New Amp Completed *PICS*

MartinLogan Audio Owners Forum

Help Support MartinLogan Audio Owners Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Dominick22

Well-known member
Joined
Aug 19, 2005
Messages
281
Reaction score
1
Location
Columbus, Ohio
*PICS BELOW*
I named her "MAHGER". If anyone knows how to go about getting a metal badge made........?

The down and dirty:
-The amp chassis is made of aromatic cedar.
-4 LM4780 chips each wired in parallel.
-2 Noble (32 step) Pots. One controls the ESL's (highs). One controls the Woofers.
-2 420 VA torroids that are stacked and 1 12V transformer.
-4 Nichicon Super Through caps 10,000uf.
-4 Nichicon Gold Tune caps 4,700uf.
-Audioquest copperhead interconnect wire throughout.


The purpose of this amp is to power my Martin Logan Ascent i speakers in stereo mode. Essentially, this amp was disigned to bi-amp them. The 10,000uf channels push the woofers, while the 4,700uf channels push the electrostatic panels.

I know that I will probably receive some criticism for buying expensive interconnects and then cutting them up, but there really is a difference in sound (I did a side by side comparison). Will I ever do it again? ABSOLUTELY NOT, unless it is to build for someone else. I spent about $400 on just the green wire (audioquest copperhead) you see in the pics, but there really is a difference in sound from my last LM4780 amp. I just think that it is getting into the "law of diminishing returns". None the less, I felt that I would try to run the exact same wire from the signal source to the speaker output. The difference can be described as the amp has taken a deep breath. The amp is more open, although not in an airy sense. Transients are lighter, but clearer.

I am very happy with this amp and it was an "all out" approach that I took in building it. A once in a lifetime thing although I will keep building.

Although this amp was originally designed for the Ascent i's, it is capable of driving much more than that. Each component (ESL/Woofer) in each speaker has 120 watts available to it. Thats 240 watts per speaker in reserve. 480 total watts for the amp. But more important is the 840 VA that the power supply is capable of. Anyway, tell me what you think either way.







MORE PICS BELOW

Dominick in New Jersey
 
Last edited:
Um, the pics you have posted appear to be the thumbnails. I guarantee that they aren't the 1600x1200 indicated on the pics :). (unless you're using micro-pixels of course :) ).

On second look, if members delete the ".th" from the URLs as posted, you'll get the full-size versions.
 
Im not experianced at posting pics, but i meant to only post the thumbs for a quick look. You can enlarge them twice if you want a close, detailed view. I think!

Dominick
 
Interesting design. How much current (amps) will it deliver to the ESL?

Also, why larger capacitance reserves for the woofer vs. the panel? The Panel typically draws more, which means it would need the greater reserves.

Even though the audioquest wire might sound good to you, I’d still recommend laying it out better within the chassis. Looks a little ‘thrown together’, also, crossing the transformers like that puts them directly in the flux field, which is not great either.

As for the box, you really need to look for a reasonable metal one. There are some nice designs and builds mentioned and available on the DIYForum.com .
 
The reason for the cap choices as seen is purely that the sonic characteristics was proven best in this configuration through trials and testing.

I am not worried about placement of the wires since they are fully shielded and I actually like the way they are placed in the case. Everything was carefully placed so if it looks haphazard, then that is just perception. If you notice, every wire is in an almost direct straight line to the point it needs to go to hence shorter lengths. Also, Some prefer the look of woodwork. And I enjoy it.

I may never use metal cases.

Thanks for the suggestions though,
Dominick
 
JonFo,

Just curious why you say I "need" to look into metal enclosures? Or were you just suggesting?

Thanks,
Dominick
 
So, I forgot the Ampere ratio and had to do my math again.

This amplifier delivers 9.5A to every channel. ESl's and woofers.

What do you think? High enough?

Dominick
 
Hi Dominic,

First, sorry about the word ‘need’, it’s your design and you drive it based on your choices.
I just feel that metal enclosures confer some advantages in terms of inbound and outbound RF shielding of the amp and surrounding gear. That’s what I was getting at.
Looks and aesthetics are very personal as well, and my preference has usually been for the ‘high-tech’ look. I do like the blue lighting you used, that’s cool.

The layout using shortest path is also understandable, I just wanted to validate the placement alongside high-flux devices which could potentially impact sonic qualities. But as thick ( and expensive) as that green wire is, it better be shielded or twisted pairs or both ;)

The current output at <10amps is what I would personally have an issue with. My experience has been that it generally takes >40amps capacity to truly wrangle the ESL into submission. Also, stable current output regardless of impedance is important as well. This is generally reflected in specs where the amp doubles its output with every halving of the impedance load.

For instance, in my rig, the Sunfires are direct coupled to the ESL transformer. No Crossovers, no resistors, just a cable. The amp output is wired to the taps on the ESL transformer. That’s a pretty reactive load, and it takes massive current to drive the big Monolith panels. Sunfires rarely get warm, but mine have gotten hot after some spirited sessions at very high SPL’s.
A Sunfire Cinema Grand signature outputs around 40amps of current per channel.

My ideal amp would still be a digital, the TacT Milleium III. That puppy delivers a full 100amps of current into low impedances. I also like its digital tracking volume control approach.
 
I am a little confused with this. I know sunfire is a reputable company, but to achieve 40A, I would need 2 Torroids each capable of 1760VA. I have never seen anywhere close to that kind of power (3520VA) in ANY amp. Check my math please.

My amp. 22V @ 420VA

420/22 =19.1A

19.1A/2 =9.5A I must divide by two because there are 2 secondary windings, right? Each chip/channel splits the current? So each channel actually sees 9.5A!??? Or do they both still see 19.1A???

Your proposed amp
?VA/22V =40A

22x40 =880VA

880x2 =1760VA (Again-2 channels per torroid)

1760x2 =3520VA (Multiply by 2 torroids to get 4 channel VA need)

880VA per channel?? That does not make sense? Even the monster Krell Amps that ML uses to demo there equipment don't come close to this rating. I don't know everything, but I remember seeing the inside of a monster 2 channel Krell and one thing noted was the huge 1000VA torroid. To get a rating of 40A in that amp, the voltage would have to be running at 12.5??

I wonder if you could find out what the actual transformer(s) is/are rated at inside your sunfire. Companies like to embelish, but watts in can NEVER exceed watts out. I'm not picking a fight at all so please do not take offense...Just trying to learn.

Thanks,
Dominick
 
Forgive me for interjecting Dominck, for my technical knowledge is by far too lacking to provide much meaningful input on this subject. However, because I saw another posting referencing Sunfire equipment, I was prompted to do a web search on Sunfire, and just happened to run across a link that may provide an answer to some of your questions about how the Sunfire works....hopefully :) If not, then perhaps JonFo will be able to get you what you need.... But hopefully this helps.

http://www.vxm.com/21R.69.html
 
Interesting, I guess a DIY'er is doomed to the inefficient realm of amplifiers. The sunfire grand cinema is a unique piece and appears to be very economical. I wonder what kind of amperes Krell and Levinson amps put out?

Dominick
 
Dominick22 said:
I am a little confused with this. I know sunfire is a reputable company, but to achieve 40A, I would need 2 Torroids each capable of 1760VA. I have never seen anywhere close to that kind of power (3520VA) in ANY amp. Check my math please.

My amp. 22V @ 420VA

420/22 =19.1A

19.1A/2 =9.5A I must divide by two because there are 2 secondary windings, right? Each chip/channel splits the current? So each channel actually sees 9.5A!??? Or do they both still see 19.1A???

Your proposed amp
?VA/22V =40A

22x40 =880VA

880x2 =1760VA (Again-2 channels per torroid)

1760x2 =3520VA (Multiply by 2 torroids to get 4 channel VA need)

880VA per channel?? That does not make sense? Even the monster Krell Amps that ML uses to demo there equipment don't come close to this rating. I don't know everything, but I remember seeing the inside of a monster 2 channel Krell and one thing noted was the huge 1000VA torroid. To get a rating of 40A in that amp, the voltage would have to be running at 12.5??

I wonder if you could find out what the actual transformer(s) is/are rated at inside your sunfire. Companies like to embelish, but watts in can NEVER exceed watts out. I'm not picking a fight at all so please do not take offense...Just trying to learn.

Thanks,
Dominick

Hi,
You are using continuous current specs. Most of the amp manufacturers base their max current specs on short term current peaks, and those are delivered by the massive main filter capacitors, augmented by the transformer.

Peter
 
So what formula can I use to obtain the proper current (peak) output?

I'll start with the 50V 10,000uf Nichicon Super Through caps. There are two for one channel.

Can we figure out the peak current?

Thanks,
Dominick
 
Tracking Downconverter

Dominick22 said:
I am a little confused with this. I know sunfire is a reputable company, but to achieve 40A, I would need 2 Torroids each capable of 1760VA. I have never seen anywhere close to that kind of power (3520VA) in ANY amp. Check my math please.

[some math, see above]

880VA per channel?? That does not make sense? Even the monster Krell Amps that ML uses to demo there equipment don't come close to this rating. I don't know everything, but I remember seeing the inside of a monster 2 channel Krell and one thing noted was the huge 1000VA torroid. To get a rating of 40A in that amp, the voltage would have to be running at 12.5??

I wonder if you could find out what the actual transformer(s) is/are rated at inside your sunfire. Companies like to embelish, but watts in can NEVER exceed watts out. I'm not picking a fight at all so please do not take offense...Just trying to learn.

Thanks,
Dominick

Dominick, not a problem, you are not the first to disbelive the abilities of the Sunfires. I myself had questions as well many moons ago.

I'll refer you to the Sunfire WhitePaper on the Tracking Downconverter technology which is at the heart of Bob Carvers design.

But not to make this about Sunfires vs your design. I think the LM based designs are cool. I just like to have more amperage available for ESL's. But then maybe I'm not understading the capabilities of the LM's. Are there comerical designs based on these?
 
Dominick22 said:
So what formula can I use to obtain the proper current (peak) output?

I'll start with the 50V 10,000uf Nichicon Super Through caps. There are two for one channel.

Can we figure out the peak current?

Thanks,
Dominick

Hi,
The two specs of the capacitors you are interested in is the ESR (equivalent series resistance), and maximum ripple current. According to the specs on this page:
http://www.panasonic.com/industrial/components/pdf/pic_ts-ha_dne.pdf
(for Panasonic caps, Nichicon doesn't seem to give these specs...but it should give you an idea, as far as typical numbers go).

The 10,000uF @ 50V cap lists as having a .033 ohm ESR. To get the max peak discharge current, divide the voltage by the ESR value. For this capacitor, 50 Volts divided by a .033 ohm ESR gives 1515.15 amps maximum.discharge current. This would be into a short circuit, and would never be reached in actual use, but it shows that the capacitor can deliver a VERY high current, for short bursts. The resistance of the output transistors (and any protection circuits), wiring, and the impedance of the speaker would limit this to a much lower value.

The second important spec is the maximum ripple current. This lets you know the maximum current that it can handle over a long period of time. Since the capacitor does have some series resistance, it is not quite 100% efficient during charge/discharge, hence some energy is converted to heat. The ripple spec gives you the maximum current it can handle, long term, without overheating. This is listed at 4.00 amps for the 10,000uF @ 50V cap. If you use an 8 ohm speaker, and ohms law, you come up with 128 watts peak output at maximum ripple current (multiply by .707 for sine wave RMS watts = 90.5). This means the cap should be safe, ripple current wise, for amps up to this many watts output (more than this, and you could overheat the filter caps with sustained full power output). Audio normally will not require sustained full power output, but you should design for worst case...this would be one difference between 'supermarket stereos', and high end gear, the cheap stuff is built to a price, and they would save money by using smaller, cheaper filter caps, and running them beyond their limits at full output...this would work OK for average audio use, but if you drive it hard for an extended time, it may fail for reasons shown above.

HTH,
Peter
 

Latest posts

Back
Top