'AK Mods' SC Class A 15W Amplifier

[Deleted]

Hi Shaun,
heat sinking........one 300mm X 100mm X 8mm full back plate for the Regulators ie. 2 PSU modules.
Two 300mm X 100mm X 8mm side plates for the OPs that's one module per side.
Each module plate will have either angle or U channel pop riveted along it's length as radiator fins. A 5000mm length of 25.4 X 12.7mm U channel will more than cover three plates at 11 sections per plate......a lot of holes and pop rivets. Lets see what transpires.
As Frans said it'll make a nice room heater.
Regards,
Alan

[Deleted]

Jan 25, 2012 23:17:21 GMT drymdrum said:

Hi Shaun,
heat sinking........one 300mm X 100mm X 8mm full back plate for the Regulators ie. 2 PSU modules.
Two 300mm X 100mm X 8mm side plates for the OPs that's one module per side.
Each module plate will have either angle or U channel pop riveted along it's length as radiator fins. A 5000mm length of 25.4 X 12.7mm U channel will more than cover three plates at 11 sections per plate......a lot of holes and pop rivets. Lets see what transpires.
As Frans said it'll make a nice room heater.
Regards,
Alan

Hi Alan

sounding good

take care

[Deleted]

Hi Shaun and Alan
I was lucky to get hold of the case for an E.A. Mosfet power amp years ago from a friend who was a Jaycar manager. It uses the heatsinks as the sides, and the heatsinks have tapped threads to take the top and bottom covers.
Regards
Alex

[Deleted]

Hi Alex,
yes heat sinks will make up the sides of my intended case, also the back, could be the front panel depends on how I see the input/output/power entry.
With regard to the JLH 2u2 caps, I have some 1u MKS 63v can I sub them for the 2u2?
Regards,
Alan

[Deleted]

With regard to the JLH 2u2 caps, I have some 1u MKS 63v can I sub them for the 2u2?

Hi Alan

They will work O.K. but 1u5 may be better and a little easier , and cheaper, to obtain than the 2u2.

Regards
Alex

[Deleted]

Hi Alex

my second 2n3811A arrived this morning HFE 342 measured with my Atlas.

so within spec

i hope yours will be OK also and I'd be interested in there HFE measurement. they took around 10 days to arrive here in the UK so you should get yours pretty soon.

just waiting on the LM394H now as i plan to do those upgrades one board at a time.

take care

[Deleted]

Jan 27, 2012 9:56:25 GMT @shaun said:

Hi Alex

my second 2n3811A arrived this morning HFE 342 measured with my Atlas.

so within spec

i hope yours will be OK also and I'd be interested in there HFE measurement. they took around 10 days to arrive here in the UK so you should get yours pretty soon.

just waiting on the LM394H now as i plan to do those upgrades one board at a time.

take care

Hi Shaun
I haven't used the LM394 in the Class A, only in a phono preamp, so I am not sure whether it will be an improvement there or not.
It does have a relatively high value of capacitance between it's 2 collectors though.The LM394 achieves it's low noise figures because it actually consists of multiple transistors in parallel.
The LS313 MAY be a better performer there, as it has a minimum hFE of 400, Ft of 250mHz at 1mA, and only 2 pf capacitance between it's collectors.
Kind Regards
Alex

[Deleted]

Jan 27, 2012 10:28:26 GMT @sandyk said:

Jan 27, 2012 9:56:25 GMT @shaun said:

Hi Alex

my second 2n3811A arrived this morning HFE 342 measured with my Atlas.

so within spec

i hope yours will be OK also and I'd be interested in there HFE measurement. they took around 10 days to arrive here in the UK so you should get yours pretty soon.

just waiting on the LM394H now as i plan to do those upgrades one board at a time.

take care

Hi Shaun
I haven't used the LM394 in the Class A, only in a phono preamp, so I am not sure whether it will be an improvement there or not.
It does have a relatively high value of capacitance between it's 2 collectors though.The LM394 achieves it's low noise figures because it actually consists of multiple transistors in parallel.
The LS313 MAY be a better performer there, as it has a minimum hFE of 400, Ft of 250mHz at 1mA, and only 2 pf capacitance between it's collectors.
Kind Regards
Alex

Hi Alex

thanks for the tip

i have some extra 313 penciled in so i'll wait for those.

I'll slip the rest in on Monday.

take care

[Deleted]

hi all

Here are some pictures of my Alex Lite SC Class A build.

Frogs in a box time again ;D

The PSU

Hifi2000 U2 dissipante steel case



The transformers are mounted on 3mm Aluminum sheet isolated from the chassis and separately earthed back to the IEC input.

Power output sockets are IEC Female to avoid any exposed voltages should the plug become disconnected.

The reg boards are my etched versions so may be a little larger than the ones Will has in mind.

I have 20.2V + & – on both reg Ops

Power chords are silicone rubber 16A mains cable (it’s sooo much less stiff than the regular stuff)

The Amp

HIFI2000 U4 dissipante Ali case

Uses 10A power cables to hook up the OP devices and 6A for the JLH’s except for the long runs from the JLH which i used screened IC cable for as per Alex's build.

The amp boards and DC offset boards are taken from Will’s designs posted on the HA/PRE thread (thanks Will).



Of all of my etches, I’m most pleased with the DC offset correctors as they are pretty fine tracks which I doubted the press and peel would cope with. It did a really good job on those boards. The etching process was fun and I was amazed to end up with working boards at the end of it.

by the way i used TL071 in the corrector which gave 26mV and 27mV which is pretty respectable. i had 10 to choose from and while some where lower i just went for the best match.

Internal signal IC are Shark wire OFC copper in a pseudo balanced configuration. It’s double screened (foil and braid) as I wanted to make sure that noise was kept to a minimum (and in keeping with the amp ‘’it’s Australian’’). I may experiment a little with that later.

I have fitted the speaker protection board which gives around 5 seconds delay on start-up and switches off pretty quickly avoiding start up and shut down thumps and bumps. Initially I was going to run the AC for it from the PSU (note UN needed XLR socket) but in reality I just soldered a link from the positive DC input to one of the AC inputs as per the testing procedure outlined in the build guide. I also fitted a quick switch off on the thermal shut down connection as per Alex's build.

Speaker OP cable is 2x 10A PTFE silver on copper stranded per run which I use for my LS cables.

JLH boards are Greg’s fantastic work. Thanks Greg.
The chassis is UN earthed but I have fitted a binding post to allow that option if needed.

I found that it was not and it runs very quiet indeed.

So as I say this amp is a real joint effort and all the better for it.
It also happens to sound superb (well stunningly good actually)

Take care

[Deleted]

Hi Shaun,
I just love that you've gone down the "Real DIY" trail, right out there in front of us all.
Frogs in a box indeed, that's worthy of a place in any modern art gallery.
Your etched boards certainly look the business I'm really pleased to see you getting first class results from your etching.
You've given us an in depth run down on your whole project that really helps so many thanks for the great tutorial, which I'll certainly refer to as I get to grips with my own.
Great work Shaun...."four in hand" going like the clappers.
Regards,
Alan

[Deleted]

Hi Alan

I'm glad you like the work but I'd never have started the etching thing without your help and encouragement.

as for the tutorial

well lets see what the boss says.

i used my usual reading plenty and wild guess work approach so i may have got some things wrong.

i did try to stick pretty much to the spirit of Alex's Class A HA/PRE method. especially with the grounding which is only common between channels at the IEC socket. so both amp boards are star grounded from their star ground points back to the IEC where they meet. with the tall case it did give me scope for plenty of separation between signal and power inputs which made life easier.

i also managed to tap some threads into all of the heatsinks and fitted the Regs and OP devices on insulated pads with 12mm M3 hex bolts. i also used some nice wide washers to distribute the pressure more evenly into the devices. I'm not sure if it made any difference but it felt good.

i was grateful to have Johns metalwork thread to guide me and boy did i need guiding on that front.

Wills artwork made life so much easier especially with the DC corrector and i would have been struggling without it.

Alex has done all of the hard work with his hours of experimenting and tweaking and spent quite a bit of cash in the process no doubt.

some may say that it was an existing design

well yes and no.

comparison to the original design shows how many changes have been made along the way.

The front end balance (which i still need to upgrade) made a noticeable difference to the HA/PRE and I'm sure that the SF diodes are going to do the same trick with the PA.

Alex's work has transformed this Amp into something of beauty.

I'd say that giving all of that away is pretty generous and shows the mark of the man.

me well i just did the easy dot to dot and follow the leader stuff.

so as i say

a real joint effort

Alan I'm looking forward to sharing the sound with you and as i say all good things taste better that way.

keep up the good work

it really is worth the effort.

take care

[Deleted]

And while I’m thinking on it

I know that all of you people involved in the group build are old lags but more people read these threads than just us

So some tips on how I build my circuits.

Do I trust myself not to make mistakes? HELL NO.

So firstly preparation is EVERYTHING.

Have the circuit diagram in front of you when building anything.

Don’t trust manufacturers labeling of resistor packs as they do get it wrong from time to time.

I use a plastic pot for each resistor value. I measure all resistors going into the pots and measure again taking them out to solder to the board.

Same thing with transistors (harrumph I did misplace one on this build so it can happen) check the numbering on the component to make sure it is what you think it is. I also use a plastic pot for each value and type.

Same with caps and diodes.

Check twice solder once

One thing at a time and don't rush it's not a race

Flatter components like links and resistors first on the board.
Then work your way up size wise

Use good quality solder it really is worth the extra money.
I use cardas for all of my builds and it is just a pleasure to use.

A 10x loop is pretty useful for getting a good look at those transistors and checking that the solder joints you’ve made are nice and bright with no dry joints (dull).

If you are not sure ask as help will always be forth coming here at least. It’s much harder to remove components than to solder them in.

I was thinking about when I first started and how lucky i was to have good advice without being made to feel foolish.

I hope that the above may help anyone stumbling on our nice group build.

Feel free to join us

Take care

[Deleted]

Hi Shaun
Very nice work, and good construction advice in your other post.
The DC offset figures are a little on the high side. Did you have the inputs terminated with a suitable value resistor when you measured them ?
Kind Regards
Alex

[jonclancy]

Utterly excellent!!

[Deleted]

Jan 27, 2012 19:55:32 GMT @sandyk said:

Hi Shaun
Very nice work, and good construction advice in your other post.
The DC offset figures are a little on the high side. Did you have the inputs terminated with a suitable value resistor when you measured them ?
Kind Regards
Alex

Hi Alex

no resistor on the IP just measured the offset at the OP.

I'll load her up and re measure,

what value would you recommend?4k7 maybe

the DC offset did go higher when the servo was disconnected so i just assumed that they where working OK.

take care

[Deleted]

Jan 27, 2012 23:20:07 GMT @shaun said:

Jan 27, 2012 19:55:32 GMT @sandyk said:

Hi Shaun
Very nice work, and good construction advice in your other post.
The DC offset figures are a little on the high side. Did you have the inputs terminated with a suitable value resistor when you measured them ?
Kind Regards
Alex

Hi Alex

no resistor on the IP just measured the offset at the OP.

I'll load her up and re measure,

what value would you recommend?4k7 maybe

the DC offset did go higher when the servo was disconnected so i just assumed that they where working OK.

take care

Hi Shaun
3.3K should get you fairly close,and be close enough to let the offset corrector take charge.
Regards
Alex

P.S.
I was thinking here of the Class A HA/preamp.

As the input has a series 1.8K after the 15K , just put a s/c across the input.

[Deleted]

Jan 28, 2012 0:07:48 GMT @sandyk said:

Jan 27, 2012 23:20:07 GMT @shaun said:

Hi Alex

no resistor on the IP just measured the offset at the OP.

I'll load her up and re measure,

what value would you recommend?4k7 maybe

the DC offset did go higher when the servo was disconnected so i just assumed that they where working OK.

take care

Hi Shaun
3.3K should get you fairly close,and be close enough to let the offset corrector take charge.
Regards
Alex

Hi Alex

yup you may be right and i have a problem with the DC offset.

i have no 3k3 so just used a 4K7 to load the IP and the offset went up to 35mV.

I've checked the voltages to the chips and all seems fine

the fact that both chips are doing the same thing suggests that it's a mistake that I've made on both correctors.

sooo could be that i have the offset corrector connected to the wrong part of the board. I'm fairly sure that i have the OP properly located connected between the base of Q2/18K/1.2K. but the IP i have connected before the 6R8 1W resistor which i thought looked right but maybe not.

any pointers of where to look would be great

take care

[Deleted]

Hi Shaun
The input is taken from the actual output of the amplifier, but I doubt that this would make much difference.
It is interesting that the offset was reduced when you connected the offset corrector. Have you measured the offset when the amplifier is working into an 8 ohm load or speakers ?
If it is reasonable, perhaps just wait until you install the duals ?
If you are concerned , you could always temp fit input caps until then .
Regards
Alex

[Deleted]

Hi Shaun
This is a closeup of that area of mine.
Regards
Alex

Attachments:

[Deleted]

Hi Alex

this is the circuit that i have worked too



it should give some idea of where my offset is connected

take care

[Deleted]

Hi Shaun
The main differences are the type of input and CM devices, and an SF12 instead of the 1N5819, and 200 ohms emitter resistors for the input pair as wello as 200 ohms for Q103 emitter resistor. The circuit you posted is the circuit I used until recently, and there shouldn't be any problems.The one in the photo is with LS devices from the original GB, and those other small changes.
I just measured my DC offsets at the speakers with DTV audio via the SC DAC, and both channels were flicking around -.1mV to
-.2mV in between voices.Your circuit is as per the previous version with Q104 base resistor as 150 ohms. Come back at it when you are refreshed.
Make sure the 2 sets of emitter resistors are the same value.
I remember that someone slipped there once, using 150 on one side and 100 on the other.That's highly unlikely though.
Regards
Alex

[Spirit]

Quick question...
Does the GB include offset corrector boards/components, or are we making/sourcing our own?

[Deleted]

Jan 28, 2012 6:24:24 GMT Spirit said:

Quick question...
Does the GB include offset corrector boards/components, or are we making/sourcing our own?

Hi Phil
I think you are given the choice of either, or both, PSU and main PCBs, LS devices, or kits for the main PCB ? Perhaps JLH PCBs too ?
You would need to go back and check that post.
Will's Class A PCBs incorporate the offset corrector as in the HA/preamp.
Alex

[Deleted]

Jan 28, 2012 3:03:28 GMT @sandyk said:

Hi Shaun
The main differences are the type of input and CM devices, and an SF12 instead of the 1N5819, and 200 ohms emitter resistors for the input pair as wello as 200 ohms for Q103 emitter resistor. The circuit you posted is the circuit I used until recently, and there shouldn't be any problems.The one in the photo is with LS devices from the original GB, and those other small changes.
I just measured my DC offsets at the speakers with DTV audio via the SC DAC, and both channels were flicking around -.1mV to
-.2mV in between voices.Your circuit is as per the previous version with Q104 base resistor as 150 ohms. Come back at it when you are refreshed.
Make sure the 2 sets of emitter resistors are the same value.
I remember that someone slipped there once, using 150 on one side and 100 on the other.That's highly unlikely though.
Regards
Alex

Hi Alex

I've had an idea for a quick fix

we could swap boards ;D ;D ;D

don't worry I'm on the case and it's going to be something small.

take care

[Deleted]

we could swap boards

Yes
Send them over with Jon. and I could send mine back with Jon.
IN YOUR DREAMS !