An amp I designed went DC this week because I was being a twonker. As luck would have it uses an output transformer. From my experiance an output capacitor of modest performance is as good as high cost transformers or if bass matters better. The performance advantages a capacitor offers are considerable. Anyone who say differently I have to respectfully say you didn't test this. Anyone should be able to see the DC servo we almost must use for a conventional amplifier is not good news. Also below 5 Hz I doubt we reproduce anything useful. How many speakers have seen DC when none suspected ?
For a subwoffer I would change the rules. The drivers I use are £30 yet of good quality. I don't mind them being a fuse. Now if you ask me can I hear a fuse? Yes I can and more so than nearly all the imagined stuff. Can I hear an output cap ? Yes I can and it sounds good. The fuse worries me more. As I don't want DC to ruin my speakers a cap seems the only nice way to do it.
A rule of thumb my brother used. Output cap should be 1/2 the value of the PSU. Thus for 4700 uF output 10 000 uF PSU minimum. If not the amp will sound boomy. The trannsformer as always at least double what the maths suggest ( assume 400 % the single output power if class B is about right ). It seems a reasonable idea. I have that choice to make when my BD140 turn up. Do I use 8 x 4700 uF and 2 x 4700 uF outputs( running in right now ). Or do I use 4 x 22 000 uF in series-paralell (centre tapped PSU) computer grade ? The latter 22 000 uF and the other 37 600 uF. I am attracted to the 22 000 uF as it is large enough by a factor of 2. Being more compact the rails are more compact. This is all for nothing as the Quad 303 creates an artificial 0 V on the PSU. I am making a clone so want Mk1 to be very like the original. Dejan that's why I don't fit the extra outputs as I want it just one step away each time. If not the learning in compromised. Remember 22 000 uF could be in marketing speak said to be 88 000 uF. That is untrue electrically. So it is the real value times the voltage used that is important. CV value which I suspect is a square law? I should use the 8 x 4700 uF for one reason. My transitors will stand 120 V and the caps 90 V ( allowing 10% ). The 22 000 uF only 80 V. Even though I have no plans to go to 100 V it would be a smalll advantage and the heatsinks big enough. That is the transistor capacitance reduces with voltage even if SOA gets worse. Did you realise most transistors only have SOA of 1 amp sine-wave at the supposed voltages they can work at? A FET will be many times better. This would be a standard TO3P device of nominal 15 amp ability. 3 amps is exceptional. I think I calculated a NAD 3020 can offer it's full 15 amp rating into 1 ohm transient. This is becaue the rail voltage it low and 3055/2955 are strong. Give it 20 watts sine-wave 8 R and it won't last long.
My tweeters use a 4.7 uF high grade. That won't chnge. I could DC couple and filter their bespoke amp. No sane engineer would. As mostly the HF is doubful on a 4700 uF cap no sane engineer should couple the 4.7 uF in line with the 4700 uF. Alas in the past we did.
Nige, my protection system, if you remember, will allow as much current as allowed by the emitter resistors in dynamic (transient) mode. Assuming Re to be 0.33R and the PSU line 50V, that works out to 15.15 Amps per output pair. I realize that I am playing with fire here, but in my view not everybody is ever going to push it that far - consider what 4 output pairs can pass on to the speaker, over 60 Amps, and few speakers could survive that without some blue smoke in the room.
Also, remember I use reguilated +/-56V for the IPS, VAS, and predriver, while the driver and the output trannies are fed off PSU lines of +/- 51V, which improves their SOAR position.
Also, remember I use reguilated +/-56V for the IPS, VAS, and predriver, while the driver and the output trannies are fed off PSU lines of +/- 51V, which improves their SOAR position.
I strongly beleive speaker cables do make a difference ( Dan that Youtube was sent to my boss, I think he will hate it ). I suspect 99% of people have the wrong ones. If a valve amp they should use the same cable as the transformer and could even make it continuous if a zero feedback option. They only calcualtion being damping factor. Against what is understood low damping factor can be very good. It can be between 3 and 16 to be a good choice. The Qts of the speaker and Q of it's way of working come into it more. Many speakers have a choke to the bass unit. If so what is the story of damping factor then ? Dejan told me I should not have an output choke on my amp. 18 inches of 1 mm wire 8 mOhms ( 16 turns 8 mm id ). It's nothing. That's circa 27 ohms per mile or 17 per kM. 0.25 ohms is where it might be heard and not always worse.
The ideal wire is about 0.6 mm . One can use 2.5 mm for subs and 0.25mm for tweeters. The sound will seem at first slightly dull. What you should notice is the music is easier to follow. The only big deal is most people have paid too much. If you take my advice buy the wire on a spool and do not be rough with it. Try not to kink it.
If I was winding transformers I might see if 0.6 mm is OK for the output side.
The ideal wire is about 0.6 mm . One can use 2.5 mm for subs and 0.25mm for tweeters. The sound will seem at first slightly dull. What you should notice is the music is easier to follow. The only big deal is most people have paid too much. If you take my advice buy the wire on a spool and do not be rough with it. Try not to kink it.
If I was winding transformers I might see if 0.6 mm is OK for the output side.
On paper what I use looks to be about 3 amps per device. Using a very unscientific test I can say 5 amps conventional sounds less loud. I suspect that is why this simple circuit of the old Quad 303 is not used as it is not like the conventional version. It could simply be it distorts in a nicer way that makes me feel I got more. I know for a fact these limiter circuits are very crude. Maybe I have found that the worst is best? I discovered how good by accident. I was trying the idea without realising where I saw it before on a power supply I was building. It was better and much easier than text book foldback limiters that to be honest are useless. Ultimately neither was OK for that. I use a comparator and crowbar on the PSU. I crowbar the reference voltage.
So what is your bosses problem...Hitler or subjective opinions ?
From experience I know that cables make differences.
Differing cable R/L/C/DA, differing (non ideal) amplifiers, differing speaker impedence/phase behaviour.
Why should cables not make any differences.
Discussion on JCBT thread is saying adding series resistance improves speaker distortion performance.
16 ohm speakers maybe not so stupid after all.
So what Dejan is actually saying is that output inductance is detrimental ?.
It's a small inductance value, should be inaudible but it's not according to his findings.
So what's happening, is it setting up of out of band ringing that causes audible artifacts ?....Dejan ?.
Easier to follow sounds like less hash/artifacts, is that what you mean ?.
0.6mm or so is telephone MDF jumper wire size is it not....Cat 5 is 0.51mm.
Dan.
Nige, a little formatting in your posts like paragraphing would make them much easier to read/digest.
I do study your posts, just that sometimes it's a bit like hard work.
No offence intended please, I say this in the interest of more effective communications. Kudos.
I take that as a sincere compliment, thank you.
If I were in Europe, in a heartbeat I would like to drop in and see/listen/enjoy your work Dejan.
Frank is expressing his desire to hear properly sounding audio reproduction.
Frank also relates that far too much of the expensive gear that he has heard at Expos is wrong, totally wrong.
Frank also expresses when he hears gear that really is good and pleasing.
The audio world is full of both degreed and self proclaimed 'experts' who don't really understand what they are hearing, and the typical consumer has less than zero clue.
I sometimes wonder about the living/working environment of many designers/users.
By this I mean general and persistent/continuous sonic/infrasonic pollution like traffic, air conditioning, general city scape omnipresent sounds etc.
In my living situation if I turn off my home appliances I have silence, outside in the back yard is silent of man made sounds.
So generally, the only sounds are natural sounds...birds, wind, trees rustling and when I go to the close by beach I hear only natural surf sounds.
I really treasure this natural sound reference, and this is one of my arbiters.
Standing outside and listening to the reproduced sound emanating from the house very quickly forms an opinion of any system/tweak.
Natural sounds are generally pleasing/comforting.
Man made sounds generally are not.
Reproduced sounds can sound natural, but generally do not.
Dan.
I know Dan. I'm the worse when reading other peoples stuff. I want it clear as clear.
The single core is good for cables as no one needs to spend much to see if they agree. I found this out by accident. A pair of Bose 901 were being used in a church. The wires were not long enough. My French wasn't up to saying exacly what I wanted. I said something like get me a reel of 1mm lighting cable. The guy comes back with 10 metres of add on and a choc bloc. I added it on and found it was as solid core sounds. This suggests the solid core is subtly a filter be it RF or AF. The solid core people would feel once ruined that's it. Not so. The solid core is doing something which might to some be worse? All I can say is my system has a way of making music easy to understand. The filter effect is the size of core. Like most things it is how you like it. 0.6 mm has bass. Cat 5 should be ideal.
What I think discredits genuine differences is people must then extract money from other people. That's when sane people run a mile. Genius is close to madness as a rule. Thus sometimes we mix up who is telling the truth. Nothing I know of need cost money to prove. If it does I don't need to know. Even with wine this is true. No one need ever spend more than £10 bottle. Often that is double the real value. Cava is good enough for anyone. Think about it, they are selling spoiled grape juice.
A painting I wanted to pinch is John Constables The Lock. I told Colleen there was a good chance the Prado ( Thysen actually I think ) will find it's gone that night. Tuner I adore. Constable I slightly prefer if that's not too stupid to say? I find out Constable was not sucessful in his own time. Tuner was. That is crazy as in many ways both Turner and Constable showed the French how to let go of the past. Constable was honered by the French in 1824 when England didn't rate him in the least. Constables paintings sold at his death for about £3 which although plenty was not a fortune. This is the subjective nature of how we see things. Turner was more bussiness like and popular. He was already accepted at age 14 as an artist. Constable was a difficult man. He was without doubt a genius and caused impressionists to evolve ( He changed the mindset . De La Croix for example was moved by what he saw and retouched his exhibit piece of 1824 ). Constables drafts were not in pencil. His drafts looked like what was to come. Turner was so concerned he changed his approach. Vincent we can all understand as unusual. A mans madness for all to see. Constable defeats me as to how he was invisible. The Lock Gate sold for £20 000 000 a year or two ago. We are the same people now as then. Our perceptions have moved on. What this really means is most people haven't a clue yet have the look of someone who does.
Dejan said I should be careful about output chokes. Yes, but not damping factor.
Dan I am nearly 60 and aged 50 I had never written in my life. It's a lot of catching up. It's ill health that started this. That seems to have reversed and I am not bad for 60. I can not reliably write with a pen as my only problem or carry a pint in my right hand. Reparing am amp last night I was using my left hand to do up a nut in a confined space and didn't realise. Give me 20 years and the left will function as the right once did.
https://gerryco23.wordpress.com/2013/02/04/constables-oil-sketches-atmosphere-and-light/
The single core is good for cables as no one needs to spend much to see if they agree. I found this out by accident. A pair of Bose 901 were being used in a church. The wires were not long enough. My French wasn't up to saying exacly what I wanted. I said something like get me a reel of 1mm lighting cable. The guy comes back with 10 metres of add on and a choc bloc. I added it on and found it was as solid core sounds. This suggests the solid core is subtly a filter be it RF or AF. The solid core people would feel once ruined that's it. Not so. The solid core is doing something which might to some be worse? All I can say is my system has a way of making music easy to understand. The filter effect is the size of core. Like most things it is how you like it. 0.6 mm has bass. Cat 5 should be ideal.
What I think discredits genuine differences is people must then extract money from other people. That's when sane people run a mile. Genius is close to madness as a rule. Thus sometimes we mix up who is telling the truth. Nothing I know of need cost money to prove. If it does I don't need to know. Even with wine this is true. No one need ever spend more than £10 bottle. Often that is double the real value. Cava is good enough for anyone. Think about it, they are selling spoiled grape juice.
A painting I wanted to pinch is John Constables The Lock. I told Colleen there was a good chance the Prado ( Thysen actually I think ) will find it's gone that night. Tuner I adore. Constable I slightly prefer if that's not too stupid to say? I find out Constable was not sucessful in his own time. Tuner was. That is crazy as in many ways both Turner and Constable showed the French how to let go of the past. Constable was honered by the French in 1824 when England didn't rate him in the least. Constables paintings sold at his death for about £3 which although plenty was not a fortune. This is the subjective nature of how we see things. Turner was more bussiness like and popular. He was already accepted at age 14 as an artist. Constable was a difficult man. He was without doubt a genius and caused impressionists to evolve ( He changed the mindset . De La Croix for example was moved by what he saw and retouched his exhibit piece of 1824 ). Constables drafts were not in pencil. His drafts looked like what was to come. Turner was so concerned he changed his approach. Vincent we can all understand as unusual. A mans madness for all to see. Constable defeats me as to how he was invisible. The Lock Gate sold for £20 000 000 a year or two ago. We are the same people now as then. Our perceptions have moved on. What this really means is most people haven't a clue yet have the look of someone who does.
Dejan said I should be careful about output chokes. Yes, but not damping factor.
Dan I am nearly 60 and aged 50 I had never written in my life. It's a lot of catching up. It's ill health that started this. That seems to have reversed and I am not bad for 60. I can not reliably write with a pen as my only problem or carry a pint in my right hand. Reparing am amp last night I was using my left hand to do up a nut in a confined space and didn't realise. Give me 20 years and the left will function as the right once did.
https://gerryco23.wordpress.com/2013/02/04/constables-oil-sketches-atmosphere-and-light/
Last edited:
Me also
.Yeah, I'm hearing you.
Curious how the last bit of cable can undo/correct for upstream cabling.
There have been lots of cable theories presented, like micro diodes, skin effect etc.
I am not clear on the reasons, but I do agree that solid cable tends to sound more 'coherent' for want of better term.
You mean sour grapes
.I have a coaster that changes flavour of wines in short order...JC has touched on this.
AUD$10.00 reds taste perfectly fine when treated...no need for $30.00 stuff, and in general I find them to be inferior to the honest but ok cheaper stuff.
There is a limit to cheapness of course.
Mr Raidho has has given input on the JCBT thread about his companies high end speakers, which are extraordinarily priced...USD$89,000 for their top model.
Not mentioned so far is Raidho cable prices....thousands of dollars for normal lengths.
You and I understand that is just taking the mickey, just like high priced wines and artworks are taking the mickey.
Like you, I respect good engineering and abhor overpriced nonsense.
£20 000 000 for an artwork is pure bovine.
I would be perfectly happy with a nice quality print/copy/forgery hanging on the back of the lav door.
Those with that amount of liquid spending power would do much better to construct local community halls or hospital wings or educational facilities, or sustainable energy research....something useful to mankind and a lasting legacy.
If Tibetan Buddhism beliefs are true these types deserve to spend eternity in the Bardo.
Dejan ?.
Soldier on, we like what you do.
Your explorations and observations are a pleasant breath of fresh air.
Nah, these paintings don't really do it for me....too blurry, indistinct.
A few years back I visited the Nymindegab Museum in DK.
Some really stunning older paintings with detail to rival the best modern digital camera, and somehow a sense of being there as the painter recording the scene.
Dan.
Appreciate the good thoughts of you both - likewise, I would have some fun listening to Dejan's efforts. Be warned though, I'm a hard taskmaster!!
- I'm very disappointing to have around, the person who has assembled the system is typically so excited by what it's doing right, for him - and all I hear are the little bits and pieces that it's doing wrong - in audio, I'm the ultimate pessimist! 
Really they culled the worst of the worst.
Don't get me wrong I have great Aussie (and Kiwi) friends.
The cable thing is all about the fact that real bits and pieces, when used in situ are not "perfect", and therefore ideal in how they behave - I get around the issues by hardwiring everything - bingo! those "problems" disappear. An audio friend who has been "contaminated" by my thinking,
, discovered this for himself by comparing expensive audiophile cable used normally, vs. hardwiring cheap, everyday wire - guess which performed better?I also use single core cable whenever possible, because it automatically will have less chance of not being being "perfect" - I have never been silly about this area, I have heard "good sound" too many times when this has been completely ignored - thus, it's not a major factor.
Last edited:
To be precise, I said that the output shoke is far too often a cover story for a poor PSU.
I can see some cases when it might be a good idea, for example, with certain amps which have a power bandwidth into the MHz range, and you can push that only so far. However, generally, it does what it does but unfortunately it also affects the damping factor, and as we have learnt from the other thread, messing with the damping factor also messes with frequency response linear delivery into real world loads. To a linearity freak like myself, that's sacriledge.
- Status
- Not open for further replies.