Please see example schematic...
Are these coupling capacitors correctly placed?
C1 is the input coupling capacitor. An electrolytic is not the best choice for here as a film cap is the better choice. I use a film cap here now.
However I have seen many schematics this C1 cap has + going to the base of Q1 and also some with - going to base. I am confused as to which it is.
C2-C4- I understand them...they serve to bypass the emitter resistors.
C3-C5-C6- Are these polarized correctly?
In my power amplifier circuit, C6 would be the output cap blocking the DC so my speakers would not get fried but today several caps got very hot and started leaking, a couple even popped. I measured 17.5v DC on the emitter of Q3 and used a 10,000uf 35v electrolytic capacitor here. I have tried many values here and a smaller capacitance attenuates 20hz too much. I have a 40V 3300uf nichicon on it now and have been using it for a while with no problems apart from much less bass.
I read somewhere that a bipolar electrolytic can be used here but have never tried or seen one at 10,000uf. Anyone know of any? Nor have I seen a film cap of that size (it would surely be enormous in size and price).
I am at a loss as to why the 10'000uf ones popped......
Maybe they were cheap...3.50 euro..
Ripple or voltage rating too low?
Not right way around?
I also used some old PYE 35v 10,000uf which I ripped out of an old transmitter....they fizzled and popped also.
Any suggestions as to what type to use here? Looks like my only choice is normal polarized electrolytic.
Decoupled max output 7.5vrms@20hz into 2ohm
Are these coupling capacitors correctly placed?
C1 is the input coupling capacitor. An electrolytic is not the best choice for here as a film cap is the better choice. I use a film cap here now.
However I have seen many schematics this C1 cap has + going to the base of Q1 and also some with - going to base. I am confused as to which it is.
C2-C4- I understand them...they serve to bypass the emitter resistors.
C3-C5-C6- Are these polarized correctly?
In my power amplifier circuit, C6 would be the output cap blocking the DC so my speakers would not get fried but today several caps got very hot and started leaking, a couple even popped. I measured 17.5v DC on the emitter of Q3 and used a 10,000uf 35v electrolytic capacitor here. I have tried many values here and a smaller capacitance attenuates 20hz too much. I have a 40V 3300uf nichicon on it now and have been using it for a while with no problems apart from much less bass.
I read somewhere that a bipolar electrolytic can be used here but have never tried or seen one at 10,000uf. Anyone know of any? Nor have I seen a film cap of that size (it would surely be enormous in size and price).
I am at a loss as to why the 10'000uf ones popped......
Maybe they were cheap...3.50 euro..
Ripple or voltage rating too low?
Not right way around?
I also used some old PYE 35v 10,000uf which I ripped out of an old transmitter....they fizzled and popped also.
Any suggestions as to what type to use here? Looks like my only choice is normal polarized electrolytic.
Decoupled max output 7.5vrms@20hz into 2ohm
Is this your power amp circuit? Resistively biased class A with single supply?
Anyway, whether C3 and C5 are correctly polarized is impossible to say. They are if the collectors are biased at higher voltages than the bases of the next stages. The others seem OK to me.
Anyway, whether C3 and C5 are correctly polarized is impossible to say. They are if the collectors are biased at higher voltages than the bases of the next stages. The others seem OK to me.
This is not the actual circuit, it's just an example. The actual needs much tidying up before it's presentable. My power amp is a dual supply class A with current biased emitter follower output. I just put a brand new 50 volt 10,000uf nichicon on C6 10 miuntes ago. It's hot and the top of it is swollen up again. Another good cap (5.50euro) in the bin. I am at a loss as to why. The 3300uf 40v nichicon is ok and has been for months. Just I was fine tuning the frequency response today I found 3300uf too small but its the only capacitor that works there so far.
Epcos and Pye got hot within 10-15 minutes and leaked....except the nichicon, I caught it on time but it was hot and top is all bulged up. The blue nichicon is fine. I got it from an old Tektronix analogue oscilloscope. I wonder whats different about this compared to the others.
5910-01-233-9285, 5910012339285, LNK1G332TSM, NKB040SM3301HB Data. Get Quote & Buy
5910-01-233-9285, 5910012339285, LNK1G332TSM, NKB040SM3301HB Data. Get Quote & Buy
Attachments
Use bipolar electrolytics if you aren't sure.
An electrolytic will take a reverse voltage up to about 1v5 then it will act like a short circuit.
Also you could use back to back electrolytics if you want.
An electrolytic will take a reverse voltage up to about 1v5 then it will act like a short circuit.
Also you could use back to back electrolytics if you want.
Parasitic oscillation? A 20W Class A audio amplifier can also be a 50W Class C RF oscillator - that might cook an electrolytic.
Back to back electrolytics is a good idea, first I ever heard of it. I will do more reserach on it and implement it if I can come up with no other solution or fix. Thanks Nigel.
DF96, I hooked the oscilloscope up to the output before and after the capacitor and there does not seem to be any oscillations that may cause concern. There is some high frequency content in the signal but they are only a few mVrms. The oscilloscope would read similar even if the amp was off. That shows it's not acting as a Class C RF amp right?
Maybe it's what Nigel said,
''An electrolytic will take a reverse voltage up to about 1v5 then it will act like a short circuit.''
That seems to be whats happening. Polarized caps are ok for small signal(pre-amp) but I would prefer film caps for small signal anyway.
I suppose there is no point in even trying a different polarized electrolytic with a higher rating because of this 1v5 limit is there?
Looks like my only choices are polarized in series(back to back) or several non polarized in parallel.
I also did a DC multi-meter test on the output, thinking that this blue nichicon is shorted already but just not leak electrolyte or get hot but it reads zero DC. Looking at the datasheet it looks like any other cap. Still wondering why it's holding up while the others are frying.
DF96, I hooked the oscilloscope up to the output before and after the capacitor and there does not seem to be any oscillations that may cause concern. There is some high frequency content in the signal but they are only a few mVrms. The oscilloscope would read similar even if the amp was off. That shows it's not acting as a Class C RF amp right?
Maybe it's what Nigel said,
''An electrolytic will take a reverse voltage up to about 1v5 then it will act like a short circuit.''
That seems to be whats happening. Polarized caps are ok for small signal(pre-amp) but I would prefer film caps for small signal anyway.
I suppose there is no point in even trying a different polarized electrolytic with a higher rating because of this 1v5 limit is there?
Looks like my only choices are polarized in series(back to back) or several non polarized in parallel.
I also did a DC multi-meter test on the output, thinking that this blue nichicon is shorted already but just not leak electrolyte or get hot but it reads zero DC. Looking at the datasheet it looks like any other cap. Still wondering why it's holding up while the others are frying.
With the example circuit, there is no way the output can go negative, so the polarity of the voltage across the electrolytic output capacitor is perfectly predictable. With a dual supply circuit (assuming that you mean +x V, 0 V and -x V) it may not be that predictable what the voltage across the cap will be, especially in the peaks of low-frequency signals. Like Nigel wrote, any peak below -1.5 V...-2 V can cause trouble unless you use bipolar capacitors or capacitors in anti-series.
PLEASE post the circuit you ACTUALLY BUILT, not the one YOU DID NOT:
How can you expect sensible answers otherwise? 😕 🙄
How can you expect sensible answers otherwise? 😕 🙄
Actually all the answers I got has been so helpful and they all make perfect sense.
Thanks.
Marcel yes, VCC, GND, VEE on a current sourced emitter follower. I think you are right. I checked just now after reading your post. The signal swings max 7.5vrms @20hz with a 10,000uf cap on the output. Surely enough to fry the cap. With the same signal strength @20hz the output with the 3300uf cap reads only 2.65vrms, which explains why the 3300uf does fry as it is not pulled down pass the -2V limit.
I think I would have kept buying more and more expensive and different 10,000uf caps and blown them all if it was not for the help I got. Now I will invest in some bipolar ones.
I did some reading up on bipolar caps today and somewhere I read that internally they are a back to back configuration anyway. Just was hard to find one large enough, and the large ones are quite expensive. I will try 2 back to back first and see how it goes.
Thanks.
Marcel yes, VCC, GND, VEE on a current sourced emitter follower. I think you are right. I checked just now after reading your post. The signal swings max 7.5vrms @20hz with a 10,000uf cap on the output. Surely enough to fry the cap. With the same signal strength @20hz the output with the 3300uf cap reads only 2.65vrms, which explains why the 3300uf does fry as it is not pulled down pass the -2V limit.
I think I would have kept buying more and more expensive and different 10,000uf caps and blown them all if it was not for the help I got. Now I will invest in some bipolar ones.
I did some reading up on bipolar caps today and somewhere I read that internally they are a back to back configuration anyway. Just was hard to find one large enough, and the large ones are quite expensive. I will try 2 back to back first and see how it goes.
Bipolar caps are not back to back normal caps. They use two anode foils. I agree with JMFahey: please show us what you built, not what you didn't build.
You need to understand that it is extremely common for someone seeking remote diagnosis from us to not only withhold information (as you have done) but also feed us misleading information i.e. to do something different from what he says he has done - either through simple mistake or inexperience leading him to think that the difference between what he has done and what he has described doesn't matter when in reality it turns out to be the crux of the matter.
What does that mean? Which signal; exactly where? We need details.
You need to understand that it is extremely common for someone seeking remote diagnosis from us to not only withhold information (as you have done) but also feed us misleading information i.e. to do something different from what he says he has done - either through simple mistake or inexperience leading him to think that the difference between what he has done and what he has described doesn't matter when in reality it turns out to be the crux of the matter.
I have no experience with bipolar electrolytic caps. It was just something I read.
The circuit is still in prototype stage on breadboard. I don't draw up the schematics until the prototype is finished and it is time to do PCB design. I really do not have the output drawn up yet. I was finalizing the prototype and adjusting the frequency response to draw up the schematic to do the pcb when the capacitor problem occurred - when I changed the 3300uf for a 10,000uf one. The preamp part is only partially drawn up and it is still a work in progress.
The question was about cap polarity. I did not think it was important to draw up the prototype as it is just a current biased emitter follower. Fair enough I probably should have also mentioned that and that I was using a dual rail power supply also but the circuit is all over the internet.
https://www.wikiwand.com/en/Common_collector
See figure.4
I hope that clears things up for you.
''The signal swings max 7.5vrms @20hz with a 10,000uf cap on the output.''
The 20Hz 'sine' signal which came from the signal generator went through the vas(voltage amplification stage) and into the base of the output transistor. Form there the output was taken from the emitter and the DC was decoupled with a 10,000uf capacitor. The amplified A/C sine signal reached a maximum magnitude of 7.5vrms while driving a 2ohm load at 1%thd. The oscilloscope probe was on the output side of the 10'000uf cap.
The 10'000uf cap was having none of it. That's what the problem was and not the emitter follower.
My theory on it at the moment is I don't think it was parasitic RF oscillations as it would see that on the scope, especially one that would be large enough to blow up a 10'000uf 40V capacitor. Nor was it the emitter follower circuit, one that is current biased or not with a single or dual supply. At the end of the day it is still basically just an emitter follower. So I did not think that would be misleading or withholding any additional information that would help with a diagnosis. I really did not think it was common emitter problem. It was specifically a capacitor problem....which polarity when DC decoupling a signal and which one or type to use for an emitter follower output. I apologize if it misled some people. It was not the polarity either as that's confirmed.
The problem is that the cap was just too big at 10'000uf.
I think that cap would have blown no matter which of the above configurations of emitter follower I connected it to at that voltage.
I could use a bipolar or back to back electrolytics but I still don't know much about them and have not tried yet.
This morning I came up with the solution of putting 2200uf polarized electrolytics in parallel. I only had 3 2200uf (6600uf) 40v only so i replaced them with the 3300uf 40v one and it's working a treat. Cool as a cucumber. Been running now for about 2 hours now and I would expect it to still function in the future. I just need to add one more or change that to 3@ 3300uf to reach my target.
Each 2200uf cap should now only see less than -2vrms or there abouts. Not enough to cause problems.
I did not know at low frequencies that polarized electrolytics had this +/- 2v limit when decoupling a large audio signal. Now I know.
Thanks
People might not be able to give you all the finer details of what they design for many reasons. Perhaps there might be intellectual property issues, they signed a non disclosure agreement, their company would not like them to disclose details of the project they are working on, they might have childhood trust issues....who knows what? Or maybe like me...just don't have it ready or the time to draw it up and thought that the information he gave was enough to diagnose a capacitor problem.
The circuit is still in prototype stage on breadboard. I don't draw up the schematics until the prototype is finished and it is time to do PCB design. I really do not have the output drawn up yet. I was finalizing the prototype and adjusting the frequency response to draw up the schematic to do the pcb when the capacitor problem occurred - when I changed the 3300uf for a 10,000uf one. The preamp part is only partially drawn up and it is still a work in progress.
The question was about cap polarity. I did not think it was important to draw up the prototype as it is just a current biased emitter follower. Fair enough I probably should have also mentioned that and that I was using a dual rail power supply also but the circuit is all over the internet.
https://www.wikiwand.com/en/Common_collector
See figure.4
I hope that clears things up for you.
''The signal swings max 7.5vrms @20hz with a 10,000uf cap on the output.''
The 20Hz 'sine' signal which came from the signal generator went through the vas(voltage amplification stage) and into the base of the output transistor. Form there the output was taken from the emitter and the DC was decoupled with a 10,000uf capacitor. The amplified A/C sine signal reached a maximum magnitude of 7.5vrms while driving a 2ohm load at 1%thd. The oscilloscope probe was on the output side of the 10'000uf cap.
The 10'000uf cap was having none of it. That's what the problem was and not the emitter follower.
My theory on it at the moment is I don't think it was parasitic RF oscillations as it would see that on the scope, especially one that would be large enough to blow up a 10'000uf 40V capacitor. Nor was it the emitter follower circuit, one that is current biased or not with a single or dual supply. At the end of the day it is still basically just an emitter follower. So I did not think that would be misleading or withholding any additional information that would help with a diagnosis. I really did not think it was common emitter problem. It was specifically a capacitor problem....which polarity when DC decoupling a signal and which one or type to use for an emitter follower output. I apologize if it misled some people. It was not the polarity either as that's confirmed.
The problem is that the cap was just too big at 10'000uf.
I think that cap would have blown no matter which of the above configurations of emitter follower I connected it to at that voltage.
I could use a bipolar or back to back electrolytics but I still don't know much about them and have not tried yet.
This morning I came up with the solution of putting 2200uf polarized electrolytics in parallel. I only had 3 2200uf (6600uf) 40v only so i replaced them with the 3300uf 40v one and it's working a treat. Cool as a cucumber. Been running now for about 2 hours now and I would expect it to still function in the future. I just need to add one more or change that to 3@ 3300uf to reach my target.
Each 2200uf cap should now only see less than -2vrms or there abouts. Not enough to cause problems.
I did not know at low frequencies that polarized electrolytics had this +/- 2v limit when decoupling a large audio signal. Now I know.
Thanks
People might not be able to give you all the finer details of what they design for many reasons. Perhaps there might be intellectual property issues, they signed a non disclosure agreement, their company would not like them to disclose details of the project they are working on, they might have childhood trust issues....who knows what? Or maybe like me...just don't have it ready or the time to draw it up and thought that the information he gave was enough to diagnose a capacitor problem.
What you describe should not cook a cap. Therefore something is happening which ought not to happen. This may mean that what you think you are doing is not what you are actually doing. Simply swapping from 3mF to 10mF will not create a problem of polarity. Swapping from a small cap to a big one (i.e. physically larger) may create instability due to capacitive feedback - that is why we need details.
Note that simply passing a 7.5V signal will not upset a cap, as the cap will not see much of the AC voltage. Only if the cap is acting as a relevant high pass filter will problems occur. If so, swapping from 3mF to 10mF should improve things.
High frequency RF is not always easy to see. Emitter followers are known to suffer from instability if they see a capacitive load, so if you think you are using "just an emitter follower" then maybe you need to do some more reading about emitter followers.
Someone seeking pro bono commercial consultancy here ought to say so, as we hope students seeking help with their assignment will do too.
Not if the problem is nothing to do with polarity. Anyway, it happens that all of us have our crystal balls away for annual servicing at present (we got a good bulk deal!) so we only know what you choose to tell us.
Whatever the reason was the problem is fixed by the 3x 2200uf caps in parallel. I have not changed the follower in any way.
I have no experience of this but I would imagine that if this was an issue, adding 4X 2200uf (8800uf) or 5 will cook also but that is not the case. I don't have more caps to add to this as I have only 3, but I am quite sure that if I add 3x3300uf (9900uf) they will not cook either. While one 10,000uf one will and have done so in my last 3 recent attempts.
I agree that emitter followers can suffer from RF oscillations and that they might be hard or not possible to see on a scope but I think that would cause distortions in the sine signal and not cook capacitors. I don't see any other than the normal harmonic distortion when the output is driven to clipping. Other times the sine wave at the output is clean. How would I see these RF oscillations? Any suggestions?
Why does it matter who is looking for consultancy? Why do they have to say even? If they wanted everyone to know they say and also have a profile to share that on. You can't charge them money for you answer here anyway. I don't really understand that. If I see a question that I might understand and I can attempt an answer it would be my pleasure regardless of who it is. If I don't understand the question there is nothing much for me to say. The focus should be on the discussion or problem to be solved and not the person.
As far as to why the solution worked I don't really know why but it worked. But it does kind of show that what others have said about the +/- 1v5- 2v limit being right.
This is exactly what I thought initially. I was so shocked that they fizzled up like that and thought I was going crazy....That is what prompted me to start this thread. I have built this circuit perhaps near a hundred times already although not at such high power. I never had a problem with it and still don't. I am quite confident that it was not the follower circuit.
I have no experience of this but I would imagine that if this was an issue, adding 4X 2200uf (8800uf) or 5 will cook also but that is not the case. I don't have more caps to add to this as I have only 3, but I am quite sure that if I add 3x3300uf (9900uf) they will not cook either. While one 10,000uf one will and have done so in my last 3 recent attempts.
I agree that emitter followers can suffer from RF oscillations and that they might be hard or not possible to see on a scope but I think that would cause distortions in the sine signal and not cook capacitors. I don't see any other than the normal harmonic distortion when the output is driven to clipping. Other times the sine wave at the output is clean. How would I see these RF oscillations? Any suggestions?
Why does it matter who is looking for consultancy? Why do they have to say even? If they wanted everyone to know they say and also have a profile to share that on. You can't charge them money for you answer here anyway. I don't really understand that. If I see a question that I might understand and I can attempt an answer it would be my pleasure regardless of who it is. If I don't understand the question there is nothing much for me to say. The focus should be on the discussion or problem to be solved and not the person.
As far as to why the solution worked I don't really know why but it worked. But it does kind of show that what others have said about the +/- 1v5- 2v limit being right.
This is exactly what I thought initially. I was so shocked that they fizzled up like that and thought I was going crazy....That is what prompted me to start this thread. I have built this circuit perhaps near a hundred times already although not at such high power. I never had a problem with it and still don't. I am quite confident that it was not the follower circuit.
[Bolding mine, in case anyone wonders]
Wait a minute, how old are these capacitors? Where did you get the others?
Part of the problem could be the age of these things. Electrolytic capacitors go bad over time.
1x 35v 10,000uf Epcos and 1x 50v 10,000uf nichicon were brand new - cooked
1x 35v 10,000uf Pye was old - cooked
1x 40v 3300uf Blue nichicon from scope is old - working (no temperature rise)
3x 40v 2200uf Fraco in parallel (old) - working (4-5 degrees C rise)
elko frako gpf 2200uf 40v 20x40 in aus Resonatoren auf AliExpress.com | Alibaba Group
These Fraco ones were salvaged from an old 1980's transmitter. I put them on yesterday and had them running for about 6-7 hours. They got a little warm after about an hour and continued working with no further temperature rise.
All the other 10,000uf ones only worked for under 20 miuntes. They all got hot and begin to either leak electrolyte or top bulged up. The epcos even had its plastic top partially cap blow off.
1x 35v 10,000uf Pye was old - cooked
1x 40v 3300uf Blue nichicon from scope is old - working (no temperature rise)
3x 40v 2200uf Fraco in parallel (old) - working (4-5 degrees C rise)
elko frako gpf 2200uf 40v 20x40 in aus Resonatoren auf AliExpress.com | Alibaba Group
These Fraco ones were salvaged from an old 1980's transmitter. I put them on yesterday and had them running for about 6-7 hours. They got a little warm after about an hour and continued working with no further temperature rise.
All the other 10,000uf ones only worked for under 20 miuntes. They all got hot and begin to either leak electrolyte or top bulged up. The epcos even had its plastic top partially cap blow off.
I´m starting to smell a troll.
Typical jumping around the main issue, vague demands, asking answers without disclosing data, mentioning problems only seen by him, etc.
Who cares that "the project is not ready yet"?
The project IS built TODAY, one way or another, and it has a problem TODAY, so if you want answers hand draw it as it is NOW.
Or wait until it´s finished to ask.
You are wasting our time.
Typical jumping around the main issue, vague demands, asking answers without disclosing data, mentioning problems only seen by him, etc.
Who cares that "the project is not ready yet"?
The project IS built TODAY, one way or another, and it has a problem TODAY, so if you want answers hand draw it as it is NOW.
Or wait until it´s finished to ask.
You are wasting our time.
Some people are happy to provide free consultancy for commercial audio; some are not. Similarly, some are happy to help a student cheat; some are not. Good manners and professional/academic integrity demand that the questioner should say. This is a DIY site, so people wanting non-DIY help should say so.
Something strange going on. Coupling caps should not be getting hot. We are not getting the whole story.
Charlie, when you are testing and the 10mF runs hot, how are you testing?
From your posts I have collected the following data:
Output voltage of the follower is 7.5Vrms. Load is 2 Ohms. Are you testing with 20Hz?
If you are testing that way there is a chance that the current in the output capacitor is too high. That would explain why the smaller cap survives. It limits the current because of its higher impedance at 20 Hz.
Paralleling two caps would mean the caps share the current.
Please be aware that there is a maximum permissible value for the current in a cap. The data sheet calls it max ripple current. The data sheet gives you a value for 100 or 120 Hz because that is what the cap will see when used as a smoothing cap in a classic power supply.
The max ripple current at 20 Hz is much lower. Maybe another forum member can explain why that is so.
Gesendet von iPad mit Tapatalk
From your posts I have collected the following data:
Output voltage of the follower is 7.5Vrms. Load is 2 Ohms. Are you testing with 20Hz?
If you are testing that way there is a chance that the current in the output capacitor is too high. That would explain why the smaller cap survives. It limits the current because of its higher impedance at 20 Hz.
Paralleling two caps would mean the caps share the current.
Please be aware that there is a maximum permissible value for the current in a cap. The data sheet calls it max ripple current. The data sheet gives you a value for 100 or 120 Hz because that is what the cap will see when used as a smoothing cap in a classic power supply.
The max ripple current at 20 Hz is much lower. Maybe another forum member can explain why that is so.
Gesendet von iPad mit Tapatalk
- Status
- Not open for further replies.