'AK Mods' SC Class A 15W Amplifier

[Deleted]

Hi Chris,

Alan,

Confirmed here too

That pleases me no end,
Regards,
Alan

[Deleted]

Feb 5, 2012 22:21:15 GMT drymdrum said:

That pleases me no end,
Regards,
Alan

You are very welcome

[Deleted]

Jan 25, 2012 13:19:18 GMT @shaun said:

Jan 19, 2012 18:33:23 GMT @shaun said:

Jan 8, 2012 11:11:25 GMT @shaun said:

Hi All

just to keep this available.

for the LS device GB so far i have

Will 2x LS352 and 2X LS313
Phil 2x LS352 and 2X LS313
Me 2x LS352 and 2X LS313
Syd 2x LS352 and 2X LS313
Alan 4X LS352 and 4X LS313
Alex 2X LS352 and 2X LS313
Mark 2X LS352 and 2X LS313 (cyteen)
Jeff 2X LS352 and 2X LS313
Jon C 2X LS352 and 2X LS313

Kits GB

Chris
Syd
Phil
Will
mark (cyteen)
Jon C
me


Boards only GB

Alex
Alan
Jeff

just to remind people

any additions just let me know.

aaaannnnnndddddddddd

after a week of having the benefit of Alex's patient guidance he found the reason for the high DC offset i was getting.

it was errrr STUPIDITY.

i soldered the MPSA in the wrong way around.

after Alex's superb tuition i now have a good grip on the circuit and most 8-)of the fault conditions to look for.

so mistakes can have a positive outcome.

I'VE LEARNED PLENTY.

I now have 1.2mV and 1.8mV respectively with TL071 so not quite as good as Alex's stellar results but nothing shoddy either.

way better than most commercial offerings

thanks for all of the help and advice Alex

I'd have been lost without it.

take care

[Deleted]

Oh and another thing

anyone thinking of building the amp without bothering with the Loudspeaker protection board needs to think hard about that course of action.

they saved my speakers from becoming TOAST when things went pair shaped.

let me see a few £ for the loudspeaker protection kit from jaycar
verses £550 to replace my speakers.

it's a no brain-er

take care

[Deleted]

let me see a few £ for the loudspeaker protection kit from jaycar verses £550 to replace my speakers.

In Australia, this kit is no longer available over the counter from Jaycar. Altronics and their distributors are still stocking it.
I am fortunate in that Altronics Sydney store and warehouse is only 10 minutes drive away from me. This kit is also recommended for any DIY high powered Class AB amplifier, not just Class A designs.
Alex

[Deleted]

Feb 6, 2012 21:26:59 GMT @sandyk said:

let me see a few £ for the loudspeaker protection kit from jaycar verses £550 to replace my speakers.

In Australia, this kit is no longer available over the counter from Jaycar. Altronics and their distributors are still stocking it.
I am fortunate in that Altronics Sydney store and warehouse is only 10 minutes drive away from me. This kit is also recommended for any DIY high powered Class AB amplifier, not just Class A designs.
Alex

Feb 6, 2012 21:26:59 GMT @sandyk said:

let me see a few £ for the loudspeaker protection kit from jaycar verses £550 to replace my speakers.

In Australia, this kit is no longer available over the counter from Jaycar. Altronics and their distributors are still stocking it.
I am fortunate in that Altronics Sydney store and warehouse is only 10 minutes drive away from me. This kit is also recommended for any DIY high powered Class AB amplifier, not just Class A designs.
Alex

Hi Alex

yup i forgot that the LS protection kit is only available from the Internet catalog. it's well worth the cash though and i'm soooo glad i used one.

i replaced the TL071's in the offset corrector with AD744 and it did make some difference to my offset figures.

i now have

-0.4mV and -0.7mV respectively so a little lower than the TL071.

IMHO the TL071 does a pretty good job especially for the money but for those wanting to drive that offset down as low as possible (and for an amp of this quality it makes sense to go for the best) the extra money is well spent on the AD744.

so still not as low as your Stella's but pretty respectable none the less.

I've also had the LM394H (legs did the twist as did the 3811A (and the MPSA )) running in the Amp for a few days now and ERR interesting.

when i first installed them they seemed IMHO a little rolled of but now they've had a few days to cook that aspect seems to be better mmmmmmmmmm .

i also noticed how much better the bass had become with the 100uf removed from the FB loop and the offset correctors installed.
overall clarity also improved quite a bit.

so now i have something to compare with once I've built Wills boards using those new fangled LS devices and SF12.

should be interesting and fun.

take care

[Deleted]

Hi All

I've had Alex's Class A on test for the last two weeks with 8R2 5W resistors strapped to it's OP (and a little music playing also )

just cycling on for a few hours and then off twice a day .

All looks good with no protests from Caps/Regs/Transistors so I'm ready for my final set up procedure.

balance front end and check OP.

mmmmmmm sounds great.

I'm now thinking on a pair of speakers to match the quality of the above.

I'd love to try some IPL's which i think would just be OK sensitivity wise and should be fairly flat impedance wise,

BUT

I'm in the small room at the moment 8FT X 12FT so I'm going to struggle to fit the IPL in. I'm hoping to annex a larger space at some point in which case sensitivity does matter.

SO

I'm looking for something either DIY (my woodwork is crap so cabs needed to be supplied with the kit) or commercial to pair with Alex's Class A.

trouble is with most small monitors is they:

A) lack sensitivity
8) tend to need more beef to drive them properly.

i used PMC OB1 and FB1 at college and I've had a sweet spot for them ever since. stunningly good speakers but that was with Bryston PA so I'd want a listen with what i have.

the FB1 are pretty sensitive 90dB 1W and a pretty flat 8ohm load so I'm tempted to look for a nicely used pair.

BUT

any other suggestions would be great as I've been a little out of tough with speakers.

My budget is around £1K5.

take care

[Deleted]

any other suggestions would be great as I've been a little out of tough with speakers.

I used to have a pair of Opera Prima stand mount 'speakers. Gorgeous little things, both in looks and sound. Solid wood, impressive bass for the size and well voiced/balanced.
At 89db and 4ohm they should be enough in the size room you have.

The latest version is also front ported, making them wall friendly.

Better than the small Sonus Fabers and at the time I bought them they were head and shoulders above anything in the same price bracket, only about £600 back then. Now over 1k eg here.

[Deleted]

Feb 18, 2012 18:31:00 GMT @cjarchez said:

any other suggestions would be great as I've been a little out of tough with speakers.

I used to have a pair of Opera Prima stand mount 'speakers. Gorgeous little things, both in looks and sound. Solid wood, impressive bass for the size and well voiced/balanced.
At 89db and 4ohm they should be enough in the size room you have.

The latest version is also front ported, making them wall friendly.

Better than the small Sonus Fabers and at the time I bought them they were head and shoulders above anything in the same price bracket, only about £600 back then. Now over 1k eg here.

Hi Chris

I've had a listen to those speakers and you are right in saying that they sound really good.

the only thing that bothers me a little is @ 4ohm they are going to need a little more power to drive them properly.

so

I'm trying to find a nice easy load for the class A which is how i arrived at TL's.

TL's tend to have a fairly flat impedance curve (mostly) but can sometimes struggle a little for sensitivity behaving a little more like a sealed box in terms of rear sound absorption from the drive units.

hope that make some sense

i guess what I'm getting at and i should maybe have asked to begin with is what are the ideal speaker specs for the class A.

I'm also trying to plan ahead a little and not have to replace the speakers again when i have more space.

sooo squeezing in some well behaved TL's is possible (just) and then Waite for space to become available (sort of jam tomorrow type thing)

take care

[Deleted]

Shaun,

what's the current (amperes) capability of the Aclass? At 15w/c driving a 4ohm load this is more important.
If the current is strong, say 30a plus, then the speakers will have roughly double the perceived volume for the same wattage.

[Deleted]

Feb 19, 2012 19:19:17 GMT @solderdude said:

Feb 19, 2012 14:14:27 GMT @cjarchez said:

Shaun,

what's the current (amperes) capability of the Aclass? At 15w/c driving a 4ohm load this is more important.
If the current is strong, say 30a plus, then the speakers will have roughly double the perceived volume for the same wattage.

care to explain ?

As our versions will be using separate PSU PCBs, and perhaps 2 transformers, we should be able to go very close to 30W into 4 ohms as stipulated .
Alex

Attachments:

[Deleted]

Hi Frans,
I’m glad you’ve jumped on this as I’ve always wanted it properly explained in terms I can grasp.

So it looks like I’m heading for stormy waters with insufficient knowledge but I’m game, full sails are up and the wind’s at my back.

Basically my understanding was that a well designed amplifier (‘speaker driving power amp) should double its measured output in watts RMS when the resistance of the loudspeakers is reduced from 8ohm to 4ohm. (assuming stable load!)

If the current available to the amplifier is sufficiently greater than that required at 8ohms then it stands a good chance of meeting the doubling of watts RMS into 4ohms, based on the extra current available preventing the voltage on the supply rails dipping below specification and in turn causing clipping.

Ref. This equation P=VI = I2R=V2/R

Before writing this I had a quick search around and can see things are not this simple but from what I’ve read the theory basically, though incomplete, held water as far as I could discern.

I look forward to your reply as this is something that has bugged me for some time.

Chris

[Deleted]

Feb 19, 2012 20:47:43 GMT @sandyk said:

Feb 19, 2012 19:19:17 GMT @solderdude said:

care to explain ?

As our versions will be using separate PSU PCBs, and perhaps 2 transformers, we should be able to go very close to 30W into 4 ohms as stipulated .
Alex

Hi Alex

thanks for the explanation which explains what i was clumsily trying to get at.

it's not only about max power but keeping the amp running in Class A for as long as possible that's on my mind.

with 4ohm as the article states 30-35W but it's going to slip out of class A into AB much earlier (7.5W) than with an 8ohm load.

I'm pretty sure that 4ohm will work just fine but I'm just thinking of the ideal here.

I have 6ohm at the moment so a half way house but I'd really like to try to optimize the speaker side of things to squeeze the max out of your Class A.

yup doubling the quietescent current would fix that up but then it's bigger heat sinks and running the regs and OP devices harder.

I'm thinking that the lighter load will keep things in the Class A sweet spot longer which i guess is what I'm looking for and why i love class A amps so much.

But

speaker matching is one of those puzzles which often involves some degree of compromise which is why i asked the question.

for me in my small room I'm thinking that a reasonable compromise sensitivity wise may be better than increasing the load seen by the amps OP devices.

a nice flat (ish) load is maybe what i should be looking for and do the best i can sensitivity wise (88-90dB @1w)

that's my guess but i may be wrong.

the main problem I'm finding is that once the requirements become fairly narrow the choice gets smaller errr exponentially.

so still thinking out loud.

Hi Chris

''Basically my understanding was that a well designed amplifier (‘speaker driving power amp) should double its measured output in watts RMS when the resistance of the loudspeakers is reduced from 8ohm to 4ohm. (assuming stable load!)''

Yup that seems about right but one of the flys in the ointment is that loudspeakers seldom provide a stable load.

nope they change the load according to frequency so rather than the fixed resistor value we tend to get one varied by frequency.

but please lets not talk about QQQQQ

take care

[Deleted]

Hi Frans

thanks for the nice lucid explanation

it helps things along

I've had a go at loudspeaker design and it seemed like a tricky can of worms to me as the poor results of my efforts revealed.

take care

[PinkFloyd]

Feb 19, 2012 22:44:19 GMT @solderdude said:

Feb 19, 2012 14:14:27 GMT @cjarchez said:

Shaun,

what's the current (amperes) capability of the Aclass? At 15w/c driving a 4ohm load this is more important.
If the current is strong, say 30a plus, then the speakers will have roughly double the perceived volume for the same wattage.

Basically the theory of doubling the power of a amp (if it can supply the current) when the load impedance halves is correct.

Suddenly you speak of current and mention 30a but you mean 30W.

Where it goes wrong is the assumption that a doubling of power constitutes a doubling of perceived volume.

a doubling of perceived volume needs a 10dB increase in power which is a 10x times higher power.
So if you have 15W on your speaker and want to double the 'volume' you need 150W
a doubling of power is only a 3dB increase.

As for class A things get different as indeed mentioned below.
an amp is class A UNTIL one of the output devices is switched off.
so with 1A class A current, due to Kirchhoff laws 2A peak can be delivered.
Since power is needs RMS (Root Mean Square) and not peak currents the peak current needs to be converted to RMS, simply by dividing through SQRT2 = 1.41ARMS
with a 4 Ohm load the class A power can be calculated:
I2xR= 1.412x4= 8W till it runs out of class-A
for 8 Ohm you can do the math and it doubles so the same amp is 16W class A in 8 Ohms.
6 Ohm = 12W class-A

To make the amp TRULY class-A in 4 Ohm the idle current must be increased.
Since the amp can deliver about 13VRMS on 20V DC supply this means the peak voltage will be around 18V peak a load of 4 Ohm will draw 4.5A.
So the idle current for a 30W class A should be at least 2.25A.
With 40V power supply voltage this means the output transistors will dissipate 90W.. per channel, not counting the losses in the regulator if used.
The power transformer (per channel) should at least be a 150W version if a constant 90W is drawn.
The size of the heatsinks must be considerably bigger if the same temperature is desired.

Any chance you can go on to explain what class "A-B" is? (before A goes to B)... class "A-B" is a zone seldom discussed but a zone that warrants further discussion as most of our listening is in the A-B zone

[PinkFloyd]

There is no such thing as pure class A (in real terms) at best we should be looking at A-B (combo!).... am I wrong?

[Deleted]

Feb 19, 2012 23:23:06 GMT PinkFloyd said:

There is no such thing as pure class A (in real terms) at best we should be looking at A-B (combo!).... am I wrong?

Hi Mike

good question

and I'm going to let my ignorance run riot here

OK I'll start with a quote

''As for class A things get different as indeed mentioned below.
an amp is class A UNTIL the drawn output current exceeds the idle current.
so with 1.5A class A current a 1.06ARMS current can be had.
Above it the amp is class B again.
with a 4 Ohm load the class A power can be calculated:
I2xR= 1.062x4= 4.5W till it runs out of class-A
for 8Ohm you can do the math and it doubles so the same amp is 9W class A in 8 Ohms.
6 Ohm = 6.75W class-A

so i think that the point is that class A (without slipping into A/B)is obtainable so long as we don't exceed the iddle current (1A in the case of Alex's class A)

Yes true most of us don't get much past 1W in our every day musical pleasures. in fact Nelson Pass has made a nice living (and rightly so) with his ''first Watt'' products.

I've used 88dB @1w 6ohm speakers with my 1.5W 45SE and it was plenty good SPL wise.

but is that only part of the story?

music is not a straight line and what about the loud bits (transients).

so I'm thinking that loudspeakers with a flat ish impedance curve and sensible sensitivity helps this amp to stay in Class A all of the time.

so i guess I'm thinking that you may be right and wrong depending on the specs of the loudspeakers used.

Class A in every circumstance and with every speaker no.

Class A with the right speaker yes

it's just where i am now thought wise

but I'm here to learn so feel free to correct me on the above.

take care

[Deleted]

Hi Frans and Shaun,

Thank you for the explanation, that's made things much clearer, especially as regards to A class itself. Guessing from manufacturer designs claiming 20/30/40/50 a class amplification there are other ways of doing things to keep the heatsinks smaller?

Suddenly you speak of current and mention 30a but you mean 30W.

Purely a coincidence of the number 30. 30a was a figure bandied about by manufacturers as a "good" level of reserve for amplifiers to handle transients, though for B class.

[Deleted]

Shaun
As an aside. My DCM QED speakers are 87dB/1W at 1M. Their rated impedance is 8 ohms, but they are actually 8 ohms minimum, and go from 8 to 12 ohms depending on frequency. I have never been able to use this amplifier at anywhere near maximum power in a normal domestic situation. A previous amplifier had a LED power display into these speakers, and rarely was more than a little over 1W exceeded.It has also been trialled in a friends larger dining room as part of a surround setup into big 4 ohm Ohm Walsh speakers with plasma tweeters, and still didn't go near maximum power.
Alex

[Deleted]

Alex,

That is reassuring on for my 'speakers rated as;

Sensitivity 88dB/1W/1m
Impedance 6 ohm (min)

Chris

[Deleted]

Feb 20, 2012 0:19:47 GMT @sandyk said:

Shaun
As an aside. My DCM QED speakers are 87dB/1W at 1M. Their rated impedance is 8 ohms, but they are actually 8 ohms minimum, and go from 8 to 12 ohms depending on frequency. I have never been able to use this amplifier at anywhere near maximum power in a normal domestic situation. A previous amplifier had a LED power display into these speakers, and rarely was more than a little over 1W exceeded.It has also been trialled in a friends larger dining room as part of a surround setup into big 4 ohm Ohm Walsh speakers with plasma tweeters, and still didn't go near maximum power.
Alex

Hi Alex

as there are so few people at the moment this side of the world that have the chance of using your lovely Class A, real life everyday experience is never an ''aside'' but most welcome.

your DCM's are Transmission lines?

Which is pretty much what i have in mind (as you've probably noticed) but the DCM are a rare beast over here in the UK and i think that the QED is quite rare full stop.

i think that Leo is using the Dancing Usher with his Class A 88dB @1w @1m 8ohm nominal beryllium or diamond tweeter (can't remember which).

so I'm still kind of torn between some small IPL TL's 86dB@1W @1m which are a good price and DIY or one of the PMC series OB1 87dB @1w @1m 8ohm or the slightly bigger brother 90dB 8ohm or the floor standing GB1i 90db 8ohm.

I've heard the PCM's and liked them much but they are pretty pricey.

it would be nice to have a longer list to think on though.

take care

[Deleted]

Feb 20, 2012 0:27:08 GMT @cjarchez said:

Alex,

That is reassuring on for my 'speakers rated as;

Sensitivity 88dB/1W/1m
Impedance 6 ohm (min)

Chris

Hi Chris

Snap on the speaker spec front

same as my Sonus Fabers which work well with Alex's class A.

take care

[Deleted]

Hi Shaun
It may be intersting to see if Leo feels the need for additional power from his Class A.
Regards
Alex

[Deleted]

Feb 20, 2012 1:18:23 GMT @sandyk said:

Hi Shaun
It may be intersting to see if Leo feels the need for additional power from his Class A.
Regards
Alex

Hi Alex

I'm pretty sure that Leo would have mentioned it if he felt under powered in any way. ;D

but as i said previously people who use WATTS as a measure of goodness just make me want to laugh.

maybe it's my tube history where matching speaker to amp was a little bit more obsessive.

for me and spending cash on speakers is just a big head ache and choice crisis soon creeps in.

so having the chat may help me avoid any BIG blunders.

so I'm really grateful to all the above contributions which have helped clear the waters a little.

yup decided on some of those nice Bose jobies ;D

take care

[Deleted]

Alex,
with regard to the air cored coil how much leeway do we have in the value around 4.7mh?
I'm heading out to get the coils I've wound measured and see if I'm in or out of the ballpark.
Regards,
Alan