Building a Monster... Class A

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
bear said:
if the thing that is getting hot is that aluminum bar that goes between the two toroids, it is because you have a magnetic circuit, not an electrical one?

It's actually an electrical circuit, magnetically induced. Didn't take too long to figure out, when the supply *used* to work before that little pcb above it was mounted. It's a relatively easy fix - at least nothing needed to be physically *moved*.
 
Adrculda said:


I look at is as constructive criticism, and also you guys are giving me great ideas on how to maximize my budget which i have set aside about $2500. if i can get even more out of it than that's even better !!!

Nope...
my grandfather before he passed away said a couple of words to me that opened m teenage eyes ( at the time ) most of the not pertaining to anything here but one:
"A smart man learns from other peoples mistakes"

:D

First off, see if you can get this thing built way under $2500. Start by searching out cheap sources of heat sinks, caps, and trafos. And you may need to adjust your plan to build around what you can actually get. If you can't do that, then you might as well go mortgage the house and buy a half dozen new amps. It will be cheaper than DIY, even if you have to baby them running them 8 ohms and doubling up on amps.

Second, if you're putting this much into it, try to refine the amp design. Project 117 with some massaging will *work*, but you can do better in both efficency and sound quality if you put some research into it. The heavy iron will be pretty much the same in any case.
 
Electrovoice is reffer that EVX180B can give a max output of 128,5 dBSPL and that can handle 600W continuous. This has not the sense that only with 600Wrms this speaker can produce its maximum SPL. Maybe a robust class AB amplifier of 400Wrms/8Ù (20-20000Hz) restricted from an electronic x-over between 250 to 500Hz can do the job very well, because these folded horn-loaded speakers can touch only the 500Hz without "shrieks" in mid frequencies. So for 4 speakers you need only 2 stereo amplifiers of 2X400Wrms/8Ù each. Be it so 2X600Wrms/8Ù.
In the past, i had a rental system - at whole Peavey - in which included four UDH band pass subs. In these speakers, contained four BW15" of 350Wrms each. But Peavey was reffered that each 15" BW speaker to give its maximum SPL it needs 350Wrms. I used these tremendous subs (which can touch notes down to 35Hz with enough SPL, we had checked this with bass guitar) in a 4 way system, bellow 110Hz! Above this, 8 HDH3 and 8 HDH4. Ah! Old good days before class D amplifiers and speakers arrays.
Well, each UDH sub was driven by a CS1200X amplifier, which can produce 2X600Wrms/4Ù. The lack of yet 100Wrms needed theoretically for driving these speakers into their maximum SPL, had no sense because the cut-off frequency placed at 110Hz from active x-over.
Summarizing: for driving PA speakers, the use of class A amp it has no sense because these speakers produce enough distortion in big SPL by alone.
You may turn in class AB amps if they are DIY, unless in ready commercial class D. For DIY 2 X 350 to 2 X 400Wrms amps are sufficient. Don't drive your speakers in their limits. Leave a margin of 100Wrms per speaker. If the SPL produced is not enough, then add 2 more bass speakers. It is better to you have a clean sound than hundreds of SPL.
I used some times EV speakers driven from Crest amps, this conjuction can gives a high quality sound... but only if is driven in 90% - as much - of its maximum power level. Above this level, the system is going to heavy distortion.

Regs
Fotios
 
fotios said:
Electrovoice is reffer that EVX180B can give a max output of 128,5 dBSPL and that can handle 600W continuous. This has not the sense that only with 600Wrms this speaker can produce its maximum SPL. Maybe a robust class AB amplifier of 400Wrms/8Ù (20-20000Hz) restricted from an electronic x-over between 250 to 500Hz can do the job very well, because these folded horn-loaded speakers can touch only the 500Hz without "shrieks" in mid frequencies. So for 4 speakers you need only 2 stereo amplifiers of 2X400Wrms/8Ù each. Be it so 2X600Wrms/8Ù.
In the past, i had a rental system - at whole Peavey - in which included four UDH band pass subs. In these speakers, contained four BW15" of 350Wrms each. But Peavey was reffered that each 15" BW speaker to give its maximum SPL it needs 350Wrms. I used these tremendous subs (which can touch notes down to 35Hz with enough SPL, we had checked this with bass guitar) in a 4 way system, bellow 110Hz! Above this, 8 HDH3 and 8 HDH4. Ah! Old good days before class D amplifiers and speakers arrays.
Well, each UDH sub was driven by a CS1200X amplifier, which can produce 2X600Wrms/4Ù. The lack of yet 100Wrms needed theoretically for driving these speakers into their maximum SPL, had no sense because the cut-off frequency placed at 110Hz from active x-over.
Summarizing: for driving PA speakers, the use of class A amp it has no sense because these speakers produce enough distortion in big SPL by alone.
You may turn in class AB amps if they are DIY, unless in ready commercial class D. For DIY 2 X 350 to 2 X 400Wrms amps are sufficient. Don't drive your speakers in their limits. Leave a margin of 100Wrms per speaker. If the SPL produced is not enough, then add 2 more bass speakers. It is better to you have a clean sound than hundreds of SPL.
I used some times EV speakers driven from Crest amps, this conjuction can gives a high quality sound... but only if is driven in 90% - as much - of its maximum power level. Above this level, the system is going to heavy distortion.

Regs
Fotios


If i can squeeze 2 x 550W RMS @ 8Ohms, i would be more than happy ;)
 

GK

Disabled Account
Joined 2006
wg_ski said:
Second, if you're putting this much into it, try to refine the amp design. Project 117 with some massaging will *work*, but you can do better in both efficency and sound quality if you put some research into it. The heavy iron will be pretty much the same in any case.


Not in class A it won’t! You’ll need a fair few extra devices to manageably dissipate that amount of power. And the proposed +/-140V rails? A bit dicey for 250V output devices.
A single tiered monster amp with these output voltage swings should really be a bridged design.

There is one valid ‘class A’ HiFi approach for an OTT amp like this though. Rate it at 1500W into 4 ohms. Make it a bridged design with; say +/-70V rails.
Use a huge number of output devices, say 15 pairs per bridge side (30 pairs per output channel).
Bias it for an easily manageable (with hefty blown heatsinks) ~20W per output transistor at idle; say 250mA per pair.
Select emitter resistors for an emitter-emitter idle voltage drop of 100mV; 0.2 ohms in this case. Optimal (D.Self) bias requires ~50mV emitter-emitter (0.1 ohms) but this isn’t practical in such an amp because with such high voltage rails bias stability will be too difficult to control/maintain and there will be an unavoidable thermal gradient across the huge heat sinks required, so higher valued emitter resistors are required to better ensure current sharing at idle.
None of this really matters though because with so many parallel output pairs the effects of gm doubling a significantly diminished anyway.

All this will give you a 1500W/4 ohm amp that will dissipate about 1000W at idle per channel. It will deliver approximately 225W into 8 ohms in class A and be able to drive ridiculously low impedances, still with very low distortion.

Cheers,
Glen
 
Of course it won't work in class A. I don't think the OP is considering that anymore - especially now that we know it's for PA bass bins. The +/-140 is for my 3-tier design. One of the earlier class B amps running on on +/-127 used a stacked S-Leach type OPS with the same number of devices as Project 117. And would happily drive 2 labhorns in parallel all night without giving a ...., at least until the plug melted off the extension cord. New cord, and the bass thumped on.

The whole idea of not making it a bridge amplifier is so that you can bridge it. 4K into 4R.
 
PERFECT..
this is what i want guys ... keep it up...
none of this is set in stone
and after readding all this i just called my PCB guy and put my "project on hold"

No parts have been ordered yet besides the transformer that has been put on hold since they are awaiting my answer.

anyways
what do you guys think of high power darlingtons ??
i found a NPN/PNP set, SGSD100/200 ....
let me know what you guys think

THANKS
and keep it up !!
 
Adrculda said:



If i can squeeze 2 x 550W RMS @ 8Ohms, i would be more than happy ;)


Piece of cake. The old design will do THAT and sit there and laugh at you.
 

Attachments

  • 1500watt.jpg
    1500watt.jpg
    86.5 KB · Views: 830

GK

Disabled Account
Joined 2006
wg_ski said:
Of course it won't work in class A. I don't think the OP is considering that anymore - especially now that we know it's for PA bass bins. The +/-140 is for my 3-tier design. One of the earlier class B amps running on on +/-127 used a stacked S-Leach type OPS with the same number of devices as Project 117. And would happily drive 2 labhorns in parallel all night without giving a ...., at least until the plug melted off the extension cord. New cord, and the bass thumped on.

The whole idea of not making it a bridge amplifier is so that you can bridge it. 4K into 4R.


I was referring to Adrculda's amp and his +/-140V rails, as proposed in the opening post.
I sincerely hope he is no longer considering 1500W class A :hot:

Cheers,
Glen
 
G.Kleinschmidt said:



I was referring to Adrculda's amp and his +/-140V rails, as proposed in the opening post.
I sincerely hope he is no longer considering 1500W class A :hot:

Cheers,
Glen


HAHA
noticed i said 550w RMS at 8Ohms, not at 4 or 2 ;)
i think everyone overlooked that fact :smash:

EDIT to be honest ...
that was supposed to be class AB , but sure as hell didnt know where the B went off to
 
Member
Joined 2005
Paid Member
those high power darlingtons are well... not high enough power.
based on their voltage rating.

Quasi has a design for this sort of power here maybe some of this thread can be useful:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=53264&perpage=25&pagenumber=18

I attached a pic of the one i built.

2 x 500W rms into 8R

Quite a simple layout with nice results.

PS, you may wanna rethink the amount of parallel devices used here
at some point the driver stage will find it difficult to cope especially at higher frequencies. although i hope you do build this amp with 140V dc rails good luck with your build!!!!!!

also in my sig line is a 1kW (2 ohm) amp design if you are interested.

-Dan
 

Attachments

  • actrk600.jpg
    actrk600.jpg
    55 KB · Views: 838
Adrculda said:



HAHA
noticed i said 550w RMS at 8Ohms, not at 4 or 2 ;)
i think everyone overlooked that fact :smash:


Did not. There are 2 versions of that +/-127V amp. Stereo and monoblock. The Stereo version puts out 680 watts per channel at 8 ohms. The monoblock (pictured) around 1100 into 4, 1500 into 2 - limited by the incoming AC mains. The 'replacements' should be tickling the 1kW level at 8 ohms.
 

GK

Disabled Account
Joined 2006
Adrculda said:



HAHA
noticed i said 550w RMS at 8Ohms, not at 4 or 2 ;)
i think everyone overlooked that fact :smash:

EDIT to be honest ...
that was supposed to be class AB , but sure as hell didnt know where the B went off to


Oh I get it now. this whole thread is supposed to be a joke.

I just read your opening post again and in it you make no mention of "550W RMS at 8 ohms". Exactly why one would seek a 6kVA transformer for a 550W/8ohms class AB amplfier is certainly strange.

Also, can we see a picture of those fans and heatsinks that can dissipate 3500W without rising above 22degC? I'd really like to see that. Cooler climate where you are I guess.............
 
Adrculda said:
PERFECT..
this is what i want guys ... keep it up...
none of this is set in stone
and after readding all this i just called my PCB guy and put my "project on hold"

No parts have been ordered yet besides the transformer that has been put on hold since they are awaiting my answer.

I hope it isn't the 2x119V one linked earlier - how are you going to handle the +-170V DC rails? I'd say class AB stops being practical at much lower power levels...


anyways
what do you guys think of high power darlingtons ??
i found a NPN/PNP set, SGSD100/200 ....
let me know what you guys think

Um, 80V Vce max and SOA almost as bad as a 2N3055. No way! In a 100W/4 ohm amp or so it might work though...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.