ETI 477 article

[sandyK]

ETI5000 mods

Anthem
Email me, and I will reply with a higher resolution photo. As for the other capacitor mod, it was as suggested by David Tillbrook in a later issue for stability purposes,IIRC.
I do not remember the details. It was quite a while back !

SandyK

alexkethel@optusnet.com.au

 

[Anthem]

Thanks SandyK but I cannot email you since I am a new user and the forum rules do not seem to allow me to do it. You may be allowed to email me since you are a more senior member. Not sure how it works, but I would appreciate the higher res photo.

Also, if you or anyone on the forum is interested, I have scanned a relatively clear pdf copy the entire ETI5000 amplifier from ETI June 1982. There seems to a few requests to get more info on it. Quite big, about 6Mb pdf file.

If I was allowed to post it on your site somewhere it would be OK with me.

 

[too_tall_kiwi]

bringing an old post back from the dead

Hi all, First time poster here.

I found this thread through googling AEM6000 - reason being that I have 4 6000 monos, which were built by someone else, and are generally very good ( albeit very, very over engineered ) for what I want, which is running fully active speakers ( mids and tweets ) I use Perreaux's finest for the bass.

Anyway, one amp has developed a 1.8 volt DC offset. This I discovered as I found one of the tweet diaphragms totally popped out from the faceplate ( Focal Audiom TD5's ) By some stroke of good luck, the tweet was absolutely fine once it was reseated.

Anyway, I have schematics for this amp, and friends who are capable of deciphering said schematics, but they are asking for the correct procedure to alter the DC offset bias ( RV1 I believe, the 20K in the feedback look? ) Apparently some of them have had bad experiences re biasing some amps, and have had the magic smoke come out.

Is anyone able to post the instructions for this, or, alternately email me the full instructions to set it up correctly.

Many thanks, I hope someone who is able to help is still reading this forum!

Attached is an image of the amp. Caps are 4 * 10000micro caps, 1.2Kva transformer potted, and some grossly oversize internal wiring!

 

[suzyj]

Yes, RV1 is indeed used to set the output DC level.

Simply leave the input and output open circuit, power the amp up and leave it to come up to temperature, and measure the output with a voltmeter while adjusting RV1 to zero it.

It's possible that the DC bias is a symptom of something else being stuffed, or it may simply be a dry trimpot. In any case give it a go and see if it comes good.

And I disagree, these amps aren't overengineered. Granted, they're a complex beast, but it's all good. I'm listening to mine (or at least my take on Tilbrook's design - see http://www.littlefishbicycles.com/poweramp/index.html ) as I write this, and they're just perfect.

 

[too_tall_kiwi]

Thanks for the reply, The amp is performing fine on a scope, so I think that its just an adjustment. It has not been used much since new, to be honest. When I said over engineered, I was referring to my amps - not the design in general. 16mm internal wiring, 1200VA transformer...........

I will take to it with my scope and a screw driver tomorrow.



Any idea how these amps perform spec wise at 4ohm? or even 2 ohm ( if they will? ) My amps have a full width 2u heatsinks along the front. Ugly, but they never get very warm.


PS. your contact link on your website wont work for me, I had seen your work there and you did a nice tidy job.

 

[too_tall_kiwi]

a (not very good) pic.

 

[too_tall_kiwi]

Thanks for the reply, The amp is performing fine on a scope, so I think that its just an adjustment. It has not been used much since new, to be honest.

Just to clarify - its performing fine bar the DC offset!.

This amp has only ever delt with 2Khz upwards, on pretty benign tweeters, so its not likely to have ever been given a true workout.
I dont know if this makes any difference to expected lifespans of devices.

Another question - the VR2 bias - is this effectively setting how much of the amps operation is in class A or am I getting the wrong end of the stick?

 

[suzyj]

Thanks for the reply, The amp is performing fine on a scope, so I think that its just an adjustment. It has not been used much since new, to be honest. When I said over engineered, I was referring to my amps - not the design in general. 16mm internal wiring, 1200VA transformer...........

Yeah, that transformer is pretty huge. You won't have much in the way of supply droop.

Originally posted by too_tall_kiwi Any idea how these amps perform spec wise at 4ohm? or even 2 ohm ( if they will? ) My amps have a full width 2u heatsinks along the front. Ugly, but they never get very warm.[/B]

Mine are fine into 4 Ohms. A lot depends on what your power supplies are. I chose +/-56V, which was a compromise with two pairs of output devices between available power at 8 Ohms and acceptable dissipation at 4 Ohms. My main speakers are Infinity RS-5b's, which from memory are around 6 Ohm.

They won't do so well into 2 Ohms. There just aren't enough output devices.

Originally posted by too_tall_kiwi PS. your contact link on your website wont work for me, I had seen your work there and you did a nice tidy job. [/B]

Yes, I'm aware of that. Unfortunately despite the disclaimer on the site, I was getting a ridiculous number of inane emails asking me for stuff, so I removed it.

 

[too_tall_kiwi]

thanks suzy

Well, thanks to you for your advice, my amp now has a DC offset of about 10mV. Nothing thats going to damage a tweeter!. I have left it running with the scope recording just to make sure nothing changes, will do so for the next 24 hours or so. I really dont want to have to fork out nearly NZ$800 for a new tweeter!.

I have been looking at your preamp... interesting that a DIY would go as far into logic as you have, very well done.

I am currently trying to figure out what I actually want / need for preamp duties, as I currently have a HFA valve preamp ( NZ made ) and a Rotel RSP 1069 processor. Lets say that the valve preamp is a LOT nicer sounding. However, it has no remote, limited inputs ( I agree with your point of having plenty! ) and of course, no digital processing or video capability.

I mostly get my music from my Logitech Transporter now.

My question for you is have you ever contemplated having an internal DAC in your preamp? Specially a buffering / re clocking DAC.

 

[VHF man]

Re: bringing an old post back from the dead

Hi all, First time poster here.

I found this thread through googling AEM6000 - reason being that I have 4 6000 monos, which were built by someone else, and are generally very good ( albeit very, very over engineered ) for what I want, which is running fully active speakers ( mids and tweets ) I use Perreaux's finest for the bass.

Anyway, one amp has developed a 1.8 volt DC offset. This I discovered as I found one of the tweet diaphragms totally popped out from the faceplate ( Focal Audiom TD5's ) By some stroke of good luck, the tweet was absolutely fine once it was reseated.

!

If you are bi-amping with an active crossover it's still a good idea to use a large Metalized Polypropylene coupling capacitor to block DC at the output. That way if your amp fails or goes faulty you don't loose the tweeters. The original 6000 kit does have DC protection but by the time DC has been detected and the relay has operated your tweeters will be history.

 

[suzyj]

Good to hear.

A DAC is most definitely on my to-do list. I've been listening to music (and movies and TV) a lot through my HTPC, which alas doesn't have room for a decent sound card, but does have an SP/DIF output.

I've been thinking of something like a TI DIR9001 receiver, and a PCM1798 or similar DAC...

I really should finish the remote control software, too, as that was my main reason for using a processor

 

[too_tall_kiwi]

Re: Re: bringing an old post back from the dead

If you are bi-amping with an active crossover it's still a good idea to use a large Metalized Polypropylene coupling capacitor to block DC at the output. That way if your amp fails or goes faulty you don't loose the tweeters. The original 6000 kit does have DC protection but by the time DC has been detected and the relay has operated your tweeters will be history.

I am aware of the fact that I have absolutely NO DC protection in the system at all, and had been wondering about how to prevent future disasters. There is also that turn on spike, its enough to throw the mids a fair few mm from rest, I cant imagine that would do tweeters any good whatsoever!. What value cap should I be looking at? Series connection I can only assume? What will this do to the phasing?

Thanks

 

[too_tall_kiwi]

Good to hear.

A DAC is most definitely on my to-do list. I've been listening to music (and movies and TV) a lot through my HTPC, which alas doesn't have room for a decent sound card, but does have an SP/DIF output.

I've been thinking of something like a TI DIR9001 receiver, and a PCM1798 or similar DAC...

I really should finish the remote control software, too, as that was my main reason for using a processor

Suzy, you might like to have a look at this Kiwi based thread, much discussion has gone into getting a good signal from a computer, and SP/DIF is probably not it!

http://audioenz.co.nz/forums/showthread.php?t=8523

I am a fan of the Logitech Transporter / Duet system, its simple, easy, works nicely and not overly expensive in the duet / squeezebox iterations. Look into it.

You could no doubt build a proper power supply for a duet or squeezebox, which apparently makes a considerable difference to the sound quality and dynamics.

 

[suzyj]

Suzy, you might like to have a look at this Kiwi based thread, much discussion has gone into getting a good signal from a computer, and SP/DIF is probably not it!

http://audioenz.co.nz/forums/showthread.php?t=8523

I am a fan of the Logitech Transporter / Duet system, its simple, easy, works nicely and not overly expensive in the duet / squeezebox iterations. Look into it.

You could no doubt build a proper power supply for a duet or squeezebox, which apparently makes a considerable difference to the sound quality and dynamics.

I must confess I disagree with Antipodes on this. By streaming the music over ethernet, you've got to have a whole lot more digital processing power, memory buffers, etc at the DAC than you do for USB, which in turn needs more logic at the DAC than SP/DIF, which only requires a simple receiver/reclocker.

In my experience, problems with PC audio are generally either digital garbage from the processor etc being picked up in the analogue side, either through poor power supply isolation, ground plane stuff-ups, etc., or else timing issues. Timing issues, in turn are either clock jitter from poor clocking practices or gross errors caused by the OS tying up the processor for long periods of time and starving the audio stream of data. Oh, and there's problems with codecs arbitrarily resampling stuff, but that's more a software problem.

The digital noise is easy to deal with. Put the minimum amount of digital logic in the box that's got your DAC in it as possible, and pay proper attention to power supply separation. think that SP/DIF can accomplish this.

Timing issues are sorted by reclocking with a reasonably low jitter clock at the DAC, and by not running windows to deal with the gross errors.

My HTPC runs mythtv on Ubuntu Linux, fwiw. I've found that to be pretty good.

Edit: Actually, reading further down the thread, it appears we might be on much the same page. Here's a pic of the innards of my HTPC:

http://www.flickr.com/photos/suzy_j/2628909361/in/set-72157603437130437/

 

[too_tall_kiwi]

I dont entirely agree with Antip on this subject either, and he appears to be taking a no holds barred approach.

He also claims that his $2K/Pair interconnects are far superior to anything available, whereas I am happy with far, far cheaper i/c's, and yes, I have tried a few!

However, Computers tend to dump far too much crap into the S/P DIF interface in my experience. I have used both ubuntu and windows, ubuntu is better than windows, but the network music players have a considerably better result _in my experience_

It depends on what your after of course, and what your system is capable of resolving. My speakers could be considered a cross between a Grande Utpoia ( Same tweeter, same wmtmw config, same midrange, 13" w cone bass drivers ) and a transmission line of some description - the mids and tweets are 1/4 wave transmission line enclosures built from carbon fibre, the bass cabinets are simple box enclosures, albeit they contain 3 sheets of 24mm HDF per box. As such, their capable of resolving the difference between most things pretty damn well ( too well sometimes ) and I am therefore very fussy about cables... But I dont see the need for a cable to cost what some people consider them to be worth!

 

[VHF man]

Re: Re: Re: bringing an old post back from the dead

I am aware of the fact that I have absolutely NO DC protection in the system at all, and had been wondering about how to prevent future disasters. There is also that turn on spike, its enough to throw the mids a fair few mm from rest, I cant imagine that would do tweeters any good whatsoever!. What value cap should I be looking at? Series connection I can only assume? What will this do to the phasing?

Thanks

The minimum value of coupling 'C' will depend to some extent on the crossover design and of course the load (tweeter) impedance but you are right to say that you want it large enough not to affect your designed crossover slope. You could use a single series cap as an element of your crossover - but if you prefer 'all active' then then a rule of thumb guide is to choose a value of C where XC at the crossover frequency is at least 1/4 the value of the tweeter impedance. For example, if your F crossover = 2500Hz and your tweeter impedance is 4 ohms then a minimum value for C is 68uF. If you are blessed with the ability to hear the affects of capacitors then you will want to choose a high quality metalized polypropylene type which will be reasonably expensive but probably much less expensive than your tweeters.

If your tweeter is fed via a series resistor greater than 3 ohms you can use a choke in parallel as an alternate measure.

If you can provide some details of the crossover and tweeter specs then I can run a quick simulation and let you know with greater accuracy what minimum value cap is recommended.

 

[too_tall_kiwi]

Crossover is a linkwitz 3 way, at 200Hz and 2kHz. The a design from "EE"mag, september 87 apparently.

The tweeter is a focal Audiom TD5
5.85ohm DC resistance and 1.15mH inductance.

As for it being fed via series resistor, well.. Is that in the amp?

I dont know if its called being blessed with being able to hear caps.

Maybe its called lunacy?

And remember its active, so the crossover will not be touched by the effects of the cap, just ( hopefully ) the DC and low frequency removed from the tweet.

 

[audiopip]

MOSFET Models

Sorry folks, sort of going back to the original thread , can SuzyJ let me know where she obtained the Spice Models for the SK/SJ Mosfets?

I am assuming they were from the audio community, rather than Hitachi or Renesas or what ever they are called this week?

My understanding is that none of the Hitachi lateral FET models in circulation are that accurate. The SemiLab/Exicon ones are just as bad, if not worse.

Hoping that Hitachi has finally come up with some good ones!

Thanks!

 

[suzyj]

I have no idea where I got them, as it was a long time ago, though it may have been from the exicon website, as they're right next to a couple of Exicon MOSFETs in my standard.mos file.

Here they are though:

.MODEL 2SK1058 NMOS (VTO=403.969M KP=20U L=2U W=29.7482M GAMMA=0 PHI=600M LAMBDA=184.988F RD=60.8251M CBD=2.56138N IS=10F CGSO=1.13517N CGDO=1.13517N TOX=0 NSUB=0 TPG=1 UO=600 RG=50 RDS=1MEG )
.MODEL 2SJ162 PMOS (VTO=842.193M KP=20U L=2U W=21.3317M GAMMA=0 PHI=600M LAMBDA=20.7067M RD=837.199M CBD=2.96862N IS=10F CGSO=1.13517N CGDO=1.13517N TOX=0 NSUB=0 TPG=1 UO=600 RG=50 RDS=1MEG )

They weren't terribly bad though.

 

[audiopip]

Spice MOSFET models

Thanks.

Same as mine.

As you say they work, probably the models are accurate, its just the parts that vary....

BTW love the bike frames- fantastic job!