• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

compression?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi!

Today i got a visit from a friend who works as a sound engineer.....
I let him listen to my amp....the first thing he said was: the sound is compressed

(like altering it with a compressor which you use for P.A)
The only thing i can think of is....the coupling caps

(0,47uf/450v)

The pre-amp/splitter has been improved to the one in the schematic (5965)
The coupling caps never have been changed....but i have no idea if this is the problem or even wich value they should be

i could really use some suggestions
 

Attachments

  • annanu.gif
    annanu.gif
    14.8 KB · Views: 371
Hi there,

@martinab2 - i looked into that but i don't see the current go down at any volume or music style (the 82uf only has to deliver 10ma)

@robert morin - hmmm, you could have a point there...strange 5ma is way in the curved area of the graph....why could he have chosen this??
I have done a try at lowering the 39k resistors to 22k. (on both sides of the amp....)
Results....well...now there runs 7,9ma per side in the 5965
bias: -1,5
Va=145v

It did change somehow...but i can't decide yet if it's better...it sounds just like if the sound has to go through a 'curtain' before it reaches your ears

Today at a radio fair someone gave me another solution....
Try an ecc83 instead of the 5965

I only had 1 ecc83 but i don't think the results are accurate because the lm317 provides current for right&left channel.
It looked like the ecc83 was 'stealing' from the 5965 because its anode voltage dropped to 77v, the ecc83 rose to 235v...and began to run quite hot. the sound was'nt good either
Hopefully i can try with 2 ecc83's sometimes.

@svokke: could you find anything in the psu???
 
hi

nothing sure but maybe:

1-your friend listen all day very neutral transistor sound

2-change 3.3k by irf820 to create a small 300v regulator
like me easy to put inside your box

3-jump up 300uf to 500 or 600

4-remove c4 c5 = -gain but +linear just for see

5-upgrade 317 ccs to a good transistor ccs like c4s
but the reverse one version for cathode side
 
@robert morin

i tried biassing the tubes to approx 8,2ma (7,9)
I lowered the anode resistances from 39k to 22k
bias is now -1,56v
VA is 155v
7,8mA per triode

I also lowered the 3k3 resistors to 1k5

The sound seems a bit clearer but still has the 'curtain' effect....bass is MUCH less, busy music gets a bit garbled up
Hmmm,after some time these 22k resistors run quite hot

Maybe it's my speakers....

@jeapel

I'm going to do the ccs for sure, just have to save a bit money :)
 
Uhhh...

Compression is noted, and the first thing you suspect...are *coupling capacitors*!?

Yeesh, what are audiophools writing about these days!? omg...(trails off mumbling)

No, compression comes from resistance. One way is power supply resistance, another is bias. A PP amp played side by side with an SE amp may be perfectly balanced at low volume, but the SE drops out at higher SPL due to the output stage beginning to cut off, moving operating point towards class AB, increasing current draw - subsequently increasing bias voltage, which reduces gain. In the case of PS resistance (excessively lossy inductors or intentional resistance), plate voltage sag occurs, though not as significant as the amplified effect at a self-biased cathode.

Fixed bias inherently does not compress as much for this reason.

All in all, I'm with SY. You get the most out of things LOUD. My Metallica collection knows this. :D

Tim
 
SY said:
Sound engineer? Your problem is probably more related to the volume control knob and its setting. Sound engineers listen to things LOUD. A wimpy little tube amp in a home setting won't satisfy them.

Slow down hot dog! When I get home from my columns of JBL VerTec line array, digital crossovers (can't believe hi-fi guys use those things by CHOICE) and mega-watt amplifiers I switch on my Darling. Only thing that keeps me sane!

Philip
 
Kathodyne,

like Tim (Sch3mat1c) already said, go for fixed bias.
Besides his arguments it will sound better cause you loose the 220uF electrolytic out of the signalpath.
No matter what TriodeDick says, fixed bias is better and sounds better.

Back to your problem, did you check your output trans. ?
Is it 6k primary ? did you connect it right ?
I spoke to another dutch diy some time ago who also build this amp and he was very disappointet with the sound. He found out that for one thing his output transformer had the wrong primary impedance. Don't remember where and how he got it, but it's worthwhile to check this.

Dick.
 
hi,

@sch3matic(@dj middelkoop for the trannies)
Well... i changed a lot to this amp. output transformers from 10k that caused oscillation to 6k that don't. upto this new splitter stage...and the power supply.
I got some REAL good technical help from mattijs de vries (www.machmat.com)
to be honest i'm just a beginner with not so much technical background and the only thing i hadn't changed where the coupling capacitors
And after reading how they can affect bandwidth....i made an assumption....
but after reading some more it seems you can just 'set' them at where they begin altering, usually dis is thone way above 20khz and somewhere around 10 to 20hz...

I am led to think this has been properly done in anna's first design so (assuming) even with these new tubes it wouldnt give a dramatic difference

@robert morin
I know now that (if i understand mattijs's explanation). the lm317 at the cathodes of the 5965 provide an almost infinite rk. The choice is to use litlle current into a big load...so that the load line gets more horizontal... and it doesn''t matter much where the tube is being biased...

changing the bias point of the 5965 didn't help, it merely got things running quite hot:
The biggest mentioned here are:

- remove autobias which puts a big elektrolytic in the signal
* ok worthwile AND nice to try...the auto-fixedbias design from www.machmat.com (diy-kit)

- use ccs with a transistor.....(or tube???!!!) instead of the lm317


It seems i now know what i have to be doing!! buy stuff!!!!

Thanx a lot!


:- my: guess....within 6 years....NO part of Anna will have been in the original one -:
 
Ok you're making progress here.
Try the fixed bias from Matthijs.
Matthijs is very knowledgable, following his advise will shurely help.

You're right with the coupling caps, the values are ok, leave them for now.
When your amp start sounding better you can try other brands to hear the sonic impact.

I don't use the 317 for CCS but use transisitors or (mos)fets.
So I cannot tell you if this will make an improvement.

Dick.

:- my: guess....within 6 years....NO part of Anna will have been in the original one -:
:D
 
The first thing I would be doing is changing C4 and C5. Spend a bit of hard earned cash on some 470uF/100V Blackgate Standards.

As noted above cathode bias is always going to impart a "compressed" sound compared to fixed bias - me, I like it that way and have converted a couple of amps from fixed bias back to cathode bias. Make sure that YOU don't like it that way before paying any attention whatsoever to your audio engineer friend.

Cheers,
Ian
 
Hi!,

Thank you all for your tips!

This morning i tried something else....i removed the 100ohm resistors from g2....

Well believe it or not, but this did do something....it clears up!
just to remove something!!:cannotbe:

this in combination with the automatic fixed bias module which will arrive next week will do a lot of good, i'm sure!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.