'AK Mods' SC Class A 15W Amplifier

[Deleted]

Hi Shaun,

i've been driving my 88 dB 6ohm Sonus Fabers with no problemo.

are you using your own boards or Will's, for above speakers?
Regards,
Alan

[Deleted]

Apr 22, 2013 14:41:46 GMT drymdrum said:

Hi Shaun,

i've been driving my 88 dB 6ohm Sonus Fabers with no problemo.

are you using your own boards or Will's, for above speakers?
Regards,
Alan

Hi Alan

My home brew etched jobbies.

Shamefully enough I’ve not built up Will’s lovely board yet but plan to soon.

But I think that we could be onto something here with Will’s boards.

I’m guessing that having identified two key differences between the original and Will’s boards that it could give us the answer.

Let’s hope so.

As I say

I’ve had no problems driving the Sonus which are a 6ohm 88DB ported design similar to your sons so let’s hope that a quick fiddle will sort things.

Take care

[Deleted]

Hi Shaun,
well that's interesting I guess then you'r point of contact would be after the zobel and right on the output.

I didn't think I'd get at the no. 2 build till Tuesday but I picked it up this afternoon along with the 6 ohm speakers.

I might get back online later, see how the evening goes.

Regards,

Alan

 

[Deleted]

Well that was quick for me, attached the Wharfedale 6 ohm speakers to my first build and they sound very good.
By that you will realise that my hand cut boards are up to the job and taking the OSC input connection
from before the Zobel seems to be okay.
In one way that pleases me as I don't have to mess with my own SandyK 15 Watt.
Now if I had some 4 ohm jobbies to test it lower down that would be good.

So on now to no.2 build which has Will's boards let's hope Alex's suggestions get to the bottom of the instability on 6 ohms
as they are okay driving the 8 ohm TLs here in the den.
Perhaps a bottle of beer will top it off later.
Regards,
Alan

Just remembered the groundlift on no. 1 build is as specified 10R installed. 

[Deleted]

Okay got to the no. 2 build with Will's boards.....................
first thing I notice, the 10R ground lift resistors are burnt on both boards, see pic below.

When did I complete this amp, must have had it running solid since first week in January so it's had a good few hours seemingly working okay
into my 8 ohm TLs so why does it burn the ground lift when it is connected to the 6 ohm speakers if that's when they over heated.
I'm going for a beer and just relax for a bit.

 


 

[Deleted]

Hi Alan

Good detective work and yes that resistor looks like it's taken a bit of a beating.

The same on both Boards?

It’s hard to tell from the photo but is it a ½ W unit

Any chance of a MM resistance reading on that resistor to see if she has drifted or gone Open circuit (which is unlikely but interesting all the same).

Looks like Alex’s resistor bypass may be the next move.

Take care

[Deleted]

Apr 22, 2013 20:35:20 GMT drymdrum said:

Okay got to the no. 2 build with Will's boards.....................
first thing I notice, the 10R ground lift resistors are burnt on both boards, see pic below.

When did I complete this amp, must have had it running solid since first week in January so it's had a good few hours seemingly working okay
into my 8 ohm TLs so why does it burn the ground lift when it is connected to the 6 ohm speakers if that's when they over heated.
I'm going for a beer and just relax for a bit.

 


 

Hi Alan
This has been worrying me since I found out about the PCB error, as it appears that the speaker return is via the 10 ohm resistors, not a direct earth.
You will need to either put a link in place of the resistors, or do some minor surgery on the PCB. Without the earth lift resistors there is the possibility of a small amount of hum.
This will apply to all those who have used Will's PCBs.

Kind Regards
Alex

[Deleted]

Thanks Alex,
but I'm tired tonight so anything to do with boards will have to wait.

Reckon it'll be surgery but I'm a bit baffled when I look at a board that's not plain vanilla and single sided.
That probably means I'll need detailed instructions but when I feel like it I'll try to decipher where all the traces go
and get the scalpel out.
Mark you it might be easier just to cut a pair of boards plain vanilla style.
Going to bed three hours before my norm.
G'dnight
Alan

[Deleted]

Yup I’d go for surgery and try to keep the ground lift in place but that’s just an opinion.

Looks like the Ground track is the one in red that runs on the top side of the board so should be quite easy to run a scalpel through it.

I’d remove the LS OP connection and get as close to the Ground pad as possible as there seems to be less chance of hitting anything else that end. then after replacing the conection turminal i'd just run a THICK wire link between LS ground and Circuit ground.

But let’s try and find the best way as I’m often wrong.

Take care

[Deleted]

Guys
Please check your emails.
Kind Regards
Alex

[Deleted]

I’d remove the LS OP connection and get as close to the Ground pad as possible as there seems to be less chance of hitting anything else that end. then after replacing the conection turminal i'd just run a THICK wire link between LS ground and Circuit ground.

Hi Shaun
That way would be ideal as it would cover up the cut, BUT if the terminal block doesn't come out easily, stop and redo the soldered joints as there is a through connection between top and bottom traces on the active side of the speaker terminal block that could possibly get damaged.
Regards
Alex

[Deleted]

Hi Alex

This is how I’d do it

it's not a guide just the start of Discussions

I’m sure that with a solder suckering and a little gentle wiggling that the terminal block should be ok to remove. I’ve had a look at my board and the holes are not too small so it should be fine.

I’ve Errrr borrowed a terminal block or two in my time and it just takes a little care and patience. From the bottom of the board I just heat the solder enough to melt (not going toooo crazy to protect the board) and solder sucker it a few times on each pin. Sometimes the block just comes right out but if not then I gentle heat one pin at a time and tease that pin up a little. Then do the same on’t other side. Once the first bit is done the rest is child’s play.

Then after identifying the track I’d just scrape a little of the solder mask off near to the LS ground pad. Do continuity test to make sure we have the correct track. then cut it as close to the LS ground pad as possible with a good quality modelling knife or scalpel with a nice fine blade (no buzz saws or can openers needed ), I’d also be inclined to make a second cut about 2-3mm away from the first further down the track.

Take the piece of track out between the cuts and dab a little black nail varnish (don’t forget to ask permission) on the cut and allow to dry. That’s the ends sealed nicely.

When the varnish has dried if you have a fresh terminal block then that would be better if not then just straighten out the old block pins with some fine nose pliers.

Once the terminal block is soldered in then just link the LS ground pad to the Circuit ground under the board.

That’s it all bodged err I mean done.

Anyone spot a mistake or two pipe up

I do make em

Just a small amount of care and patience needed.

and the ground lift stays in

Take care

[Deleted]

I have now modified both PCBs as per the method that Shaun and myself described. You can't see the cut in the track because it is covered by the reinstalled speaker terminal block.
I ran a length of tinned copper wire underneath the PCB, between the earth terminal of the speaker terminal block and both central earth terminals.

You will need to check with your DMM afterwards, that the earth lift terminal is no longer connected to speaker earth.
Both my speaker terminal blocks took quite a bit of gentle persuasion, and several attempts with a solder sucker to free them, so be careful, and if looks like you may cause damage, then cut the track further past the terminal block and resolder the terminal blocks again.If the 10 ohm earth lift resistors appear to have been overheated, they should be replaced.
Cutting their leads near the body and gently pulling on the cut leads while the solder joint underneath is heated, will make them easy to remove.
Alex

P:.S.
I used the tip of a small rat tail file to make the cut in the PCB track , in case I slipped with a sharp blade and damaged the track immediately above it.

[Deleted]

Apr 23, 2013 3:19:27 GMT @sandyk said:

I have now modified both PCBs as per the method that Shaun and myself described. You can't see the cut in the track because it is covered by the reinstalled speaker terminal block.
I ran a length of tinned copper wire underneath the PCB, between the earth terminal of the speaker terminal block and both central earth terminals.

You will need to check with your DMM afterwards, that the earth lift terminal is no longer connected to speaker earth.
Both my speaker terminal blocks took quite a bit of gentle persuasion, and several attempts with a solder sucker to free them, so be careful, and if looks like you may cause damage, then cut the track further past the terminal block and resolder the terminal blocks again.If the 10 ohm earth lift resistors appear to have been overheated, they should be replaced.
Cutting their leads near the body and gently pulling on the cut leads while the solder joint underneath is heated, will make them easy to remove.
Alex

P:.S.
I used the tip of a small rat tail file to make the cut in the PCB track , in case I slipped with a sharp blade and damaged the track immediately above it.

Good work Alex

Nope not easy peasy but not everything is in this DIY lark.

But it works and the GL is still sitting pretty.

''Both my speaker terminal blocks took quite a bit of gentle persuasion, and several attempts with a solder sucker to free them, so be careful, and if looks like you may cause damage, then cut the track further past the terminal block and resolder the terminal blocks again.If the 10 ohm earth lift resistors appear to have been overheated, they should be replaced.''

All good advice and luckily enough the boards are good quality so should with care survive the surgery.

Onwards and upwards.

Take care

[Deleted]

Hi Shaun,
I've only got PINK will that still work?
About to take the boards out but not much time today to myself.
The 10Rs are burnt but still functional reading at 12.6R which is what the last one in the pack also reads.
However I'll replace any way,
thanks for all the detailed instructions Shaun/Alex it's a PITA pulling these boards as I've got them hardwired
and I was hoping to move on to pastures new.
Regards,
Alan

[Deleted]

I've only got PINK will that still work?

Yes but you may end up with a slightly coloured sound.

Pastures new?

Good luck with the mod

Take care

[Crispy]

Apr 22, 2013 20:35:20 GMT drymdrum said:

Okay got to the no. 2 build with Will's boards.....................
first thing I notice, the 10R ground lift resistors are burnt on both boards, see pic below.

When did I complete this amp, must have had it running solid since first week in January so it's had a good few hours seemingly working okay
into my 8 ohm TLs so why does it burn the ground lift when it is connected to the 6 ohm speakers if that's when they over heated.
I'm going for a beer and just relax for a bit.

 


 

I don't know much about electronics Alan, but sometimes someone can see things differently and come up with a solution.

Would simply changing the resistor for one of a higher wattage rating work?

It could just be that it has to work a lot harder at 6ohm than 8ohm and so getting hotter - only a thought.

Good move though having a beer and relaxing, then come back to it another time and grab it by the balls and sort it out

One for you to try if you have not already.  Abbot Ale - a really nice bottle of beer.

[Deleted]

Would simply changing the resistor for one of a higher wattage rating work?

It could just be that it has to work a lot harder at 6ohm than 8ohm and so getting hotter - only a thought.

Hi Chris
That won't help as the 10 ohm earth lift resistor is at present effectively in series with the speakers, and will reduce power output markedly,
as well as degrade performance.
Kind Regards
Alex

[Deleted]

Hi Chris/Alex,
boards are out but got to go pick up granddaughter at school, hope I get a bit of time later.
Alan

[Crispy]

Apr 23, 2013 9:20:45 GMT @sandyk said:

Would simply changing the resistor for one of a higher wattage rating work?

It could just be that it has to work a lot harder at 6ohm than 8ohm and so getting hotter - only a thought.

Hi Chris
That won't help as the 10 ohm earth lift resistor is at present effectively in series with the speakers, and will reduce power output markedly,
as well as degrade performance.
Kind Regards
Alex

Hi Alex,

I thought it couldn't be that simple , but sometimes simple things get overlooked and I was 99.99% sure that if it was that simple you would have noticed it and said so. You're the man to sort this problem Alex and I am sure that now Alan has had a sleep and a think about it, his son's amp will soon be singing beautiful music again.

Regards

Chris

[sean]

Hurrah, some progress - I bypassed the 10R ground lift resistor (wire link) and took the input to the DC corrector from the other side of the Zobel network, and there's no oscillation on my 4 Ohm speakers now No hum either.

There's about 30mV of DC offset on the output though, so the corrector still isn't quite right. Alex, i'll take some of those measurements as requested and report back on here...

[sean]

Right, here's some measurements - with a 1k resistor across the audio input and a 4R7 dummy load fitted.

DC offset on speaker is 79mV (ouch!) - had a preamp plugged in instead of 1k before when I got 30mV offset)

Voltage across R3 (220R) = 653mV
R4 (2k2) = 6.44V
R5,6 (200R) = 291mV, 296mV
R7,8 (100R) = 145mV, 147mV
R14 (120R) = 639mV
R18 (130R) = 696mV

Base of T7 - 1.337V w.r.t. negative rail

-all values are about what you suggested Alex, any ideas?

edit- just measured pin 6 of the op amp, it's at -13.9V w.r.t. ground!

[Deleted]

Apr 23, 2013 10:39:31 GMT sean said:

Hurrah, some progress - I bypassed the 10R ground lift resistor (wire link) and took the input to the DC corrector from the other side of the Zobel network, and there's no oscillation on my 4 Ohm speakers now No hum either.

There's about 30mV of DC offset on the output though, so the corrector still isn't quite right. Alex, i'll take some of those measurements as requested and report back on here...

Hi Sean
You should be able to drive those speakers much harder too, as well as sounding much better.
It would be interesting to know how well that low a power from a Class A amp can actually drive them.
We should be able to get close to 30W in Class AB into 4 ohms using those beefed up PSUs, and around 7.5W in pure Class A into 4 ohms.

I replied to your email before reading this post. Yes, your DC offset is way too high, as the rest of us normally get < +- 1mV .
You are getting there though!
Kind Regards
Alex

[Deleted]

Apr 23, 2013 11:08:17 GMT sean said:

Right, here's some measurements - with a 1k resistor across the audio input and a 4R7 dummy load fitted.

DC offset on speaker is 79mV (ouch!) - had a preamp plugged in instead of 1k before when I got 30mV offset)

Voltage across R3 (220R) = 653mV
R4 (2k2) = 6.44V
R5,6 (200R) = 291mV, 296mV
R7,8 (100R) = 145mV, 147mV
R14 (120R) = 639mV
R18 (130R) = 696mV

Base of T7 - 1.337V w.r.t. negative rail

-all values are about what you suggested Alex, any ideas?

edit- just measured pin 6 of the op amp, it's at -13.9V w.r.t. ground!

Hi Sean
Looks like opamp problems. Are you sure they are OPA134 ? Perhaps the fault may have destroyed them,
although that's hard to imagine with such a high input impedance, and an 86K load to earth.
With the opamps removed, and the load either removed or much higher, how is the offset ?
Regards
Alex

[Deleted]

Apr 23, 2013 11:25:51 GMT @sandyk said:

Apr 23, 2013 11:08:17 GMT sean said:

Right, here's some measurements - with a 1k resistor across the audio input and a 4R7 dummy load fitted.

DC offset on speaker is 79mV (ouch!) - had a preamp plugged in instead of 1k before when I got 30mV offset)

Voltage across R3 (220R) = 653mV
R4 (2k2) = 6.44V
R5,6 (200R) = 291mV, 296mV
R7,8 (100R) = 145mV, 147mV
R14 (120R) = 639mV
R18 (130R) = 696mV

Base of T7 - 1.337V w.r.t. negative rail

-all values are about what you suggested Alex, any ideas?

edit- just measured pin 6 of the op amp, it's at -13.9V w.r.t. ground!

Hi Sean
Looks like opamp problems. Are you sure they are OPA134 ? Perhaps the fault may have destroyed them,
although that's hard to imagine with such a high input impedance, and an 86K load to earth.
With the opamps removed, and the load either removed or much higher, how is the offset with the same preamp, which I presume has miniscule DC out.?
You should also check that you still have + and - 15V into the opamps sockets.
Regards
Alex