| tryalx |
Hi,
I would like to test the MOSFET amplifier designed by M. Renardson, from
http://www.angelfire.com/ab3/mjramp/amp7.html.
http://www.angelfire.com/ab3/mjramp/construct.html
http://www.angelfire.com/ab3/mjramp/simple5.html
Did anybody experiment with it? How does it sound?
I would like to ask some advice regarding the implications of increasing the voltage from 60V to 90V and eventually adding a second pair of MOSFETS for extra output power.
Besides choosing capacitors of higher voltage for the power supply and for the 4700uF output (100V instead of 63V), do we also need to increase the output capacitor from 4700uF to 6800uf or higher in order to accomodate the extra power?
If we increase the value of the output capacitor, since it is part of the feedback loop, do we also need to adjust the values of any other components ?
Do we need to increase the value of the power supply capacitors from 12000uf to let say 20000uf?
To add a second pair of MOSFETS, I assume that I would need to add another pair of resistors (330R) and 12V zener diodes to drive these MOSFETS. Is this correct?
Any other advice is welcome :-)
Thanks! |
|
|
| Mooly |
Hi,
As a huge fan of Lateral FET's I have to add my two penneth :D
Easy bits first, yes just add an extra pair of 330 ohms and parallel the outputs. You could always use double die MOSFET's instead such as BUZ900D and BUZ905D (TO3 Package). The 12 volt zeners, it's not usual to use back to back zeners, normally just a zener and series In4148 if you want to include them at all !! I normally do not.
The value of the output cap is unrelated to power output.
I imagine it sounds -- well -- very nice :)
Any reason why this particular design ?
Regards Karl |
|
|
| tryalx |
Hi Karl,
Thank you for your feedback.
I looked at this design because of its simplicity, yet very low noise/distortions and good linearity.
I would like to build a multi-channel home theater amplifier, and eventually pair it with one of these speakers: the Gallo reference speakers (http://www.roundsound.com/reference-speakers.htm) or the Druid MK4 (http://www.zuaudio.com/druid_2.asp).
I think that increasing the voltage I may be able to get at least 80W of power, which should be enough. Maybe I would need a more powerful amplifier for powering the woofers' second second coil of the Gallo speakers.
What do you think about a design that uses symetrical voltage? Could we get similar performances using only 7 transistors? Such a design may have an advantage that the capacitors will need lower voltage, and I could also get more power, especially for the woofer speakers.
Regards,
Alexandru |
|
|
| tryalx |
Hi Karl,
There is something more... I looked again at the schematics, and the zeners are after the 330 ohms resitors, which makes it more difficult to add another pair of transistors, if we keep the zeners. Therefore your idea to use the BUZ900D and BUZ905D MOSFETS would be better.
Do you know where to buy a matched pair of BUZ900D and BUZ905D?
Thanks,
Alexandru |
|
|
| lineup |
A general comment on Mike J. Renardson.
The guy obviously knows a lot about amplifier designs.
As a matter of facts, I was browsing his website a couple of weeks ago
and downloaded some of his best stuff .......
I would say his later amplifiers, like MJR-6, MJR-7 are 'his best'.
( Because usually people progress, no matter from which starting level )
He has got a lot of interesting knowledge Articles, besides amp projects:| quote: | DESIGNING AUDIO POWER AMPLIFIERS
http://www.angelfire.com/ab3/mjramp/designing.html
# EFFECT OF FEEDBACK ON HARMONICS. Updated 13-May-2006.
It is well known that a square-law amplifier adds only second harmonics with no feedback, but many high order harmonics with feedback applied, but this is almost entirely misleading, as explained here.
# INPUT STAGE DISTORTION.
Seven of the most common one and two transistor input stages are compared by calculating their levels of intermodulation distortion under identical stage output current levels.
# SLEW RATE AND T.I.D. (Updated 12 June 2006)
The maximum slew rate requirement and the design techniques for low t.i.d. are investigated. Now includes measurement results for maximum CD slew rate.
# MEASURING DISTORTION.
To test our low distortion designs an alternative to buying expensive distortion measuring instruments is to use the bridge-nulling method in which the input and output are directly compared.
# PHASE INTERMODULATION.
# Explanation, reduction and measurement of phase intermodulation distortion.
# COMMON-MODE DISTORTION.
Common-mode distortion is examined, a test method described, and the performance of typical input stage transistors tested and compared.
# MEASURING CAPACITOR DISTORTION. (Updated 22-Apr-2006)
Capacitors are known to add a little low order harmonic distortion, but does transient testing reveal something a lot worse than this?
# ERROR FEEDBACK AND NESTED LOOPS.
An examination of how effective these methods can be compared to overall negative feedback.
# SYMMETRY. (Updated 01-May-2006)
Symmetry can be good or bad, but just aiming for maximum circuit symmetry may add unnecessary complexity with no real benefit. It is possible in principle to achieve zero distortion in some circuits with correct use of symmetry.
DISTORTION - HOW LOW CAN YOU GO?
Is there any limit to how far power amplifier distortion can be reduced?
|
The startpage of his website, with amplifier project and articles
is:
http://www.angelfire.com/ab3/mjramp/index.html
Regars Lineup |
|
|
| nigelwright7557 |
Looks very similar to an old Maplin 75WRMS amplifier.
There seems to be quite a few more components than is absolutely necesary in this design. Typical of a designer who has been around a while !
I prefer to stick to simple designs. I dont use zeners on the MOSFET gates. I lose the output inductor. I keep the zobel network tho. I would only use one low pass filter on the input.
I probably wouldnt use the 1000uF on the output stage. |
|
|
| Mooly |
Hi
Have you checked out my design on this forum. :) |
|
|
| Mooly |
I saw you were online. I haven't figured out how to link to other threads -- it never works for me !!!
Look in the solid state forum, it's called "My MOSFET Amplifier designed for music". Have a read :) |
|
|
| h_a |
| quote: | | Implication 60V to 90V |
Now that's a drastic change. You would at least need to check the biasing and check the SOAR of all transistors. I don't have the time at the moment to have a more closer look to the schematic, but from experience I doubt that you can increase rail voltage by that amount without reworking the circuit a bit.
Have fun, Hannes |
|
|
| tryalx |
Hi Mooly,
I have read through your "My MOSFET Amplifier designed for music" topic, and there was a lot of excitement around.
There were several comments and possibly some changes in the original design and board layout. Can you please post the latest schematic and PCB layout?
You have also given a lot of information about measurements and components types. Do you have any website with all these details summarized? Otherwise, would it be possible to have the final conclusions compacted in one post? This would be very helpful!
Thank you also for the link for the components!
Cheers,
Alexandru |
|
|
| tryalx |
| Mooly, this is the schematic I have found in your thread. Is this the latest? |
|
|
| tryalx |
| And here is the PCB designed by Alex mm, is this the latest version? |
|
|
| Mooly |
Hi,
Sorry for not replying sooner --- for some reason not notified of new posts on this thread -- Gremlins ;)
The circuit is unchanged component wise apart from a 10 ohm 2 watt resistor across the output inductor. I kept missing it off the CAD program when I copied it over from the hand drawn originals.
Alex looks to have done a fantastic job with the PCB layout. I have "visually" checked it over and it looks fine. As you will have gathered the originals were very much a one off.
A website :D :D I have only had a PC at home for 18 months or so. It was hard enough figuring out how to send an E-Mail.
Final conclusions, thats easy,
BUILD IT. LISTEN TO IT. REDISCOVER YOUR MUSIC COLLECTION.
This amp does truly sound wonderful, you can close your eyes and be transported away with the music. You will find you are not listening to the amp, the music shines through. And that is the problem for me in that I wonder now how to improve on it subjectively, after 4.5 years it still has all the magic, nothing I have heard has made me want to change it.
If you do intend building it I can give any advice etc via this forum.
Regards Karl |
|
|
| tryalx |
Hi Hannes,
Thank you for your feedback, it is a good point. So I guess, first I will build the amplifier as stated on the schema and messure the B-E voltages of the transistors. The transistors will handle a 90-100 volt rail voltage. After I increase the voltage (maybe in steps) I will messure these voltages and if needed adjust the values of the resistors. I am not sure whether this is the right way of doing it, but I can try :)
Regards Alexandru |
|
|
| nigelwright7557 |
| quote: | Originally posted by tryalx
Hi Hannes,
Thank you for your feedback, it is a good point. So I guess, first I will build the amplifier as stated on the schema and messure the B-E voltages of the transistors. The transistors will handle a 90-100 volt rail voltage. After I increase the voltage (maybe in steps) I will messure these voltages and if needed adjust the values of the resistors. I am not sure whether this is the right way of doing it, but I can try :)
Regards Alexandru |
The ratios of the resistors will stay the same so you might not need to adjust any.
Just watch the pwoer rating of the resistors at higher voltages.
Check the transistors in the driver stage are correctly rated for higher voltage. |
|
|
| tryalx |
Hi Karl, everybody,
| quote: | | The circuit is unchanged component wise apart from a 10 ohm 2 watt resistor across the output inductor. I kept missing it off the CAD program when I copied it over from the hand drawn originals. |
I am not sure I have understood correctly. Are you mentioning that the resistor R25 which is a 10 ohm should be 2Watt, or is there another resistor missing from the schema?
So apparently there are at least two schemas that are intriguing, which is yours and also the MJR-7 of Mike Renardson. I may also try another one from AmpsLab: http://www.ampslab.com/lm60.htm or http://www.ampslab.com/lm60_schematics.htm
Before ordering the components, my I ask you a question about your choise of transistors. The schema contains 2N5401 and 2N5551, instead of BCxxx or BDxxx series (BC549 or BD139, BD140) transistors, or 2SCxxxx or 2SAxxxx series (2SC2547E, 2SC2911, 2SA1209) transistors found in the MJR-7 schema. Do the 2Nxxxx transistors have better audio properties?
If I will use two of the final MOSFETS in paralel or a BUZ900D/BUZ905D or equivalent, for more power into 4 ohm do we need to replace the transistors Q6 and Q7 with a medium power version? Which transistors should we use in that case for Q6 and Q7?
Do you have any suggestions on what type of capacitors to use, ceramic, mica, polyester, polypropylene? Which one did you use?
Yet another question, maybe for everybody: do we need to pair any of the transistors? I assume that using transistors in pairs with similar gains would improve the sound, wouldn't it?
Does anybody has a simple circuit that I could use to pair the transistors (Biploar an MOSFETS) by measuring the gain (output/input)?
Is there any website where we can buy paired transistors?
Thanks, Alexandru |
|
|
| tryalx |
| quote: | The ratios of the resistors will stay the same so you might not need to adjust any.
Just watch the pwoer rating of the resistors at higher voltages.
Check the transistors in the driver stage are correctly rated for higher voltage. |
Thank you for your feedback I appreciate all the input I get.
Regards, Alexandru |
|
|
| tryalx |
I have noticed two trends regarding the smoothing capacitors from the power supply:
1. Use several lower value (e.g. 1000uf or 2200uf) in parallel to yield the desired capacity and current, let's say (10000-20000uf)
2. Use one or two of 6700uf with very high ripple effect for each +- rail.
Using one/two capacitors allows to build a more compact PSU, but what about filtering and hummm noise?
From experience what is the best approach? What is the trend in professional audio equipment?
Thanks, Alexandru |
|
|
| lineup |
| quote: | Originally posted by Mooly
I saw you were online. I haven't figured out how to link to other threads -- it never works for me !!!
Look in the solid state forum, it's called "My MOSFET Amplifier designed for music". Have a read :)
A website :D :D I have only had a PC at home for 18 months or so. It was hard enough figuring out how to send an E-Mail.
Final conclusions, thats easy,
BUILD IT. LISTEN TO IT. REDISCOVER YOUR MUSIC COLLECTION.
This amp does truly sound wonderful, you can close your eyes and be transported away with the music. You will find you are not listening to the amp, the music shines through. And that is the problem for me in that I wonder now how to improve on it subjectively, after 4.5 years it still has all the magic, nothing I have heard has made me want to change it.
If you do intend building it I can give any advice etc via this forum.
Regards Karl |
Mooly, my diyaudio fellow friend, somewhere in this world.'
I have followed The thread with your MOSFET amplifier
(only posted one post, about GATE Resistor values, compared to Nelson Pass ).
I think your amplifer deserves a bit more attention.
And this is what you want to share with us.
I will put your topic alive again.
Because several diyaudio builders may discover something Very Good.
Mooly: My MOSFET amplifier designed for music
:)
Lineup - your friendly man in Sweden
------------------------------
------------------------------
PS.
Get yourself a free website.
I use www.freehostia.com
It is a easy to sign up for free.
You can have upto 10 different 'websites URL' from one account.
You can put your description + schematic images into there.
"My MOSFET Amplifier designed for music"
If you need some web advices from me, i may be able to advice you well.
I am a PHP Web programmer since like 6-.7 years back ....
besides a good knowledge audio man.
For example I have designed this startpage 100% myself
setting up this startpage for Lineup Audio:
http://lineupaudio.freehostia.com
DS. |
|
|
| Mooly |
Hello Lineup,
Thank you for your kind comments :) and putting the link in.
Regards Karl |
|
|
| lineup |
| quote: | Originally posted by Mooly
Hello Lineup,
Thank you for your kind comments :) and putting the link in.
Regards Karl |
:) thanks for saying, tank u, you English Gentle Man
... sometimes, I think me actually even deserves one Muchos Gratias |
|
|
| Mooly |
Hello Tryalx,
Why 2N5551 and 2N5401 ? They are just highly suitable really -- the fact I can get them locally for around £0.04 each has nothing at all to do with it :Pinoc: -- no, they are very suitable with a 0.6 amp current rating and 160 volt Vce. Do transistors sound different ? such as the 2SA and 2SD types etc.
Erm, probably going to ruffle a few feathers now. Different transistors have different characteristics as we know, junction capacitance, the actual size of the wafer of silicon etc. These characteristics do influence the way a circuit behaves. There is no magic audio property however, the laws of physics determine transistor operation. Now I class this amp as a low frequency device, and transistor type (general purpose small signal types) should have very little bearing on performance.
I have never seen so called matched pairs of laterals offered for sale, however laterals are acknowledged to be "more complementary" than most so called complementary Bjts. The price of laterals as well makes it a bit impractical to to have a lot to hand to sort through. The very nature of laterals as well means you can parrallel them without the problem of unequal current sharing that would happen with Bjts. It's not a problem.
The drivers will be fine as well with two pairs of outputs. If you use the BUZxxx remember these are TO3 package -- not as easy to mount perhaps.
Hum and noise is mostly down to correct layout and grounding.
Low rail ripple is a major concern in a Class A amp, much less so in Class B or AB. Oversize caps cause massive charging currents in the bridge rectifier, wiring and transformer windings due to the very short angle of conduction of the bridge--- it has to put all the energy taken from the reservoir caps back, and in a very short time. Remember the bridge conducts only when the incoming AC exceeds the rail voltage - for most of the time the bridge is non conducting.
Were you thinking of making your own PCB's.
Here is the missing 10 ohm. |
|
|
| homemodder |
Hi Karl
Yes you ruffelled some feathers allright,
Transistors influence the performance quite a lot. The magic audio property is low cob, high early voltage,low nose for input, the use of specific transistor accordingly to design operating Ic parameters .
Since this thread was initially about renardsons amp, i suggest you go read up on those pages lineup has posted.
You wanted to know how to improve your amp, a good start would be better trannies. Better types are more expensive offcourse but would inhance performance without more circuit complexity. Ballance the cost of better parts versus complexity factor. Look at renardsons input stage, the cfp, try it, but first read up on Input stage distortion by renardson to understand why. There are trannies that far outperform the 2n devices you use in all the above mentioned respects and yes they are all 2sc, 2sa types. As for the mosfets look at the attached pdf.
I have built his mjr6 as prototype which is very musical indeed, i just dont like the single supply rail for the simple reason of the introduction of caps on the output although his circuit can be taillored for dual supply.
I like your design, it should have simillar characterestic sound ableit less warm.
Alex |
|
|
| homemodder |
Hi Karl
Yes you ruffelled some feathers allright,
Transistors influence the performance quite a lot. The magic audio property is low cob, high early voltage,low nose for input, the use of specific transistor accordingly to design operating Ic parameters .
Since this thread was initially about renardsons amp, i suggest you go read up on those pages lineup has posted.
You wanted to know how to improve your amp, a good start would be better trannies. Better types are more expensive offcourse but would inhance performance without more circuit complexity. Ballance the cost of better parts versus complexity factor. Look at renardsons input stage, the cfp, try it, but first read up on Input stage distortion by renardson to understand why. There are trannies that far outperform the 2n devices you use in all the above mentioned respects and yes they are all 2sc, 2sa types. As for the mosfets look at the AN1645 application notes by national.
I have built his mjr6 as prototype which is very musical indeed, i just dont like the single supply rail for the simple reason of the introduction of caps on the output although his circuit can be taillored for dual supply.
I like your design, it should have simillar characterestic sound ableit less warm.
Alex |
|
|
| Mooly |
Hello Alex,
It's good to have discussion on these things,
| quote: | Originally posted by homemodder
Yes you ruffelled some feathers allright,
Transistors influence the performance quite a lot. The magic audio property is low cob, high early voltage,low nose for input, the use of specific transistor accordingly to design operating Ic parameters .
Alex |
:) All you have mentioned are, 1. Identifiable. 2. Measurable. 3. Not magic.
My personal feeling is that substituting the transistors would not make an audible difference. Personal opinion as I say :)
In what way would you expect the performance of the amp to differ if other devices were used. I accept the noise figure of the 2N5401 as an input device could be bettered, but I have long since concluded low noise (hiss) is way down the list of what makes an amp sound good. In any case noise is more a function of the various impedances in the circuit-- the feedback network, input resistance etc. You are probably looking at under 0.5 dB improvement in overall SNR by going for a device with a lower noise figure.
I genuinely believe my circuit is pretty well immune to differences in the semiconductors used (within reason of course), the exception being the TLO71 which isn't in the audio path anyway.
I will certainly have a read at the articles you mention and the National Application note.
Thanks for the interest and comments,
Regards Karl |
|
|
| homemodder |
hi Karl
True its not magic, but its part of what makes some amps sound magically good and it takes months, sometimes years of research and subjective listening to get right. Details are very important, and thats how guys like H Dean, N Pass, J curl make very fine amps.
Concerning Audible differences maybe you should play a bit around with diffrent trannies and look at harmonic spectrums you get. Maybe just with a simple circuit. It is audible. As an example a input stage anomoly because of being at a very low amplitude might not be audible yes, but keep in mind that this signal is going to be amplified with that same anomoly and it will then most certainly be audible, the same like low noise resistors, why use them??? / simple answer because they have lower noise and its audible. Using improved trannies will most certainly drop your THD figures somewhat but may also change your harmonic spectrum for the better. I think we all know by now THD gives little bearing on what a amp sounds like and what matters is the spectrum of this. Harmonic spectrum can be manipulated by not only topology but also device selection. When you combine low THD with favourable spectrum thats when you start getting into very good sounding amps, Renardson has done it well.;) After youve read Renardson youll understand why i suggested the cfp and note how much effort he put into chosing the right trannies to make up that cfp. All small details. The parts you have used are by no means bad at all, ive seen bad selections here on this forum but its my two cents on how to improve your amp. This single transitor input topology can sound wonderful, is one of my favourite topologies and since i like to keep things as simple as possible my reasoning for better trannies. You may find that you can do better than .5 db improvements and definetly audible. Biggest improvements on subjective side comes from wider more precise sounstaging, black backgrounds and fine detail.
Alex |
|
|
| Lumba Ogir |
| quote: | | The magic audio property is low cob, high early voltage,low nose for input, the use of specific transistor accordingly to design operating Ic parameters . |
Hannes, are you listening? |
|
|
| Mooly |
Hi Alex,
I do have an open mind on what you say :) I agree that it's the spectrum of any harmonics that make can amp sound better one over another. I found reverting back to the single ended input configuration to make the most difference to me.
Your last sentence. EVERYTHING you say-- this amp delivers--
Now thats magic ;) |
|
|
| tryalx |
Hi Karl, guys
Thanks for the updated schema and for your explanations. I am listening and learning :)
I will soon start purchasing the componets. Nikko (http://www.dalbani.co.uk/) has the transistors and also http://www.profusionplc.com/. Yet, I haven't found any US distributor, I was hoping to get cheeper rates. Has anybody any idea where to buy them in US?
Also does anybody know any website for purchasing the heatsinks?
Thanks Alexandru |
|
|
| tryalx |
| quote: | | I have built his mjr6 as prototype which is very musical indeed, i just dont like the single supply rail for the simple reason of the introduction of caps on the output although his circuit can be taillored for dual supply. |
Hi Alex (homemodder),
I am wondering if the caps on the output would increase the harmonic spectrums and so affect in a negative way the clarity of the sound.
Thanks, Alexandru |
|
|
| homemodder |
Hi Alexandru
Interesting name you have, my full name is alexandre, thats the portugeuse way.
This caps thing is a very interesting subject. I hope you have read renardsons views on this, he explains the effects that caps have on the harmonic spectrum. The effects it has Im pretty sure is correct as you will notice other investigators have delved even deeper into this and have same conclusions. Look at the refence he gives, thats a very thorough investigation done. As the caps would produce very low lower order harmonic distortions which to the ear is pleasant sounding it would be difficult to hear but now you said the right word CLARITY. Studies done have shown that low order harmonics tend to mask. I agree with the other investigators, it is audible, for me it is anyway. He claims that differences are so small and presents a blind test of way back of the seventies to reinforce his views but that was nearly 30 years ago, amps and components are so much better now that differences can be heard. I must state that he very cleverly places those caps inside the feedback loop to minimise those effects but i think he is also not convinced of himself as he states to use the best quality caps that you can find. If inaudible why the best caps??? I prefer to not use capacitor coupled outputs.
For parts look here www.bdent.com
Alex |
|
|
| tryalx |
Hi Alex,
I was born in Romania and somehow ended up in the middle of the Pacific, which isn't bad at all :)
Yes, I read again through the information provide by Renardson... I guess I will need to try and see how it sounds. Maybe the clever tricks with the feedback do a good job in reducing the lower order harmonic. I would like to have an amplifier that has a warm and very clear sound allowing to distinguish the fine details in the music fabric.
An amplifier with symmetrical rail voltage would be indeed preferable, and there is Mooly's amplifier.
Thanks, Alexandru |
|
|
| Mooly |
Hello Alex,
Been having a look at M Reynardsons site. Very interesting. I see in his article on earthing around the volume control he came against the same problems I found. I had to smile at his balance control. I did the same (well not quite) as his implementation will cause response and phase differences if the time constants seen by the two channels are not equal. I called mine an "Image center or localisation" control :)
I even went as far as modeling the grounding with two opamp based power amps and analysed it as a DC problem to try and see what really goes on. I deliberately scaled all the supply and ground resistances up to around 10 ohms to make it easier to appreciate.
I shall continue reading.
Regards Karl
P.S. I never mentioned the "balance" control but here it is. It's a multiturn pot. |
|
|
| G.Kleinschmidt |
| quote: | Originally posted by homemodder
hi Karl
True its not magic, but its part of what makes some amps sound magically good and it takes months, sometimes years of research and subjective listening to get right. Details are very important, and thats how guys like H Dean, N Pass, J curl make very fine amps.
Concerning Audible differences maybe you should play a bit around with diffrent trannies and look at harmonic spectrums you get. Maybe just with a simple circuit. It is audible. As an example a input stage anomoly because of being at a very low amplitude might not be audible yes, but keep in mind that this signal is going to be amplified with that same anomoly and it will then most certainly be audible, the same like low noise resistors, why use them??? / simple answer because they have lower noise and its audible. Using improved trannies will most certainly drop your THD figures somewhat but may also change your harmonic spectrum for the better. I think we all know by now THD gives little bearing on what a amp sounds like and what matters is the spectrum of this. Harmonic spectrum can be manipulated by not only topology but also device selection. When you combine low THD with favourable spectrum thats when you start getting into very good sounding amps, Renardson has done it well.;) After youve read Renardson youll understand why i suggested the cfp and note how much effort he put into chosing the right trannies to make up that cfp. All small details. The parts you have used are by no means bad at all, ive seen bad selections here on this forum but its my two cents on how to improve your amp. This single transitor input topology can sound wonderful, is one of my favourite topologies and since i like to keep things as simple as possible my reasoning for better trannies. You may find that you can do better than .5 db improvements and definetly audible. Biggest improvements on subjective side comes from wider more precise sounstaging, black backgrounds and fine detail.
Alex |
The importance of the harmonic spectrum is inversely proportional to the THD.
It will not make a jot of difference, sound wise, if the distortion of an amplifier with .001% THD is predominantly second or third, but it can make a significant difference to an amplifier with high THD - such as valve amps. |
|
|
| h_a |
| quote: | | the same like low noise resistors |
The funny point - that is constantly overlooked - is that no resistor in the world can be lower than the thermal noise it produces. And that thermal noise is for usual values much higher than the additional noise introduced in the manufacturing.
What 'low noise' really claims is that the total noise of the resistor is dominated by its thermal noise - and not even higher by crappy manufacturing.
That's the reason every high quality resistor is also 'low noise'.
All the best, Hannes
| quote: | | Hannes, are you listening? |
So what?
You're comparing apples and oranges here as I very well explained you the reasoning for my choice regarding the special amp you're referring to. |
|
|
| homemodder |
Hi Mooly
Indeed profusion keep some good transistors around, i purchase output bjts and now lately have ordered some small signal trannies which i use to purchase from the states there. For lateral fets the 2sa1209 are indeed good. There are others that outperform all those mentioned there mainly from toshiba and sanyo. Theres one thing they dont mention there though, and that you dont find on datasheets and that is the importance of high early voltage. Have a look at 2sa1016k as small signal input transistors and compare to 2n5401, Ive been using them for over 15 years, you wont find anything better on the market.
Alex |
|
|
| Lumba Ogir |
Hi,| quote: | | The importance of the harmonic spectrum is inversely proportional to the THD. |
a confusing statement, I certainly would not put it that way. Apropriate harmonic spectrum is a basic condition for natural sound. Numerous commonly used topologies cause severe departures from it, however some people are careful about the methods to achieve a close approximation.
Electrolytics are not nearly ideal capacitors having high dielectric absorption, small operating frequency range and poor linearity. |
|
|
| Mooly |
Hi Alex and Lumba Ogir,
All this talk of early voltage, transistors getting a cob on -- sorry having a low cob- is there a difference ;) , output inductors in another thread, Space with his flying saucer heatsink- yes really.
You know what. I do something different with my amp-- I use it to listen to and enjoy music :D :D :D |
|
|
| h_a |
| quote: | | I do something different with my amp-- I use it to listen to and enjoy music |
Perfectly well said!
Enjoy, Hannes |
|
|
| tryalx |
| quote: | | The 12 volt zeners, it's not usual to use back to back zeners, normally just a zener and series In4148 if you want to include them at all !! I normally do not. |
In Renardson amp the zener's are included after the 300 ohm resistors while in other schematics they are before the resistors, or are missing.
Is it ok to move these zeners before the 300 ohm resistors. This way I could easily add another pair of transistors and 300 ohm resistors.
Thanks Alexandru |
|
|
| tryalx |
I am trying to find a place where to buy a toroidal transformer 120V primar with 2 X 30V sec. www.qkits.com has them but the shiping costs $118 :xeye:
Who knows another place?
Thanks Alexandru |
|
|
| tryalx |
| quote: | | http://export.farnell.com/jsp/home/...en_EX_DIRECTEXP |
Thanks, I will see how much they charge for the shipping.
I have a question. The primary is 2 X 115V, and since the power line is 110-120 V I assume that I need to connect them in parallel, is that true?
Alexandru |
|
|
| tryalx |
What power supply do I need?
Considering that the the amp provides at ful power a peak voltage of 30V on 4 ohm, this means 7.5 amps peak current. The average current should be 7.5 divided by the square root of 2 = 5.3 amps. Is this correct?
Considering that the rail to rail voltage is 85V than the average power consumed is 85 * 5.3 = 400.5. Does this mean that I would need a power supply that has at least 400 watt of sustained power for one mono amplifier? Is my computation correct?
I am trying to order a 30 x 30V or 33 x 33V toroidal transformer and I would appreciate your input.
Thanks, Alexandru |
|
|
| nigelwright7557 |
| quote: | Originally posted by tryalx
What power supply do I need?
Considering that the the amp provides at ful power a peak voltage of 30V on 4 ohm, this means 7.5 amps peak current. The average current should be 7.5 divided by the square root of 2 = 5.3 amps. Is this correct?
Considering that the rail to rail voltage is 85V than the average power consumed is 85 * 5.3 = 400.5. Does this mean that I would need a power supply that has at least 400 watt of sustained power for one mono amplifier? Is my computation correct?
I am trying to order a 30 x 30V or 33 x 33V toroidal transformer and I would appreciate your input.
Thanks, Alexandru |
You have to be careful with supply lines on amps.
While its easy to think 45 volts is OK for an amp it doesnt take into consideration transients in the music. Music has a huge dynmaic range.
I was shocked just how quickly the transients were clipped on my amp. So in the my next design I went for a 55 volt power line.
This also fits in well with 63 volt electrolytics which are in abundance.
I have even seen on this forum people using upto 130 volt power lines !!!!!
|
|
|
| tryalx |
| quote: | | I was shocked just how quickly the transients were clipped on my amp. So in the my next design I went for a 55 volt power line. |
That is a good thought, I can go with higher voltage.
What about the power of the transformer do you think my computation from post #47 is correct. I can redo it for a higher voltage.
Thanks, Alexandru |
|
|
| nigelwright7557 |
| quote: | Originally posted by tryalx
That is a good thought, I can go with higher voltage.
What about the power of the transformer do you think my computation from post #47 is correct. I can redo it for a higher voltage.
Thanks, Alexandru |
I have read from other sources that you need work out what power goes into the speaker, then double it for power lost in the output transistors then add another 50% so the transformer isnt cooking on full load over long periods.
So for a 1000VA transformer you should get out about 330WRMS |
|
|
| tryalx |
| quote: | | I have read from other sources that you need work out what power goes into the speaker, then double it for power lost in the output transistors then add another 50% so the transformer isnt cooking on full load over long periods. |
That's a more straightforward way to figure out the transformer's power.
I won't be listening in my room at very high power levels, but as you mentioned the dynamics may jump high at moments. It looks like I may need a 500VA transformer to occasionally allow a max of 200W RMS burst of power into 4 ohm (also taking into account an 85% transformer efficiency).
Thanks, Alexandru |
|
|
| Mooly |
Hello Alexandru,
Have a look here www.signaltransfer.freeuk.com/powerout.htm it's quite well written and easy to follow.
The gate protection diodes ( if the amp is for your own use and you are not going to be shorting the outputs etc) I personally would not fit them. It's different in a commercial product that has to be as abuse proof as possible. If you move the zeners, there would be problems, the other side of the resistors is at a lower impedance, so more, possibly destructive current could be drawn + if they were triggered into conduction and the "characteristics" if you can call it that of the limiting action would be severely compromised.
Your transformer-- always use a bulb (mains filament say 60 to 100 watt) in series with the mains for testing, and not just for the tranny wiring but the whole amp as well.
Oh yes the primary windings, connect in parrallel. Which way round you ask --- use the bulb method, the correct way will draw no (minimal) current, the incorrect way and the bulb will be at high brightness. Now imagine trying that without the bulb :hot: |
|
|
| Mooly |
Forgot --- power output. Do you know the efficiency of your speakers in dB/WATT. The Forum Web page rearranges things, arghhh, first column is a 92 db speaker 2nd columnn an 82 db speaker.
Remember that doubling the power in the speaker increases the level by 3 db. So a 92 dB sensitivity speaker will give,
92 db/watt speaker.........and........an 82 db/watt speaker
92db for 1 watt rms..............82db for 1 watt rms
95db for 2 watt rms..............85db for 2 watt rms
98db for 4 watt rms..............88db for 4 watt rms
101db for 8 watt rms..............91db for 8 watt rms
104db for 16 watt rms..............94db for 16 watt rms
107db for 32 watt rms..............97db for 32 watt rms
110db for 64 watt rms..............100db for 64 watt rms
113db for 128 watt rms..............103db for 128 watt rms
The 82 db speaker is a real example -- Celestion SL6 -- used to use these a few years ago.
Speaker sensitivity is far more important than power. The first example is "louder" on 4 watts than the other is on 32 watts.
Even then it's not that straightfoward as a 3 db increase in level is barely perceptible to many. To get that extra 3 db which may not even be noticable adds hugely to the cost /size of everything.
I always ask this :) DO YOU HAVE A 'SCOPE. |
|
|
| tryalx |
Hi Karl,
That's a very nice example, thank you, you saved me from searching through my electricity books from highschool .:D
Ok, so the gate protection diodes are not really critical in my case, that's great.
The trick with the bulb is also nice, I will remeber it!
| quote: | | Do you know the efficiency of your speakers in dB/WATT. |
I looked at different speakers, and probably I will settle on these one, once I get some more cash:
http://www.definitivetech.com/louds...supertower.html
These speakers got some awards and also have an efficiency of 93dB, which isn't bad. Your dB comparision table is very helpful.
| quote: | | I always ask this DO YOU HAVE A 'SCOPE. |
I looked a while ago for a scope or eventually a pc scope such that I can apply some fourier to see the spectrum, and also to check the distortion level. Do you have any suggestions?
So will read the article and make up my mind about the PSU.
Thanks again,
Alexandru |
|
|
| Mooly |
Hi Alex,
Nice speakers :) I think a MOSFET amp would be great with these.
Oscilloscopes, get real one :D you don't need anything too extreme for audio, even a 20Mhz dual trace is fine. Make sure it has a good bright trace preferably one where the CRT has a high final anode voltage -- 8 to 12 kv or so as these are much better than ones with around 1 to 2 kv final anode volts - much brighter and sharper. |
|
|
| Mooly |
Hi Alex,
Afraid not, I use a 100Mhz 'scope. This is dual trace model together with a delayed sweep timebase, had it about 14 years now.
The thing with an analogue scope is it's all very real, there are no digital artefacts to confuse things, what you see on the screen is very real.
I am sure some of the modern LCD display types are very good, and they have features that would never be available on an analogue one.
Anyone else any views on this ?
Don't know what to advise really :) |
|
|
| tryalx |
Hi Karl,
I have found an article about digital oscilloscopes here: http://www.picotech.com/application...e_tutorial.html
They suggest that the resolution of a digital oscilloscope should preferably be higher than 8bit.
The OWON ones are 8 bit so I will do some more search around.
Thanks, Alexandru |
|
|
| Mooly |
Looks interesting- will have a read later.
A lot depends on how much you want to pay. Audio is all low frequency stuff really, but my feeling is still to go with analogue. Unless your paying big bucks, digital will not offer the bandwidth I suspect. Viewing 20Khz is one thing, a stability problem in the Mhz range something else.
It may be worth you posting a new thread on 'scopes asking for recommendations. Ask yourself this though. "Will the model I am looking at display a 1 Mhz squarewave accurately" what about at 10 Mhz. That's a walk in the park for a good analogue 'scope.
Think of "Digital TV" & LCD monitors etc and the effects of pixalation. Then think of the resolution you can get with an old monochrome CRT, no contest !! |
|
|
| Mooly |
| Your link to the tutorial. Figure 2b and 2c. I can relate to this, the colour burst signal of a composite video waveform. In the UK our (soon to be switched off --- yes yes yes :) ) analogue TV service uses the PAL system and the burst is at 4.433 Mhz. Is the NTSC 3.58 Mhz ? from memory. Anyway the representation shows "digital artefacts" even at this frequency, now imagine expanding the trace so one cycle fills the screen. It should be a nice clean sinewave, but you can see it "stepped" by the sampling of the 'scope. |
|
|
| tryalx |
Yes, I will need to do some more research. If it is a digital one I will need at least 12 bit resolution and probably a bandwidth of at least 250 MegaSamples/sec for 10MHz. Maybe I can find something on ebay.
Thanks. |
|
|
| Mooly |
Hi Alex,
If you can get a good Analogue 'scope on Ebay, personally I would go for that.
I know the DSO's look attractive but for seeing microscopic detail and not wondering whether what you are seeing is caused by the effects of the sampling go for analogue.
That's my personal opinion though. Try and get some other opinions perhaps on this Forum. |
|
|
|
