Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Hello, bridging -IN and GND right at the terminal did not make a difference. Here are my measurements. True RMS ac values. I got sparks flying once :( Seems ok though. Edited by user Wednesday, May 30, 2012 12:56:21 AM(UTC)
| Reason: Not specified File Attachment(s):  sympatico_bom (1) measurements.xls (26kb) downloaded 8 time(s).You cannot view/download attachments. Try to login or register.
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
I tried the following today: I removed R10 (2k) thereby reducing the gain of the input stage by 6dB. The instability occurred at approximately the same input level (around -9dB). Can this tell us anything?
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
- Removed everything from enclosure - Replaced speaker load by a 100W 8ohm resistor - Reduced set-up to its simplest expression - Results shown below with a 1dB difference at the SE input Russ, can you advise on the next step? Sympatico design issue? - I can do scope captures at various resistor locations on the board - I'm ready to send one of my boards back to you for troubleshooting Regards, Martin mcoteca attached the following image(s): IMG00005-20120531-1915.jpg (131kb) downloaded 19 time(s). IMG00004-20120531-1911.jpg (62kb) downloaded 25 time(s). IMG00003-20120531-1910.jpg (60kb) downloaded 29 time(s).You cannot view/download attachments. Try to login or register.
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Sympatico is actually a rather old and proven design.
What you are describing does not actually fit the description of motor-boating that I am familiar with. Its actually a form of clipping.
Classical motor-boating is a self sustaining low frequency harmonic that grows on its own. That is not what I see you describing.
If actually looks to me like it could be that one half of the LM4780 is dead. That could be from operating it as you did.
I will look a bit closer at your measurements when I have time.
How is the other amplifier behaving?
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
One other theory is that you also could simply be pushing the LM4780 beyond it current limiting threshold. Or at least getting the junction temperature to where it triggers the protection circuit. This does not seem likely unless the thermal resistance to your heat sink is very bad.
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Please measure the DC voltage across C5 and C6.
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Yes, from the search I did afterwards, motorboating indeed means something else. It's now just an old title to an issue I'm doing my best to characterize as best I can with the tests I've done to date. I'm not sure if you have anything in mind when you write "operating as you did". If you imply that I might have shorted or overdriven the amps, yes that it is definitely possible, although I tried to be very careful. Both amps are behaving the same way (by sound), but I will test the other one further with scope measurements. Could a short to the heat sink via the screw have caused the death of one half of the LM4780? Or perhaps I temporarily grounded one of the outputs by mistake? Or let's say I overloaded the input, could this damage the LM4780?
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Yes shorting the back of the package (with a screw or otherwise) to GND can kill/damage the LM4780 because it is not insulated.
Overloading it is less likely to kill it.
It could also be something else, it difficult to say.
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Hi Russ, 14.81V across C5 and 14.80V across C6.
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Ok, thats good. One guess is you are just asking for more current than the AMP is designed to deliver. Sympatico is not designed as a subwoofer amp for certain. It is more for normal full range use. :) Make sure your rails are not sagging with the signal under test. If they are you will need more capacitance. Also make sure your not asking the LM4780 to deliver more current then it was designed for. Refer to the datasheet. You may also want to decrease the gain. Up to you. Cheers! Russ Edited by user Friday, June 1, 2012 3:43:31 AM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Hi Russ, thanks for all your answers. My understanding is that the Sympatico can provide 4 amps comfortably into any reasonable impedance load (say 4 ohms and up). With a 6 ohm load (DC impedance of my sub), the max power of the Sympatico would be 94W and max voltage output 24V. Perhaps the inductance of the sub load changes that conclusion somewhat. What I observe is that the instability starts somewhere at between 5 and 10V at the output of the amp. I'll check the screws to make sure I have a good thermal contact between the pad and the heat sink. Don't want to overtighten though! Rgds, Martin Edited by user Friday, June 1, 2012 5:21:18 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Are you measuring the differential voltage? Or the voltage to GND at one end of the output?
Remember the outputs are not ground referenced. They are referenced to each other. This means you need a scope that can do truly differential measurements to get an accurate voltage
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Hi Russ,
you're right. I was measuring the voltage to GND at one end of the output. I'm now using my scope with two probes (one at each end) to measure Vdiff = V+ minus V-.
I've done more tests to consolidate my results. In summary:
- everything behaves ok provided Vdiff is less than 20V pk-to-pk
- when the amp is hot, the problem occurs a couple of dBs lower
- the problem is NOT frequency specific; sorry for the erroneous reports early on, I was initially testing using the output of my sub active filter which boosts the lower frequencies below 40Hz); in reality, what I observe using my DAC output fed directly into the Sympatico is very output level consistent at any frequency (tested up to 1kHz)
- the rails are not sagging
- I tightened further the heat sink screws
So strange. Just can't put my finger on it.
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
I keep staring at my amps, but no bright thoughts of mine have yet emerged. I have sent PM and email to Brian. Awaiting response. To add to my previous post, I have now unscrewed the heat sinks from both amps and performed a visual inspection: the LM4780 of the right channel amp (on which I reported flying sparks in post 21) has pin 11 melted (V-, shared with several other pins). This can put into question test results reported afterwards in post 23 which were based on this amp although everything seems to be same as before. Thankfully, my left channel amp appears unscathed from any improper handling on my part; results reported in post 33 are based on this amp. As I indicated early on, both amps seem to go unstable when the sine output reaches around 20V pk-to-pk. As a further answer to Russ' post 30, I did check the LM4780 datasheet and I'm positive that I'm well under the maximum current capability; a quick ohm's law calculation with 10V voltage and 6 ohm impedance confirms it. I'm also positive that heat sink was tightly screwed on. I could send any time you like my left channel amp for a quick test if you would be willing to check it. A side by side comparison test with your own amp might quickly show any problem I might have created with mine and would confirm whether I installed something incorrectly (doubtful at this point) or damaged one of the ICs through improper handling. In the mean time, I will order a new LM4780 part to repair my right channel amp and for the time being I'm running my subs with my 20-year old Nakamichi 50W TA-2A Receiver and preamp bypassed at comfortably high levels. Finally, I would like to apologize in advance for any measurements mistakes and damage I have caused. Even though I'm far from an experienced DIYer, this is not my first project. I built one year ago the Linkwitz Pluto project using bare PCBs for the active crossover and featuring bridged LM3886s for the woofer amp. Generally speaking, my soldering skills and electrical measurements skills are not totally lacking. I'm also an electrical engineer by education, but I have only designed circuits very early on in my career and only for a short period of time. :) Best Regards, Martin Edited by user Friday, June 8, 2012 11:52:20 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
Unfortunately it simply is not possible for me to test your work.
Your calculation for power is not correct, You need to remember that each half of of the LM4780 actually see half of the load impedance, because the load is bridge tied. Back of the envelope calc on my part and I come up with something like 120W per side, 200W total.
Do you really plan on pushing the amp that hard? Do you really need to?
|
 1 user thanked Russ White for this useful post.
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Perhaps it was all a power calculation mistake on my part then. Beginners mistake :(
My application does not require to push the amp very hard in normal circumstances. The instructions I followed to build the sub and its associated active filter that is inserted at my preamp output recommend an amp 125W per channel to have some capability in reserve. Most of the time, I really don't listen very loud. The "motorboating" sound (I'm using quotes because I don't want to insinuate it is really that, as was established in previous posts) first occured when I was watching a movie. The active filter behind the amp boosts the signal quite a bit to extend the bass down to 20Hz; I've also measured a DC offset at the output of the filter which I really need to eliminate. Perhaps these factors combined to push the amp beyond its normal limit and hence the loud pops.
After this first occurence and the initial measurements I made early on, I switched to testing with pure tones directly from a volume-controlled source. If I was driving the amp too hard during those tests, it could have been because I was not relating correctly the calculated output power and current to the measured output voltage. I also now think that the gain of the amp is set too high and I should be removing R10.
I will revise my power calculation (feel free to share with me your closed form equation).
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Russ, I just wanted to add to my last post that I appreciate your patience and all your replies over the last 3 weeks. Hopefully, I'm not abusing too much. I'm hopeful I'm getting near the end of my "issue".
|
|
|
|
|
|
Rank: Administration
Groups: Administration, Customer
Joined: 10/24/2006(UTC)
Posts: 3,979
Location: Nashville, TN
Thanks: 25 times
Was thanked: 89 time(s) in 83 post(s)
|
The Sympatico is not at all designed to be pushed to clipping. It is also not really designed as a subwoofer amp, you would be better off with an amp designed for that purpose. It has no protection circuitry to deal with the nasty things that can happen when one side of the bridge tied amp looses steam (thus depriving one half of the fully symmetrical feedback). That is why you saw the nasty effects that you did. If the current protection circuitry decides to trip because of high junction temp or excessive current you will get some nasty behavior as you did. The good news is the very likely on the damaged amp all you need do is replace the LM4780, at least it's worth a shot. I am always glad to help. I appreciate the tone with which you have presented you questions and findings. People like you will always get further with me than the people who lose their temper. :) Edited by user Saturday, June 9, 2012 8:21:49 PM(UTC)
| Reason: Not specified
|
|
|
|
|
|
Rank: Member
Groups: Member
Joined: 10/11/2011(UTC)
Posts: 40
Location: Montreal, Canada
Thanks: 1 times
|
Hi Russ and All,
just a brief report to let you know that I replaced a while ago the damaged LM4780. I damaged a pad in the process and installed a wire from an SMT resistor directly to the pin. Overall the repair went quite well. The board is playing nicely and everything is back into my enclosure with transformers and all. I fixed permanently an issue I had at the output of the active filter circuit (DC offset) which feeds the Sympatico. I think that initially this DC offset was causing the amp to clip (while music playing) and afterwards when I was investigating the Sympatico's on their own I was probably trying to push them too far with my low frequency single tones. In any case, with the DC offset out of the equation so to speak, the Sympatico's warm up very little. I haven't encountered any clipping at medium volume levels. I didn't try to push the amps to their limits as there is no need for that; the sound level is plenty high to my taste. Given the purpose of the subs (bass extension mainly for music material), the Sympatico's seem to be a good match for the application.
The whole thing was a good learning experience for me.
Cheers,
Martin
|
|
|
|
|
|
Forum Jump
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.
