[Seano]

....I have some queries and observations.

I spent all day yesterday on the damn thing.  Which was way too long.  Especially when it was so pleasant outside.  But it is done. For now.

But I have two observations that concern me and force me to ask the obvious question....why?

First question. Gain. More precisely the adjustment thereof. When you are gunning for your 55mV gain do you just take the initial gain at switch on (ie.cold) or do you let it idle for a bit? After all the surgery, the gain started at 43 and I bumped it up to 55. Checked T2 and all was sweet then went back to T1 and we were up over 60 and climbing.....eventually got it down to 55 after ten minutes or so but nearly wound P1 back to where I started (or so I thought). This happened to both boards.  Feed it some tunes (Massive Attack's 'Mezzanine') and managed to get the heatsinks well over room temperature. After resting for ten minutes, checked the reading across T1........30mV punched it back up to 45mV as suggested. And that's about it. Have done no more adjustment. Should I? Do I need to? Basically I don't understand the gain adjustment process so I have no idea whether all is good/bad/indifferent. Can someone explain in laymans terms what the go is just to set my mind at rest?

Second question/observation. I'm currently using a Yamaha integrated amp as a pre-amp. The AKSA is now hooked up as the amp (previously I'd bi-amped using the Yammy for the high range). If the AKSA is on and I kick the Yammy over I get a very sharp crack through the speaker - not loud but not real soft either - definitely not pleasant. Same goes when I switch off. Turning the CD on produces the same soft whump it always has. Turning the AKSA on last stops the crack but turning it off last doesn't (that said though there's still power in the AKSA even after a couple of minutes).  What the heck is going on? This crack is something I've not heard before despite using the Yammy for quite some time as a pre-amp. It appears to crack in time with the Yammy switch relay but why would this be a drama now and not before.  And this feature did not present itself at anytime during testing. Any ideas?

Other observations? Well it looks very sweet. It still hums a little at idle. The Jaycar binding posts are four pieces of crap. Neatly wiring an AKSA is obviously an art. The bespoke AKSA case makes for a lousy testbed. The amp is good enough straight off the bat to flatter a very cheap and well battered Akai portable CD player (I always thought the drum rim beats that starts MA's 'Teardrops' were accompanied by almost complete silence - they are not and I think you can now hear a lot of the artefacts within the samples used in many of the tracks - vinyl hiss anyone?).

Overall happy enough but still to close to the frustration of yesterday to be overjoyed. When I get a chance to give it a listen and feel the results I'm sure my mood will change.

[AKSA]

Hi Sean,

Thanks for your post.  Your problems are minor, nothing significant.

First question. Gain. More precisely the adjustment thereof. When you are gunning for your 55mV gain do you just take the initial gain at switch on (ie.cold) or do you let it idle for a bit? After all the surgery, the gain started at 43 and I bumped it up to 55. Checked T2 and all was sweet then went back to T1 and we were up over 60 and climbing.....eventually got it down to 55 after ten minutes or so but nearly wound P1 back to where I started (or so I thought). This happened to both boards. Feed it some tunes (Massive Attack's 'Mezzanine') and managed to get the heatsinks well over room temperature. After resting for ten minutes, checked the reading across T1........30mV  punched it back up to 45mV as suggested. And that's about it. Have done no more adjustment. Should I? Do I need to? Basically I don't understand the gain adjustment process so I have no idea whether all is good/bad/indifferent. Can someone explain in laymans terms what the go is just to set my mind at rest?

I believe the word you want here is bias;  gain is fixed by the feedback resistors and bias is the standing current which flows at idle in the output stage.

First, you must ensure that the thermal coupling between each and every output device and the heatsink ledge is good.  You need a thin film of the paste on both sides of the mica washer, and the same torque on each device's 3 mm bolt which holds the transistor, ledge and pcb together.   And the Vbe transistor, the BD139, must be on top of T7 and screwed down with the metal portion against the power transistor.  (No paste required here.)  So, check all are evenly tightened.

To set bias, switch on, let it idle for about an hour, THEN set the level.  Ensure there are no strong draughts around the heatsinks;  don't do it with the back door open and a gale roaring through the house!

Bias does change slightly depending on ambient temperature.  Set it as near as possible at 25C (77F) as this is a good compromise.

Second question/observation. I'm currently using a Yamaha integrated amp as a pre-amp. The AKSA is now hooked up as the amp (previously I'd bi-amped using the Yammy for the high range). If the AKSA is on and I kick the Yammy over I get a very sharp crack through the speaker - not loud but not real soft either - definitely not pleasant. Same goes when I switch off. Turning the CD on produces the same soft whump it always has. Turning the AKSA on last stops the crack but turning it off last doesn't (that said though there's still power in the AKSA even after a couple of minutes). What the heck is going on? This crack is something I've not heard before despite using the Yammy for quite some time as a pre-amp. It appears to crack in time with the Yammy switch relay but why would this be a drama now and not before. And this feature did not present itself at anytime during testing. Any ideas?

The input cap on the AKSA is now larger.  This lessens phase shift at very low frequencies, another way of saying you hear low frequencies better.  Downside is that this makes the amp more susceptible to low frequency thumps at switch-on.  Your Yamaha preamp has a voltage spike at its output at switch-on.  That's all.  It won't do any damage, and my advice would be to leave everything on all the time.  It might cost you $5 a year in electricity, but the advantage is that your amp will always be hot to trot.

Now, be prepared for some strange sounds as the upgrade burns in.  This takes about 120 hours and is normal.  But the sounds the amp will make will be quite strange;  muffled bass, strident top end, etc.  It starts out fine, then goes bad.  But at the finish it will be much better than now.  And detail - you will be astonished.

The AKSA will have power for some time after switch off because of the considerable energy stored in the filter caps.  This is normal.  With the new power supply pcb you should have bleeder resistors (8K2) for each of the four rails.  This will take about ten minutes to bleed off all voltage from these caps at switch off.  Again, normal.  The bleeders run warmish, of course.

Cheers,

Hugh

[Rocket]

Hi Hugh,

I am just curious to know how long does it take when the amp has been switched off to sound it's best ?

Regards

Rod

[EchiDna]

Hi Hugh,

I am just curious to know how long does it take when the amp has been switched off to sound it's best ?

Regards

Rod

what do your ears tell you Rocket? mine tell me about 20+ minutes from 'cold' - if you can call it that here in Singapore where everything is 32 degrees celcius all the time

[AKSA]

Yes Rod,

I'd say around 20 minutes;  that's been my experience too.

Cheers,

Hugh

[Seano]

To set bias, switch on, let it idle for about an hour, THEN set the level. Ensure there are no strong draughts around the heatsinks; don't do it with the back door open and a gale roaring through the house!

OK.  Now we've got it sorted. Hugh, I suspect this might be where your instructions need some more detail. This ideal of letting the amp idle for a half hour before fiddling with bias is not mentioned in the upgrade instructions but did kind of occur to me at the time. I was also hampered by my inability to find the original construction instructions (too many moves between then and now).

Given your input I suspect that all is good then.

As for leaving the power on...........not a chance. Especially during summer. We are 50km from the nearest sub-station, we have numerous properties between us and the sub-station - many with substantial electrical plant (irrigation pumps, coolrooms, machinery etc.), it gets ridiculously hot, summer storms are common-ish........basically the supply wobbles all over the place, brownouts and blackouts are frequent, line noise is likely (hundreds of k's of wire makes for a great aerial) and we are vulnerable to lightning strikes.  I've already lost a TV to the power. I am not losing an amp if I can help it.

Will the GK-1 give me the same hassle with voltage spike?

[kyrill]

OK.  

As for leaving the power on...........not a chance. specially during summer. We are 50km from the nearest sub-station, we have numerous properties between us and the sub-station - many with substantial electrical plant (irrigation pumps, coolrooms, machinery etc.), it gets ridiculously hot,.

Wow
What would you benefit from http://www.transcendentsound.com/power_supply.htm
or
http://www.bluecircle.com/index.php?menu_id=70
or when money is no object
http://www.soundapplication.com/html/review2.html

Can tea taste purer than the water you made it from?

[Seano]

Yep I'd probably benefit from one of those toys. But at this point it would be cheaper to simply turn the amp off. Until I can find someone in Oz who can provide something similar as a kit or even just a pcb and a parts list then it's not going to happen. I am not spending $1000 to fix something that I can fix just as easily by switching the amp off. I'm more inclined to spend the money if I can 'hear' the result but I suspect I won't.  

I'm flat out hearing the break-in process of the N+ version.  So far all have stated that some madness is expected but all I think I can 'hear' is a lack of oomph or passion or body. At the same time though we think wecan both discern some more detail...overall we think it sounds better but we can't put our finger onwhy. So how would we be with something more subtle such as a power conditioner?

Thing is, all the stereo gear does go through one of these kits which is a really basic noise filter kit that costs AU$42 from Altronics.
http://www.altronics.com.au/index.asp?area=item&id=K6045

It'll have to do until someone in Oz comes up with something better.

[Rod]

Hi Seano,

Yes I'm surprised Rod Elliot hasn't done anything in this regard. Perhaps because it's mains related, although some of his other kits are.

I wanted to ask you, how does the Altronics filter perform in relation to voltage spikes and ground loop?  I have some severe ground loop issues with  several of the GPO's in the house, and short of rewiring them something like this might make them available for use again.

Regards
Rod

[kyrill]

..1)But at this point it would be cheaper to simply turn the amp off. ... I am not spending $1000 to fix something that I can fix just as easily by switching the amp off.
2)  I'm more inclined to spend the money if I can 'hear' the result but I suspect I won't.  

3)I'm flat out hearing the break-in process of the N+ version.  So fa ...

Most pwr line conditioner will NOT protect you for summer storms. They are build fo a complete different reason. To turn the amp of is NOT an alternative approach
2) If you hear the differences between capacitors brand, Sean. you will
3) So you do hear differences.., so you will  

[Seano]

Yes I'm surprised Rod Elliot hasn't done anything in this regard. Perhaps because it's mains related, although some of his other kits are.

So am I but I suspect it has something to do with demand rather than the mains drama.  I reckon power tends to be one of those black magic/snake oil things for most people and if not enough people want them then why design/build them

I wanted to ask you, how does the Altronics filter perform in relation to voltage spikes and ground loop?

To be honest, I have absolutely no idea.  Not a clue. I have tried both with and without and not heard nothing.  To be true I'm not sure I even have a problem in the first case.  I built the thing three houses ago on a whim - I was learning myself to solder - and figured it couldn't hurt. Might as well build something I can use as opposed to a 'Clifford the Cricket' or some such. I suspect you'd have to be dead sure you have a drama in the first place and then just try it.  www.soundlabsgroup.com.au offer a slightly fancier LC Audio filter kit for $100 too.  In the end, $42 and some soldering practice is a pretty cheap test though cause how much would it cost to get a techie to find out for you?

[rabbitz]

The AKSA will have power for some time after switch off because of the considerable energy stored in the filter caps. This is normal. With the new power supply pcb you should have bleeder resistors (8K2) for each of the four rails. This will take about ten minutes to bleed off all voltage from these caps at switch off. Again, normal. The bleeders run warmish, of course.

When you put bleeders on the power supply, just make sure you disconnect the power supply from the amp PCB. I learnt the hard way and had to get some new transistors from Hugh as I ended up spiking a couple. Thanks to Hugh's great diagnostic instructions I was able to find the problem quickly.

[ginger]

Seano,
No dramas with switching your gear on/off as often as you like.

The big trick (well small trick really) - the AKSA should be the last thing you turn ON and the first thing you turn OFF. That way you get AKSA'a own "Mini-Thump"s BUT none from your other gear.

Some posts above say your AKSA power supplies will hold up for minutes after turn OFF suggesting that the amp remains "active" for this time. That doesn't in fact happen. The postive power supply rail collapses nearly all the way quite fast (less than 1 second) which effectively makes the amp OFF in this same short time. Thumps from other gear being turned OFF WON'T get thru' the amp to the speakers.

The negative power supply rail DOES hold up for quite some time which was a problem for guys like me who want to turn it off and shove a soldering iron in its guts 10 seconds later. This is why the bleeder resistors are there.

Cheers,
Ginger

[andyr]

... The negative power supply rail DOES hold up for quite some time ...

Hi, Ginger,

I had noticed that my '-' rail LEDs stayed on for a lot longer than the '+' ones did and wondered why.  Are U able to hexplain to us ignorant folk?  With my limited understanding of such matters I would've thought '+' and '-' should behave the same way??

Regards,

Andy

[Seano]

Thanks Ginger
This is the approach I've been taking and it works. Funnily enough I get way less thumping than I used to when I turn on and the amp is actually way quieter at idle than it used to be. I would only have to spin the volume on the Yammy a short way before to get a background hiss/hum. Now I have to go quite a way round the dial to get anything. Very very nice.

I suspect that the upgrade of the PS alone would be a very worthwhile (but still rather pricey) step even before the rest of the Plus.  

S'funny. When I was testing the new power supply on its own it took ages for the LEDs to go out yet now that it is connected to the boards the LEDs collapse in mere seconds. That took a while to notice cause it is not easy to see the LEDs when they are lit so I had the notion that it took some time for the amp to draw down...

[ginger]

Andy said:

I had noticed that my '-' rail LEDs stayed on for a lot longer than the '+' ones did and wondered why.  Are U able to hexplain to us ignorant folk?  With my limited understanding of such matters I would've thought '+' and '-' should behave the same way??


Well here is what I think happens:
The +ve rail is different in that we have D1, R10, C3 network which doesn't have a Complement on the negative rail.

When power is removed the main +ve and -ve rails start to collapse (evenly) as output transistors continue to conduct their bias currents.

Initially this has no real effect on the Long Tail Pair (LTP) consisting of T1 and T2 or the Volatge Amplifying Stage (VAS) consisting of T3 as their positive supply is prevented from collapsing with the main +ve rail by the action of D1 - which won't let current flow back to the output transistors.

As the -ve rail collapses further and the positive supply from C3 starts to collapse (at a slower rate) T3 (the VAS) will start to turn off - there is less voltage drop across R5 and R6 and so the positive side driver T5 will turn on harder and therefore the postive side output transistors will be turned on helping collapse the main positive rail faster than the main -ve rail.

Why doesn't this result in a thunderous great "Thump" with voice coils being lauched into orbit as they achieve escape velocity?

We have already noted that the LTP and VAS positive supply is from C3 and therfore holds up much longer  (due to D1) - also the main negative rail tends to hold up longer than the main positive rail. The Amp is therefore still under at least partial control as DC feedback from the output to the LTP to keep the output at zero volts is still attempting to operate.

The result is that the amp tends to shut down "gracefully" rather than quit with a bang when the power is turned off.

While in this process it tends to dump stored charge from the main positive supply much more quickly than the main negative rail.  

Some of you may have noted the "Zero Cost Improvement" thread where one of the changes we attempted was to delete D1. One of the MAJOR problems we encountered (and still haven't solved) is that this mod results in a BIG turn off THUMP. That mod is accordingly "still on the drawing board" and is NOT recommended. Aside: We haven't quit on it totally BUT just haven't solved all the problems associated with it yet.

Hope some of this makes sense AND a note of caution - these are my thoughts based upon 5 minutes thinking about Andy's question. If I'm wrong it wouldn't be the first time and certainly won't be the last.

Cheers,
Ginger

[andyr]

Thanks, Ginger,

Makes sense ... and I ain't gonna do a thing about that "Zero Cost Improvement" until I hear that it doesn't produce negative results!

Regards,

Andy

[Rom]

Ginger,
very nicely explained and I been learning a lot through most of your post
and of course Mr AKSA as well and most of the AKSAphile.
Cheers mate,
Rom