analog_sa said:
Thank you for your kind offer. I would have built one long ago if it wasn't for the relay - i really hate passing audio through relays. Maybe i've just invested too much into decent speaker terminals
Yeah, not to mention speaker cables!!
(mine are Straightwire Crescendo... a few cm of which costs about as much as a complete "My_Ref"!
)Indeed, removing the relay on the signal path (together with more output current capability and maybe something like a couple of Zeners across the LM318 for soft-clipping...
) is something I would really really like Mauro to change in a next revision of "My_Ref"... if there will ever be one! 
I would not suggest leaving out the relay though, unless you want to remove one speaker wire during turn-off, and turn-on. Honestly, I was not crazy about the relay either, but in fact I am glad it is there, it makes the amp virtually bullet proof and safe to drive almost any speaker, it also protects your speakers should something bad happen with your source. I have lots of amps, and lots of cable, and I would not say the MyRef lags behind anything I have, relay not withstanding. Anyone within earshot of one will see exactly what I mean. 
Cheers!
Russ
Honestly, I was not crazy about the relay either, but in fact I am glad it is there, it makes the amp virtually bullet proof and safe to drive almost any speaker, it also protects your speakers should something bad happen with your source. I have lots of amps, and lots of cable, and I would not say the MyRef lags behind anything I have, relay not withstanding. Anyone within earshot of one will see exactly what I mean. Cheers!
Russ
Russ White said:
me not, either! of course I was talking about a different protection scheme (as we discussed already here some time ago), and not just to leave out the protection altogether!
Moreover, Mauro already said that he have taken the relay contact resistance into account in his design. Thus simply removing the relay is quite likely NOT the right thing to do. And that's why I'm somewhat trying to push Mauro himself to do some investigation in that direction...
Well, I know. He also said several times that he is completely satisfied with his design as it is... but I count on the fact that, if he is even remotely affected by the audiophile syndrome, he can't be satisfied too long before start thinking about possible improvements...
don't take me wrong! I never meant to say that "MyRef" isn't good as it is!
I just wonder that _MAYBE_ it could be made even better. Or maybe not.
But I think that -exactly because it is so good as it is- it's worth trying any possible solution which may made it better!
In my humble opinion, there are a few directions that may be explored for possible improvements. Here they are (in no special order):
- try to allow for a small amount of current feedback to remain in effect (rising the internal impedance to a few tens of an ohm, say 0.3~0.5 ohm, not much...).
That could be done in several ways... the most obvious was suggested by Mauro long ago (just add a resistor across LM318). But as he clearly stated, that would deeply alter the principle of operation of the circuit, thus I don't like the idea.
A possible (less invasive?) alternative would be to try to reduce the overall NFB... but that may have nasty side effects, too.
Yet another would be to try to reduce the gain of the _output_ stage of the LM318 alone (acting on its pin 8), while keeping the input floating et all.
To me this seems the most promising. That would also improve overall stability and reduce the output impedance of LM318, possibly improving the interface with the current pump. Of course it would also reduce the overall feedback, so it may be bad, too (though we are talking about quite a small change, so it should not be bad, at least not due to the reduced overall NFB).
Needles to say this of course would require an almost complete re-design. I would have tried to do that myself but, honestly, I have not enough skill and experience to do it properly on such a fine-tuned design. But I hope that Mauro will find something interesting in the idea, and get it...
- try to increase load insensitivity by increasing the output current capability.
IMHO, the ideal (for ANY amplifier!) is having Imax = Vmax / (Ri + Rs), where Ri is the internal impedance of the amplifier (closed loop) and Rs is the sum of the unavoidable stray resistances such as contacts+cables+...
This would guarantee that the output _current_ would remain linear even if at times the load presents itself as a momentary short circuit. That is, it will never clip due to any load, no matter how insane.
Of course, with Vmax ~32V, even assuming Ri+Rs=0.8 ohm that would call for some 32/0.8=40A. Using 3886s, that would require at least 4 (!) parallel current pumps...
- try to ensure a non-blocking, soft clipping in any condition anyway. As already hinted by Mauro, and easily seen by carefully looking at the circuit, that could be as easy as adding a couple of Zeners across the LM318. The only "minor" problem (which require a lot of skill and expertise) is how to adjust the circuit (specially the compensations!) to avoid nasty side effects.
- try to design an alternative protection scheme to avoid the relay contact in series with the output;
Well... that's all, folks!
I want to stress that this are simply rough ideas, which may or may not work. Of course there's some background behind them, but they can be completely wrong. Anyway, I want to made it clear that I presented them here as a positive contribution to stimulate further developments (that is, just in case some of them were not already thought out by Mauro!
).Again, I do NOT intend to criticize in any way Mauro's excellent work.
No worried UnixMan your posts have been some of the best educated and intended posts on the thread!UnixMan said:
Oh no, I actually was in total agreement with you, I just did not want anyone to missunderstand some of the other posts. I think your ideas are very good. I am sure Mauro will chime in at some point and validate that.
I can't wait to see the fruit of that discussion!
Cheers!
Russ
I know I am going to be shot down for this but I feel that we have to remove the blockage between the psychological and physical. The biggest problem with our audio equipment is between our ears (and I include myself because I like to tweak)
There are a couple of changes that you can change in your system that will make a big difference
I will start with the source
1.Change that factory supplied power cable on your CD player for something decent if you can’t afford a fancy power cable just use a 16A cable that will give you a big difference. (Btw power cables also work well on DAC and preamp but not much of a effect on amplifiers)
2.You got to have decent interconnects. You need not by the most expensive. You need to test them in your system. Lots of name brands are real rubbish. If you are unsure what to get just try some RG58 cable till you get a cable that sounds good for you. You will be surprised
3.Speaker cables and special connectors do make a difference. But maybe 1%. If you are unsure what to by. Use some CAT5 used in a platted configuration.
4.If your system sounds better with a preamp. You have a serious problem, it means that there is something wrong in the rest of your system and the preamp is fixing it
So some of us are worried about a relay that might make a 0.5% difference to the sound, the relay will not change your signal that much. The inferior x-over components are of a much bigger worry. Use a decent relay and you will not have a problem (something like a Finder). Rather worry about the other rubbish in your system that really makes a difference, such a power cables. Ill matched interconnects and my favourite culprit the dome tweeter.
Rudi
There are a couple of changes that you can change in your system that will make a big difference
I will start with the source
1.Change that factory supplied power cable on your CD player for something decent if you can’t afford a fancy power cable just use a 16A cable that will give you a big difference. (Btw power cables also work well on DAC and preamp but not much of a effect on amplifiers)
2.You got to have decent interconnects. You need not by the most expensive. You need to test them in your system. Lots of name brands are real rubbish. If you are unsure what to get just try some RG58 cable till you get a cable that sounds good for you. You will be surprised
3.Speaker cables and special connectors do make a difference. But maybe 1%. If you are unsure what to by. Use some CAT5 used in a platted configuration.
4.If your system sounds better with a preamp. You have a serious problem, it means that there is something wrong in the rest of your system and the preamp is fixing it
So some of us are worried about a relay that might make a 0.5% difference to the sound, the relay will not change your signal that much. The inferior x-over components are of a much bigger worry. Use a decent relay and you will not have a problem (something like a Finder). Rather worry about the other rubbish in your system that really makes a difference, such a power cables. Ill matched interconnects and my favourite culprit the dome tweeter.
Rudi
Well I know they are good but once you have heard a very good ribbon there is just no going back. Ribbons don't have the same flat response of a dome BUT they are much smoother to listen to. So for a manufacturer to show the 2db dip of a ribbon just does not make good marketing sense. Ribbons are also a lot lighter that domes and moves much quicker giving a better transient response
I have found that even a cheap ribbon generally sounds better that a expensive dome
The one I use have a frequency response from 500hz at –6db and goes up to 44khz (no x-over point to content with in the most important ranges)
Have a look here http://www.pinsh.com/introduction.htm
I have found that even a cheap ribbon generally sounds better that a expensive dome
The one I use have a frequency response from 500hz at –6db and goes up to 44khz (no x-over point to content with in the most important ranges)
Have a look here http://www.pinsh.com/introduction.htm
rudi said:
Hmm. I don't doubt that they are good, but, but...
Scanspeak say, their 9500/9700/9800/9900 'Revelator' tweeters have a moving mass 0f 0.45G (other of their ranges are less, but still more than .006G)
Pinsh say:
My calculator says that .006 x 167 = 1.002G
I'm missing something here, what were they comparing with?
maf_au said:
your guess is as good as mine. not for a moment do i think that the scan and other top makes are bad. I just think once you've heard a ribbon you will not go back if you want real cheapies with GREAT sound try these http://www.madisound.com/cgi-bin/index.cgi?keywords=hi-vi they are made in china but with great sound. i have tried the cheapest one before ( RT1.3 Isodynamic tweeter 92dB $ 42.75) and even that sounded fantastic. for bass there is this one (D10.8 11" Poly Cone Woofer, 4" Voice Coil $ 104.15) just can't be beaten
Ok enough about speakers
sorry guys for hijacking the topic. Maf_au. you can mail me if you want more detail. i can also put you into contact with the pinsh manufacturers
rudi said:
This point is is not precisely correct I don't think, as many amps are designed with very low Zin, and if you pair that amp with a source that has high Zout you will absolutely need a preamp/buffer. Thats not really a flaw in the power amp, its just a design choice. Would I make that design choice? Well probably not, but that does not make it "bad".
Ok now back to the topic.
Cheers!
Russ
Russ / Mauro-
I am using the mono brds each with their own 100va transformer on a pair of 6.5" 2 way book shelf speakers and it sounds REALLY good. Actually FANTASTIC.. Because of this I have a thought.. Oh oh.
I know this may appear to be a silly question but it is for general knowledge not fact and figures...
What is the "wattage" rating of the Rev_C?
I know the figures of the LM3886 but I do not know if it is capable of those numbers in this configuration.
The reason I ask is below if you want / need history.
I am having cabinets made for my new speakers now which each side will consist of the Scan Speak 9900 revelator tweeter, B&W 6.5" Kevlar mid (drivers measured and crossovers designed and built by Meniscus), and a pair of Parts Express RS 10" Aluminum cone woofers.
The speaker is bi-amped with the Hypex UcD180 for the bass and soon to be completed KrellClone KSA-50 on the top. There is fancy woofer signal manipulating done by Mr. Erath's (remember LWE speakers? The Ferrari of speakers in the 60's / 70's) Feedback Controller Network.
Not a simple system I know, but it really does sound musical. Flat bass from 125 to 20Hz and "satellite" type imaging and detail.
So another question...
Do you think I could use the Rev_C on the top driving the B&W mid / SS tweeter as a partner to the Hypex UcD180?
Would it be "powerful" enough to drive a 87db efficient top end with a 92 db bottom with larger amplifier?
Edit: name seperation
I am using the mono brds each with their own 100va transformer on a pair of 6.5" 2 way book shelf speakers and it sounds REALLY good. Actually FANTASTIC.. Because of this I have a thought.. Oh oh.
I know this may appear to be a silly question but it is for general knowledge not fact and figures...
What is the "wattage" rating of the Rev_C?
I know the figures of the LM3886 but I do not know if it is capable of those numbers in this configuration.
The reason I ask is below if you want / need history.
I am having cabinets made for my new speakers now which each side will consist of the Scan Speak 9900 revelator tweeter, B&W 6.5" Kevlar mid (drivers measured and crossovers designed and built by Meniscus), and a pair of Parts Express RS 10" Aluminum cone woofers.
The speaker is bi-amped with the Hypex UcD180 for the bass and soon to be completed KrellClone KSA-50 on the top. There is fancy woofer signal manipulating done by Mr. Erath's (remember LWE speakers? The Ferrari of speakers in the 60's / 70's) Feedback Controller Network.
Not a simple system I know, but it really does sound musical. Flat bass from 125 to 20Hz and "satellite" type imaging and detail.
So another question...
Do you think I could use the Rev_C on the top driving the B&W mid / SS tweeter as a partner to the Hypex UcD180?
Would it be "powerful" enough to drive a 87db efficient top end with a 92 db bottom with larger amplifier?
Edit: name seperation
rabstg said:
Power output of My-Ref is 40W/8ohm – 56W/4ohm
If you are driving good speakers you will never even come close to maxing out the My-Ref, which it sounds like you are. Being able to supply the amp with 100VA torroids does not surprise me at all. In fact I have actually run a pair of them in stereo on one 150VA trafo with very good reults, I just would not push it really hard with 4ohm speakers.
If you can set the gain correctly on both amps the My-Ref driving the top end should work well, if fact I could have a very similar configuration to yours if I chose, as I have a KSA-50. Maybe I will try it!
The My-Ref is powerful eneough for 90% of high end speakers, the only speakers I would stay away from with My-Ref are those that are ridiculously inefficient or very hard to drive like 2-3 ohm loads. 4ohm is the lower limit really.
I have bi-amped with My-Ref usuing a crossoverless Cryolite with excellent results. I am working on a digitally controlled x-over using maxim digitally controlled analog pots to help me with other projects so I can have an easily adjustable x-over.
Cheers!
Russ
Russ White said:
Sorry Russ. yes you are 100% correct.
I tend to match everything before I buy or build. but if you get much better results with a pre in a matched system them you need to go an look for the source that needs fixing. once you have it running at its best with just a volume control then you know you are spot-on
cheers
Rudi