This maybe nothing. But I don't think I like it so much the input point being between V+ in and V-in.
If you follow my suggestion to add an input resistor, why don't you lay it parallel to R1, to the right?
Also add a film cap between V+ and V-, something like 1uF. That is besides the other bypasses to ground.
Create a star ground point. That might cure the noise problems. The input and output grounds should be soldered to it. Basic things that always work.
If you follow my suggestion to add an input resistor, why don't you lay it parallel to R1, to the right?
Also add a film cap between V+ and V-, something like 1uF. That is besides the other bypasses to ground.
Create a star ground point. That might cure the noise problems. The input and output grounds should be soldered to it. Basic things that always work.
peranders said:
Bypass Vcc to Gnd, Vee to Gnd and Vcc to Vee.
With respect to groundplanes -- I used this technique in my low noise instrument amplifier -- all of the signal grounds are tied to a central point -- I pour copper on the board (well, not literally, but in software ! ) on a fictitious network -- and tie this copper to the signal ground at one central point. All of the bypass caps are attached to this fictitious ground plane.
Guido Tent has a really excellent discussion of grounding, decoupling, component layout and bypass. http://www.tentlabs.com/Products/cdupgrade/shunt/assets/Supply_decoupling.pdf
I would rather not post my PCB, as I don't feel like too much negative feedback today... but I do in general try to keep input output adn power ground seperated in a sense by varying track thicknesses... never had a grounding problem, save for the first night I built a GC.. also have a thick, thin, thinner wire scheme I like to use for the ground wires connecting these points.
I have just completed the 100nf cap addition, 4 in total, for the 2 channels from V+ and V- to ground... and guess what... I think we solved the problem - I am a noob.
It makes sense now, why bypassing caps on the PSU made such a phenominal change in sound everytime... I will now have to remove bypass caps and start PSU bypassing from scratch to get the sound right...
PSU is LM317/337 running now at 17V as I set it to test the chip preamp boards, it is seperated by less than an inch of wire from the preamp boards...
I have no hum or RF issues as far as I can HEAR.
for testing I used BC547, BC547 and BC639 after all the complaints I saw about my transistor choice....
Think I will try the BD135/139 combo again tonight as the smaller transistors can not keep up with the neutrality of the sound I got from the medium power transistors, especialy with peaks and high input levels I can hear audiable distortion in treble and bass peaks with the BC547
I have just completed the 100nf cap addition, 4 in total, for the 2 channels from V+ and V- to ground... and guess what... I think we solved the problem - I am a noob.
It makes sense now, why bypassing caps on the PSU made such a phenominal change in sound everytime... I will now have to remove bypass caps and start PSU bypassing from scratch to get the sound right...
PSU is LM317/337 running now at 17V as I set it to test the chip preamp boards, it is seperated by less than an inch of wire from the preamp boards...
I have no hum or RF issues as far as I can HEAR.
for testing I used BC547, BC547 and BC639 after all the complaints I saw about my transistor choice....
Think I will try the BD135/139 combo again tonight as the smaller transistors can not keep up with the neutrality of the sound I got from the medium power transistors, especialy with peaks and high input levels I can hear audiable distortion in treble and bass peaks with the BC547
Hi,
I guess the chip inside your To92 devices will be much bigger than ANY of the individual transistors inside an opamp.
Why should you need to go to medium power to achieve peak ability? and neutrality?
There must be something else going on, that the slow, big area, low gain devices are hiding.
I guess the chip inside your To92 devices will be much bigger than ANY of the individual transistors inside an opamp.
Why should you need to go to medium power to achieve peak ability? and neutrality?
There must be something else going on, that the slow, big area, low gain devices are hiding.
So, you think I should look at other low power transistors?
A few others were recommended, but I am haveing a hard time hunting them down localy.
But yeah, its a day and night diffirence between the small and medium transistors I used...in favour of the larger...
I can live with a little dstortion at full blast.. never turn the volume up anyway.... the noise was a diffirent beast alltogether.. all crackily wet sound... not nice..
A few others were recommended, but I am haveing a hard time hunting them down localy.
But yeah, its a day and night diffirence between the small and medium transistors I used...in favour of the larger...
I can live with a little dstortion at full blast.. never turn the volume up anyway.... the noise was a diffirent beast alltogether.. all crackily wet sound... not nice..
Nordic said:
As I said before, medium power transistors seem to bring some benefits to certain applications like this buffer.
You seemed to have confirmed this with your tests, so don't be ashamed to favour them.
In fact I think you might try others, like the MJE200/210 pair, if you can find them. Or the MJE243/253. But the BDs are very good choice.
I am very glad you could solve the noise problem. In any case, the star grounding might be beneficial too.
OK, now that problem one is out of the way, may we please move on to no. two...
I want to use 2 of these buffers in series with the following attenuator config between twem.....
tests so far gave better reslts with it placed before the buffers than between.. especialy takeing a knock on the volume, way above the expected 3db....
I expect an impedance mismatch with my current setup, but am to green to imagine the next step to solveing the problem.
It looks like the buffer is designed to see 100K on input and output.... the attenuator would surely affect this...
I want to use 2 of these buffers in series with the following attenuator config between twem.....
An externally hosted image should be here but it was not working when we last tested it.
tests so far gave better reslts with it placed before the buffers than between.. especialy takeing a knock on the volume, way above the expected 3db....
I expect an impedance mismatch with my current setup, but am to green to imagine the next step to solveing the problem.
It looks like the buffer is designed to see 100K on input and output.... the attenuator would surely affect this...
Nordic said:
I have some suggestions that might improve your supply, some of them from the datasheet itself:
1) Put a smaller cap at the output, like 100uF. See if the audio quality improves. If it does, then change the .680uF caps for .100uF film types.
2) Put those 470uF caps to filter the adjusting pin, in parallel with the trimpot. You will have to put protection diodes there too.
3) See what value the trimpot is and swap it for the closest 1% resistor value you can find. A resistor will not change value, while a trimpot can.
4) Add two series 1ohm/3W resistors between C15/C11 and between C16/C12. That should clean a bit garbage from what's going into the regulators.
Remember that capacitor quality may be better than capacitor quantity. High temp and low ESR seem hints that a cap is a bit better and may improve things for you.
Some may change the diodes too, but others may not think so. If you can (and want) try faster types: fast/slow or Schottky.
Hi Carl, I have only 1 trimpot, (the other died and was replaced with a metal foil resistor like you said)
Also, the board has holes for the series resistors, but I have not bought em yet, as ppayday is 2 days away... they are just jumpered right now, I will try upping the cap in parallel with the resistor from 10uf to 470uf... I just went with 10uf as that is what the datasheet suggests. the output was 10uf, but on the problematic circuit the larger cap was better.
Regulation is pretty good, if I bypass the output relay, I can not even hear when I flick the power switch on and off , while it keeps running off the caps...
Tommorow I want to strip it back to 1x 2200uf before and 10uf after regulators and work my way back up...
I tested the BD135 with the adjusted PCB tonight.. it played MUCH better than the BC547, but after a while went noisy again.. maybe because its only running on 17V rails now.. previous tests were done at over 20V.
Also, the board has holes for the series resistors, but I have not bought em yet, as ppayday is 2 days away... they are just jumpered right now, I will try upping the cap in parallel with the resistor from 10uf to 470uf... I just went with 10uf as that is what the datasheet suggests. the output was 10uf, but on the problematic circuit the larger cap was better.
Regulation is pretty good, if I bypass the output relay, I can not even hear when I flick the power switch on and off , while it keeps running off the caps...
Tommorow I want to strip it back to 1x 2200uf before and 10uf after regulators and work my way back up...
I tested the BD135 with the adjusted PCB tonight.. it played MUCH better than the BC547, but after a while went noisy again.. maybe because its only running on 17V rails now.. previous tests were done at over 20V.
Nordic said:
Don't need to be metal foil. Just precision metal type.
What board is that? It's unusual that it has holes for series resistors. Good!
Yes, but noise will be much be better if you filter the bypass pin. Try it with the pin bypassed and just 47uF to 100uF at the output, and see what happens.
Which output relay? The supply has one? Why? You don't need it. Relays shouldn't be used in regulators, except at the input.
That's the way to do it. You probably won't need more than 2200uF, but you may need another one for a better RC filter.
No, the voltage shouldn't be causing any noise. +/-17v is quite enough. Do they 547s get noisy too? Can you identify it much better when the noise comes on? Looks like a cold solder point to me. Did you go over all solder joints?
Carlos
The PCB I made for the PSU has the holes...
Also had some extra .1uf caps for bypassing, but I think at the time I had too much flux left on the board... as they arced over, so I removed them temporarily...till I can go the shop again...
There is a relay on the output of the preamp to prevent it from playing on on the capacitors... well it does, but the signal is prevented from reaching the amplifier after it.
I suspect I won't get away with les than 2 of the 2200uf caps, they were very cheap, and not nearly as big, tall and nicely finnished as the phillips caps I got last month... will reorder a packet from the UK... only have one left.
I suspect I need to remve the transistor sockets.. very hard to get the BD transistors in there... more likely the problem than cold solder...
The bc547 stayed quiet, no noise, tiny smidgeon of distortion at top levels, but that could have been from the headphones I used to test, they don't like playing too loudly without protesting...
But those transistors just sound a bit too sweet...
My speakers are allready treble and bass heavy due to a less than stellar driver swap... Very nice mellow bass though (the speakers).
STILL NEED HELP WITH QUESTIONS FROM POST 28
Also had some extra .1uf caps for bypassing, but I think at the time I had too much flux left on the board... as they arced over, so I removed them temporarily...till I can go the shop again...
There is a relay on the output of the preamp to prevent it from playing on on the capacitors... well it does, but the signal is prevented from reaching the amplifier after it.
I suspect I won't get away with les than 2 of the 2200uf caps, they were very cheap, and not nearly as big, tall and nicely finnished as the phillips caps I got last month... will reorder a packet from the UK... only have one left.
I suspect I need to remve the transistor sockets.. very hard to get the BD transistors in there... more likely the problem than cold solder...
The bc547 stayed quiet, no noise, tiny smidgeon of distortion at top levels, but that could have been from the headphones I used to test, they don't like playing too loudly without protesting...
But those transistors just sound a bit too sweet...
My speakers are allready treble and bass heavy due to a less than stellar driver swap... Very nice mellow bass though (the speakers).
STILL NEED HELP WITH QUESTIONS FROM POST 28
Nordic said:
Why do you need such configuration?
If you told me you would be using some kind of filtering in between I would understand, but you don't need two of them for that.
You answered your own question. You got the better results with that arrangement because that's the way to do it. No need to look somewhere else. Use just one after that arrangement, which is really a pan control.
Nordic said:
You mean too much solder? Please remove it if you can, except if the copper path is too thin, which might make things worst. Solder flux can be capacitive, so clean it after all it's done. Use quality solder.
What solder pen size are you using? Get yourself a 30W one and be quick to solder. If the pcb is arcing you may be overheating your components, which is very bad.
Remove that relay if you can. In fact, if your power amp has an input capacitor, try removing the one at the preamp's output.
Where you should have a protection relay is at the poweramp's output, to prevent it from sending DC to the speakers if it fails.
What brand are them? If they are Philips they are good, others who knows?
Transistor sockets? That explains it. Of course you must remove them. They are good just for comparisons, then solder the part.
They are quiet because the contact is fine. Are you powering the headphones from the BC547s?
What poweramp are you using? What speakers are those? Home made?
Carlos
I think I got the soldering quite allraait (SA term) was a bit thick at fisrt but I did a sloppy job cleaning the flux off as I am out acetone...
the little ceramics didn't like it one bit, all three arched nice and blue and one carbonised the pcb between it's rails (i cleaned it up real nice, in the sun where I can see). Yep its incredibly hot here today
Just a tad windy.
I don't think output relays are at all that unusual in a preamp design.
The transistors seem to have no problem haveing their outputs shorted to ground, I don't trust my wife to stick to the right switch on/off sequence...I'd rather have it do no harm first.
regarding the placement of the vol/pan control... your opinion is counter intuitive, so you will have to explain more.
Everything I read up to now suggested that putting the controls between the buffers isolates both the output of the source and the input of the amp from impendance chancges as you turn the volume... I more honestly suspect the ratio between the pot and the output stage is just incorrect...
Something I do question though... the buffer has 100k in paralel with input and output... would it make sense to loose the one on the output of the first stage? Otherwise it sees 50k and so would the next stage... (disregarding the volume control now)
I finger tested all the components and its running nice and cool... I suspect no oscilation at this stage.
The speakers used to be 40 Sintronics (local brand from the 70s), one woofer died, so I had to replace both...
Still building up the courage to improve the crossover - which is actualy childishly simple... 1 resistor and a Wima cap as far as I can see...
The new drivers are 70W units (140W peak), but they are not as efficient as the old set, needs more juice to do the same job, and I think anly starts leveling off at much higher volume.. tho old set was much nicer at low levels.. yet the bass from the new drivers are realy nice... I ought to build them their own proper enclosures... but that would put me in an area where I need powerfull amps again... I'd rather want to progress to higher efficiency (maybe fullrange) with low power high quality amp.
Currently I have the Mauro Pennasa, My_Ref C chimpamp (not a gainclone, but a current pump), and the current buffer is a ECC88 valve one I built last year... very nice sound, just not as transparent in the top octaves... ( the buffer - very hard to fault the amp). Current speakers and preamp < my_ref, so they need most work first...
the little ceramics didn't like it one bit, all three arched nice and blue and one carbonised the pcb between it's rails (i cleaned it up real nice, in the sun where I can see). Yep its incredibly hot here today
Just a tad windy.I don't think output relays are at all that unusual in a preamp design.
The transistors seem to have no problem haveing their outputs shorted to ground, I don't trust my wife to stick to the right switch on/off sequence...I'd rather have it do no harm first.
regarding the placement of the vol/pan control... your opinion is counter intuitive, so you will have to explain more.
Everything I read up to now suggested that putting the controls between the buffers isolates both the output of the source and the input of the amp from impendance chancges as you turn the volume... I more honestly suspect the ratio between the pot and the output stage is just incorrect...
Something I do question though... the buffer has 100k in paralel with input and output... would it make sense to loose the one on the output of the first stage? Otherwise it sees 50k and so would the next stage... (disregarding the volume control now)
I finger tested all the components and its running nice and cool... I suspect no oscilation at this stage.
The speakers used to be 40 Sintronics (local brand from the 70s), one woofer died, so I had to replace both...
Still building up the courage to improve the crossover - which is actualy childishly simple... 1 resistor and a Wima cap as far as I can see...
The new drivers are 70W units (140W peak), but they are not as efficient as the old set, needs more juice to do the same job, and I think anly starts leveling off at much higher volume.. tho old set was much nicer at low levels.. yet the bass from the new drivers are realy nice... I ought to build them their own proper enclosures... but that would put me in an area where I need powerfull amps again... I'd rather want to progress to higher efficiency (maybe fullrange) with low power high quality amp.
Currently I have the Mauro Pennasa, My_Ref C chimpamp (not a gainclone, but a current pump), and the current buffer is a ECC88 valve one I built last year... very nice sound, just not as transparent in the top octaves... ( the buffer - very hard to fault the amp). Current speakers and preamp < my_ref, so they need most work first...
Nordic said:I think I got the soldering quite allraait (SA term) was a bit thick at fisrt but I did a sloppy job cleaning the flux off as I am out acetone...
the little ceramics didn't like it one bit, all three arched nice and blue and one carbonised the pcb between it's rails (i cleaned it up real nice, in the sun where I can see). Yep its incredibly hot here today
You are using acetone to clean the pcb? You can do it better with isopropilic alcohol or even water with some solder types.
Of course acetone will affect the components. They may dissolve them!
Also I didn't know you were using ceramic caps. Get rid of them as soon as you can! Use any film type to replace them, polypropylenes if you can find them.
Many things are not unusual. That doesn't mean they are good. If it's a question of on and off, leave the preamp always on.
My philosophy is to have as less contacts as possible to be made in the whole amp line, and the relay is one more you don't really need.
If you want a muting control, put a switch. Except if you are using a remote, which sometimes need some kind of relay switching.
I won't explain much.
The function of a buffer is to provide several things, the main ones being isolating one stage from the other, as well as providing a high impedance input and a low impedance output.
The volume turns would affect the impedance in the following input, if it's high impedance. So why add more active parts in between?
The Penasa amp is very good.
I am not so sure about the speakers and/or drivers you are using. Just one thing: you can't simply replace a speaker for another, except if it's the same sold by the speaker manufacturer.
The box was certainly computed for the original driver. If you change it don't expect too much coherence. They will play, of course, but a well designed system will make a hell of a difference.
Don't "improve" the crossover with different parts or re-design it. You have to know a lot. The capacitor is probably used to filter low frequencies from reaching the tweeter and the resistor to attenuate it. So the main driver is working unfiltered, practically like a full-range.
Qualoity full-range drivers are alright when working at low levels, but they don't go too high. They may need a well designed big box to get there.
Also they are good mostly in mids, so don't expect too many highs or too many lows.
Better get a driver that goes low enough, like an 8", and help it to get higher with a tweeter. Try to find a good design project, with measurements and all.
I noticed I have pretty bad hearing in one ear while building headphone amps, had multiple grommets as a kid - doctor said I have between 40 and 60% scarring on the eardrum... its realy starting to irritate the hell out of me to have to ask my wife to come and listen when I make changes I expect to change the top registers..... with the headphones I can at least still swap them around... although everyone I had swap their earphones around, were quite astounded by the asymetry of their hearing.
For me it is asif one side is louder and have more high frequency information (in fact I think its almost extra sensitive due to compensating for the other side) your mileage may vary...
This is also why the pan control is a neccecity, although not a perfect fix... as the one ear insists on hearing louder than the other.. i.e. my perception has to be diffirent from someone with closer matching ears. Best solution is to combine a little panning with a little body positioning... i.e. sitting dead centre = no good for me...
I had exceptional hearing as a kid I think, and was often praised for perfect pitch during trinity college of london exams (studied classical guitar as a school subject). I have what I call photographic hearing... the equivalent of people being able to remember what they saw... I can remember catching mosquitos just using my hearing at night... it was always just one grab and I had em. Now they laugh at me... but it is like you remember your wife's face, just in hearing.... its annoying and disheartening, but nevertheless my system sounds better month after month as I keep on working on it, and reworking with better quality components.
I only clean the bottom using acetone and some kitchen towels.. fully aware how it eats some plastics..
I can't realy leave it on....
Government has been asked to come clean about Eskom's latest blackouts, as a power shortage grips the country. Full Story...
http://iol.co.za/index.php?set_id=1&click_id=13&art_id=vn20070119113806355C567321
ALso I'm aware of the driver thing, but it was either no speakers or those drivers...
I will build/buy the needed test equipment to set up my speakers nicely, and want to go and have a hearing test done so I can optimise settings for listening alone.
Thats the beauty of DIY, can't buy anything that will just cater for this....
Oh I preffer a good neutral mid to a bass or treble heavy system... Bass gets all its detail from the midrange anyway... just go listen to a subwoofer on its own... almost hard to recognise some songs..
For me it is asif one side is louder and have more high frequency information (in fact I think its almost extra sensitive due to compensating for the other side) your mileage may vary...
This is also why the pan control is a neccecity, although not a perfect fix... as the one ear insists on hearing louder than the other.. i.e. my perception has to be diffirent from someone with closer matching ears. Best solution is to combine a little panning with a little body positioning... i.e. sitting dead centre = no good for me...
I had exceptional hearing as a kid I think, and was often praised for perfect pitch during trinity college of london exams (studied classical guitar as a school subject). I have what I call photographic hearing... the equivalent of people being able to remember what they saw... I can remember catching mosquitos just using my hearing at night... it was always just one grab and I had em. Now they laugh at me... but it is like you remember your wife's face, just in hearing.... its annoying and disheartening, but nevertheless my system sounds better month after month as I keep on working on it, and reworking with better quality components.
I only clean the bottom using acetone and some kitchen towels.. fully aware how it eats some plastics..
I can't realy leave it on....
Government has been asked to come clean about Eskom's latest blackouts, as a power shortage grips the country. Full Story...
http://iol.co.za/index.php?set_id=1&click_id=13&art_id=vn20070119113806355C567321
ALso I'm aware of the driver thing, but it was either no speakers or those drivers...
I will build/buy the needed test equipment to set up my speakers nicely, and want to go and have a hearing test done so I can optimise settings for listening alone.
Thats the beauty of DIY, can't buy anything that will just cater for this....
Oh I preffer a good neutral mid to a bass or treble heavy system... Bass gets all its detail from the midrange anyway... just go listen to a subwoofer on its own... almost hard to recognise some songs..
Sorry to hear about your ear problem.
In any case what you describe is quite common:
1) Women have a better ear than we do. They hear more distortion than we do and hear more highs than we do. But I think we can be more analytical than they are.
2) Most places have "unbalanced acoustics". That is higher on one side than on the other. A balance or pan control usually takes care of that.
So don't feel too bad about it. You are not deaf.
What you may need, perhaps, instead of that control you describe, is something different. Something like a giant "hearing aid" system. You may need a three-band parametric to get there and a frequency generator. That would help balance the frequencies that are diminished in your bad ear. After you set it up you can leave it there. What do you think?
Try alcohol instead of acetone. Afterwards apply some protective varnish. I use a home made one.
Leave speaker design alone. Get a ready-made decent pair by some renowned brand. Pick rubber-surrounded drivers that need no future maintenance. Speaker systems are very hard to design or modify, and you have to know a lot to do it. Not hobby people like us. It's perhaps the hardest piece to design in the audio chain.
In any case what you describe is quite common:
1) Women have a better ear than we do. They hear more distortion than we do and hear more highs than we do. But I think we can be more analytical than they are.
2) Most places have "unbalanced acoustics". That is higher on one side than on the other. A balance or pan control usually takes care of that.
So don't feel too bad about it. You are not deaf.
What you may need, perhaps, instead of that control you describe, is something different. Something like a giant "hearing aid" system. You may need a three-band parametric to get there and a frequency generator. That would help balance the frequencies that are diminished in your bad ear. After you set it up you can leave it there. What do you think?
Try alcohol instead of acetone. Afterwards apply some protective varnish. I use a home made one.
Leave speaker design alone. Get a ready-made decent pair by some renowned brand. Pick rubber-surrounded drivers that need no future maintenance. Speaker systems are very hard to design or modify, and you have to know a lot to do it. Not hobby people like us. It's perhaps the hardest piece to design in the audio chain.
Im pretty good at woodwork (another school subject of mine -don't worry the rest were science, biology and two languages...it wasn't all holiday)
So I expect to be able to follow a pretty well established design with minimal costs above the materials and components... money is always tight, as I am medicaly boarded.... at least I have lotsa time.... way too much, concidering I only sleep about 5 hours a night.
Yep I like the equaliser idea, I have the scematics and PCB plans for a 9 band stereo unit, just have to take the plunge and buy the 18 slider pots at some stage (mondo expensive).
So I expect to be able to follow a pretty well established design with minimal costs above the materials and components... money is always tight, as I am medicaly boarded.... at least I have lotsa time.... way too much, concidering I only sleep about 5 hours a night.
Yep I like the equaliser idea, I have the scematics and PCB plans for a 9 band stereo unit, just have to take the plunge and buy the 18 slider pots at some stage (mondo expensive).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.