I was going to build the latest Piersma tweaked Circlophone amplifier but after this experience I'm thinking of selling my boards and trying the Cambridge audio 651a. If I buy it online I can return it after a lengthy audition (up to a month), if I don't get on with it.
The cheap one I bought from ebay was "untested". The first thing I did was to test the DC offset at the speaker outputs. My original one was around 1.3mv on both channels. The ebay one was 50mv on one and 70mv on the other.
I plugged it into my system and noticed that the upper midrange was a bit thin/rough sounding on the ebay AM1, so removed the lid and cleaned out a ton of crap with a toothbrush and vacuum cleaner. - Not sure if this was the cause but I found a toenail
wrapped around the legs of two components (could this have caused a short or something?)I then tested the DC offset and it was around 1mv on both channels. I then plugged it back in and it sounded the same as the other one to my ears.
Every time I take the lid off and mess with the circuit, I test the DC offset afterwards and it always been fine.
EDIT: Forgot to say that since removing a lot of the mods they now sound VERY similar.
The times when I have noticed a big improvement in sound with cap rolling is when the caps were really old and "dried out." A dumpster diva Pioneer receiver was recently brought back to life by simply replacing the big power supply caps.
The fact is that engineers are well versed in turd polishing. They really know how to get the maximum performance out of the cheapest parts. Changing cap values for lower F3 etc can cause big problems too. I've even made tone controls oscillate by swapping the caps for film caps before.
That's why your own design and build, painstakingly prototyped and tested, is your best bet for a custom amp.
The fact is that engineers are well versed in turd polishing. They really know how to get the maximum performance out of the cheapest parts. Changing cap values for lower F3 etc can cause big problems too. I've even made tone controls oscillate by swapping the caps for film caps before.
That's why your own design and build, painstakingly prototyped and tested, is your best bet for a custom amp.
A good example of turd polishing is an old Mitsubishi receiver my buddy had in his garage for at least 15 years. It worked really well and sounded very good. When it self destructed (STK module smoked) I stripped it for parts. It had what would be considered boutique capacitors for its era (mid 80s) and what I would consider "deluxe" tone controls for a cheap receiver (no ceramic or electrolytic caps in the tone shaping circuit). It made me think that they must have hidden this design from the bean counters.
Maybe i am wrong, if audio circuits are concerned only, but from my experiments(actual testing) a linear power supply could not deliver small ripple(load regulation) until i added 2200uF Nichicon PW or panasonic FC in parallel with 2x10000uF 63V samwha capacitors. And the more i soldered these 2.2mF 50V capacitors in parallel, the more stable voltage on the output i got
That was simple plane test with the serious (6.6OHM) load on the output, measuring output voltage with Multimeter.
Transformer 300VA, 2x30V AC for 2x30V DC output, 5A per channel.
They are not really that expensive too, 2.2mF capacitors, especially samwha or nichicon brand.
Which is it, ripple or load regulation? They are not the same. How exactly did you measure these things?
In overall, with 6.6R load, the voltage was always closer to 30V when paralleled.
Started with 2x10000uF samwha's
http://www.kreocen.pl/img/p/1552614...T-10000uF-63V-40x40mm-a20-HE1J109M40040HA.jpg
I cant give you exact numbers, it was few years ago, i reported it in here:
Electronics Projects and Circuits - Electronics-Lab, with all the new webpage and community i cant log in with old user and email is not recognized.
This is the linear PSU:
DIYfan: Adjustable Lab Power Supply - take two
Method: Turn on psu, connect load and measure voltage turn off.
Solder 2.2mF in parallel and turn on again, procedure is same.
Solder 2.2mF again and test again.
It was 100% clear, that these samwha's were not good alone.
The most effect was from two first 2.2mF capacitors that i soldered, third one did influence the voltage on the output, but with a smaller fraction.
There's an old definition out there:
"an Engineer is the guy who can make for 50 cents what any fool can do for $10"
A couple personal examples:
I make Guitar and Bass amps since forever (1969) , local competitors *hate* me, mainly because I can sell a good working reasonably good looking amp for about their parts cost
One reason is that I invested heavily in machinery and can make in house: chassis/PCBs/transformers/speakers ...besides metal corner protectors/rubber legs/strip handles/reverb tanks/knobs/horn drivers ... even publicity T Shirts .
One of them complained about my low prices and that I was "stealing the market", somewhat accusing me of unfair dumping.
I asked him
- "how much do you pay for a PCB?"
- "U$50"
mind you, he orders one-offs from the best Argentine Factory,
.::Circuitos Impresos - Ernesto Mayer S.A.::.
has *them* design the layout, he only specifies where should pots/jacks/switches be to match front panel (which he also orders outside, fully finished) , and said PCBs are double sided, plated through hole, fiberglass base
- " well, for U$47 I buy the full 3 ft by 4 ft FR2 single sided copper clad phenolic paper material, and in about 2 hours work have 40 (forty) preamp/power amp/supply sets which are fully equivalent to your gorgeous all in one PCB .
FWIW what I do is same as established manufacturers do, such as Fender/Peavey/Laney/etc.
Peavey Bandit: notice use of phenolic material (in this case Micarta) , single face board and quite a few wire links to avoid running traces on top.
Oh, it was definitely approved
Fact is in late 70's to late 80's Japanese Hi Fi was "king of the hill" , the undisputed Market leader (Sony / Yamaha / Pioneer / Luxman / Onkyo / JVC / Mitsubishi / Sansui /Akai / etc.) they run incredible ads, such as magazine fold out 3 page magazine back covers, etc.
They made innovative FET amps, Class A , even the first Class D , also the best commercial cassette decks, turntables, the works, and to enhance perceived Audio quality they also were very creative on the "mechanical" side, gorgeous complex cabinets and hardware, even liquid cooled heatsinks.
The Empire came down when people massively started using cheap digital audio, headphones and 2" plastic PC speakers as their source of Music ... trend that stays today.
People truly interested in Audio Quality (ahem !!!! , as found here
) is about 0.00001% of the World market.But yes, do not be surprised at 80's Japanese high quality: it was there and there was a reason for it.
FWIW same happened with high quality film cameras, does anybody remember Nikon / Canon / Minolta / Olympus / Konica / ... even "middle class" Pentax?
That's interesting about Japanese hi-fi, Mr. Fahey. I didn't think that Mitsubishi was a big player in the hi-fi market. I must give credit, though; the unit provided very good performance until it emitted the magic smoke. The only thing it didn't have was protection circuitry, which cost my buddy a woofer. And if they hadn't used those stupid STK modules (ever cut one open? They're stupid inside) it might still be working.
Some of the film caps have been reincarnated in active filters I've built. I can buy better (and I do all the time), but I doubt it would make an audible difference. Changing ceramic and electrolytic capacitors for film capacitors in tone shaping circuits makes a very audible difference.
I own some good old equipment, and I've worked on a lot too. The difference between a mid-80s era Sony amplifier and modern Sony equipment is stark. The modern stuff is so awful, in spite of technological advances and much better parts that are available today. There seems to be a huge gap between consumer grade equipment, which is truly horrible (but still pretty good for the price I guess) and anything remotely worthy of being called hi-fi. In my community, hi-fi is only available in snooty, ridiculously overpriced boutique stores (and at resale shops if you're lucky).
I have old Luxman, Nakamichi, Fischer, Sony, and Pioneer equipment that I rotate in and out of service like Jay Leno rotates his cars. Maybe I'm eccentric but it makes me happy.
Some of the film caps have been reincarnated in active filters I've built. I can buy better (and I do all the time), but I doubt it would make an audible difference. Changing ceramic and electrolytic capacitors for film capacitors in tone shaping circuits makes a very audible difference.
I own some good old equipment, and I've worked on a lot too. The difference between a mid-80s era Sony amplifier and modern Sony equipment is stark. The modern stuff is so awful, in spite of technological advances and much better parts that are available today. There seems to be a huge gap between consumer grade equipment, which is truly horrible (but still pretty good for the price I guess) and anything remotely worthy of being called hi-fi. In my community, hi-fi is only available in snooty, ridiculously overpriced boutique stores (and at resale shops if you're lucky).
I have old Luxman, Nakamichi, Fischer, Sony, and Pioneer equipment that I rotate in and out of service like Jay Leno rotates his cars. Maybe I'm eccentric but it makes me happy.
That is exactly how it would sound if you use the ultra speed TO-220 diodes. Congratulations!
From what I have yes. They are super fast to hyper fast, designed for switching supply. Including MUR860 (U860) from Motorola.
I have no idea yet. May be AndrewT knows:
With diodes we deal with spike/ringing/oscillation/EMI. But I guess there are more to it (from listening experience). Most of those fast diodes have fast recovery time, which tend to be complemented by ringing.
Soft recovery diodes do not usually have this problem.
It looks nice. There are compromises in diodes so whatever the diodes, manufacturers can always write a good marketing publicity (by focusing on the positives of course).
And most diodes are marketed as a product for switching with no direct relation to audio...
And the reason I don't want to conduct my own experiment is because I don't have a plan to purchase the diode that will appear to be the best from my experiment. I prefer to use what I have on hand, so direct listening is what I prefer instead of finding out why they sound like they do (or even what type they are).
But from guessing (and relating perceived sound with a few "specs"), I think I might know what seems to be important for audio...
First, I think the Q or reverse voltage spike is the most important. The Q of the Qspeed looks like a schottky, which is the best if we want low Q.
Second, the ringing is not good as they can introduce EMI. But I consider this as not very critical, especially if we can find work-around such as finding the right amount of capacitance (tho very expensive). The ringing can be solved in 2 ways. First is using RC snubber, but I think it is its effect in increasing Q that brings negative effect to sound (or may be there are others that I don't know). Second, is expecting a soft recovery behaviour. But I think it has very slightly negative effect to sound, which is masking the details.
And I think there are others that I still have no idea yet. I believe it is related to frequency and phase.
Another way to avoid ringing diode is I think by chosen diode with the right Imax. If we look at the diode spec, the higher the temperature, the diode starts to behave like soft recovery ones. So we want a "warm" diode. Not a cool one because of high current capability. Diodes that require heatsink tends to sound good (tho may be just because it is a schottky, or other parameters) than diodes with hundreds Ampere of current capability.
To be honest I was playing devils advocate regarding the Qspeed diodes (was curious to hear your opinion). If you read back through the thread you'll see that the lower mid suckout was caused by the polyprop coupling caps and the slightly soft sound was caused by the Mills emitter resistors.
The amps sound remarkably similar now and It takes extended listening to lots of music to notice a difference (very slight upper mid brightness on only a very few select tracks).
Now It could be the diodes causing this but my money would be on something else. The Nichicon feedback cap is suspect number one.
The amps sound remarkably similar now and It takes extended listening to lots of music to notice a difference (very slight upper mid brightness on only a very few select tracks).
Now It could be the diodes causing this but my money would be on something else. The Nichicon feedback cap is suspect number one.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.