are: will an increase of the gain to around 30db make a noticeable difference? If yes, should I go for a bit more than 30db?
Also, is it the same to alter either 2k RF (R3 in schematics see below) or the ground 100R resistor R2 to achieve the desired gain. Basically I could go for 3k for RF and leave R2 untouched or change R2 to 65R and leave RF untouched. So, is it the same to change one or the other? Or should I change both and go for 20k and 680R . The designer did suggest to keep the resistors low but I would like to understand the logic and benefits keeping values low or high if in the end the result (ie 30db gain) is the same.
If needed these are the amp and PSU schematics: http://www.vikash.info/audio/p2plm3886/ I am using 2x22 300v toroids (one per channel) and a 10k pot. My speakers are Jordan mltl full range driver : 6 Ohms and below 88 sensitivity
Many thanks
Antonio
john blackburn said:
John
All I ignore about gainclones and electronics I know it about saxophones and I can tell you that you should not even try to re pad one.
Unless you want to learn, in which case I suggest you to buy the cheapest second hand chinese sax plus some good books and a mentor, it is not worth even trying. The cost of the pads and the tools, plus the time and the almost certain bad result should discourage anyone.
A sax is much more complex than it looks as soon as you start removing screws.
What I suggest you is to really shop around in order to find a good repairer with reasonable pricing (which is not easy in the UK- but I can help you if you want). The good thing is that a good re-pad and proper care of the sax can last many years.
Antonio
Raising the gain from 20dB to 30dB means ~3,16 times higher output voltage, 10 times higher output power and the acoustic impression of double the volume for the same input voltage. This is of course only true as long as you don't reach the amplifiers voltage and power limits.
680/22k is widely used, and is a reasonable value for your rail voltage. It will give you nominal output power into 8 Ohms at somewhere between -0,4dB and 0,5dB input level, depending on the size and quality of your transformer and the amount of µF in your power supply.
680/22k is widely used, and is a reasonable value for your rail voltage. It will give you nominal output power into 8 Ohms at somewhere between -0,4dB and 0,5dB input level, depending on the size and quality of your transformer and the amount of µF in your power supply.
pacificblue said:
Thanks Pacific,
So, increasing the gain would indeed make a noticeable difference. Can you guess why the designer of the schematics recommended not to go above 4.7k resistor values when increasing the gain?
By the way I have around 24 700 uf per rail and decent quality 300va toroids.
Antonio
using 2k0 & 100r (from Vikash) gives a gain of +26.4dB (21times). This is well above the minimum gain of +20dB (10times) stated by National.A Sanchez said:
Your speakers @ 88dB/2.83V/m will probably accept peak voltage of about 20 to 40Vpk. With an amp gain of +26dB you need a maximum input signal of 1Vpk to 2Vpk to get maximum power into your speakers.
The amp will overload @ ~1.3Vpk~=900mVac
Try measuring the source voltages for typical music signals.
I would expect 100mVac to 200mVac of signal.
Which model of LM3886 do you use ? T or TF ?
If you have the T-Model, did you add a washer ?
Did you touch the heatsink with any sort of signal cable when the amp get in to spike protection?
Have you assembled an Lowpass filter at the Input of the LM ? (Oscillations à Spike)
Where the ground connections of the signal cables solid connected when it happened ?
One LM3886 per speaker ergo nothing paralleled nor bridged; right ?
artQuake
If you have the T-Model, did you add a washer ?
Did you touch the heatsink with any sort of signal cable when the amp get in to spike protection?
Have you assembled an Lowpass filter at the Input of the LM ? (Oscillations à Spike)
Where the ground connections of the signal cables solid connected when it happened ?
One LM3886 per speaker ergo nothing paralleled nor bridged; right ?
artQuake
AndrewT said:
Thanks Andrew. All your posts are not only helpful but they also require thinking and encourage learning.
There seems to be a big confusion in many threads regarding the use of terminology in connection to amplifers' "gain". You mention a gain in dB and a "times" gain. I believe people usually mention dB gain while they mean "times" gain. So, the usual formula 1+Rf/R3 seem to give the "Times gain" but people actually say dB gain. For example in the chipamp and audisector manuals it is said that their products gain is set to 33dB (1+22k/680) , but according with what you are saying they mean 33 times or somewhere above 35db (sorry I could not figure out the formula to reach an exact number). Is this right?
Anyway, I will check my CDs mVac signal and if it is standard 100 to 200 mVac I understand I could go up to 30 or 33 times gain without problem.
John,
It is a bit of both, vodoo and mechanics.... Getting the pads to seal perfectly and the keys to move and feel evenly and smooth is something that only experienced people can do. So, whenever your sax is done by a good repairer you feel that magic....vodooo.
The main problem is that a saxophone that does not work perfectly is virtually useless regardless of your level of playing, and it is specially bad for beginners.
Do you know the model, year, serial number of the Conn? I might be interested in buying it if you were planning to sell it.
Antonio
It is a bit of both, vodoo and mechanics.... Getting the pads to seal perfectly and the keys to move and feel evenly and smooth is something that only experienced people can do. So, whenever your sax is done by a good repairer you feel that magic....vodooo.
The main problem is that a saxophone that does not work perfectly is virtually useless regardless of your level of playing, and it is specially bad for beginners.
Do you know the model, year, serial number of the Conn? I might be interested in buying it if you were planning to sell it.
Antonio
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.