'AK Mods' SC Class A 15W Amplifier

[Deleted]

BTW, the small PSU PCB size means it will just clear the chunk of aluminium I cut for mounting transformers

Hi Jeff
Not a lucky guess. That's why we decided to go with just 2 larger value electros. Giving provision for a another couple of parallel ones would have made the PCB length too long for some.
Kind Regards
Alex

[Will]

Hi,

Just doing a list of 'things to-do' for the amp pcb, and have the following list.

• Change the 78L15 and 79L15 regs on the offset corrector for to220 7815/7915
  
• Change the post 7815/7915 reg cap to 10uF
  
• Put wire pads in to allow the input/output to the offset corrector to be disconnected
  
• Make provision for the feedback capacitor
  

Any other ideas are welcome, as long as they don't mean huge amounts of work.

One thing I'd like to ask of Alex - as at least one person (looking at you, Chris ) will most likely want to run at a higher bias current, would it be a good idea to allow for larger resistors than the current 0.5W ones?

[Deleted]

Hi Will,

I think, after the recent discussion of higher bias pushing the amps into AB earlier in the wattage scale, that I would benefit more from the standard arrangement.
Especially as I may be taking my SBLs down the active route, meaning 4 channels firing them up!
If I plump for this (subject to sorting an active cross-over) I'll adjust my order accordingly, i.e. four monoblocks so I can experiment with both horizontal and vertical configurations.

It's either this or flog the SBLs and get something else, hmmmmmmm.

Edit = talking bow-locks again!

[Deleted]

I think, after the recent discussion of higher bias pushing the amps into AB earlier in the wattage scale,

Hi Chris
Higher bias will do the opposite. Perhaps you are thinking about lower impedance speakers such as 4 ohms ?
Regards
Alex

[Deleted]

One thing I'd like to ask of Alex - as at least one person (looking at you, Chris ) will most likely want to run at a higher bias current, would it be a good idea to allow for larger resistors than the current 0.5W ones?

Hi Will
If you use Ohms law you will see that even at 150mA bias and .25 ohms (4 x 1 ohm in parallel) that you are still well within the total power rating of 2W.
Kind Regards
Alex

[Deleted]

Apr 29, 2012 21:44:35 GMT @sandyk said:

I think, after the recent discussion of higher bias pushing the amps into AB earlier in the wattage scale,

Hi Chris
Higher bias will do the opposite. Perhaps you are thinking about lower impedance speakers such as 4 ohms ?
Regards
Alex

Er, yeah, I remembered it all back to front

Had another read through, back on course....

[Deleted]

Make provision for the feedback capacitor

Hi Will
Perhaps make those pads beefier than normal to allow for possible damage due to replacement of the electro by a link later on.
Kind Regards
Alex

[Deleted]

Put wire pads in to allow the input/output to the offset corrector to be disconnected

Hi Will
I am open to suggestions on this one.
Do we need this when we can simply not plug in the I.C. ?
It may not hurt to fit a bleed resistor across both C1 and C2 to give a small load on the VRegs, and ensure that C1 and C2 are discharged before the I.C. is installed.(e.g. 3.3K?)
Alex

Attachments:

[Will]

Hi Alex

Thanks for the feedback. Wasn't sure of the bias current, so good to know that all;s good in that part. Fitting the resistors is easy enough, as the pcb will have extra room now that I have to make it big enough to to take the zobel and offset corrector. It'll essentially mean that the little offset pcb will be part of the same pcb as the amp, tagged on the side.

[Will]

Progress report.





No where near finished, but an indicator of where I am with things. Mr Clancy has been a star again, and created the custom files required for the output devices and output inductor. Cheers, Jon!

As you can see, provision has been made for the feedback capacitor (the two wirepads above the groundlift). The output zobel has been implemented, which has created a lot of space on the RH side for the offset corrector. This is the next area I want to sort out. Tidying up and compaction to come after the layout is tidied up.

We're getting there!

If someone could compare the cct diagram with Alex's latest cct, posted on page 18(?), for component type, value and general cock-ups, I'd appreciate it

[Deleted]

Hi Will

Looking very good indeed well done

had a quick check of the above circuit and it looks spot on to me but another set of eyes on it would be good just to make sure the Gin has not effected my eyesight to much

take care

[Will]

Thanks Shaun!

I've a suggestion to put out to you guys.

It was suggested that being able to isolate the offset corrector (OC) from the rest of the circuit is useful if you need to fault find.

Something that I've implemented on the pcb this evening is wirepads on the input and output of the offset corrector. These wirepad are joined by a fine trace, that will enable the circuit to work straight off, but could be cut with a sharp knife if you want to isolate the circuit, of you prefer to use a capacitor in the feedback to reduce dc level. The wirepads can be linked (obviously using 99.999999999999999999% oxygen free copper, gold plated and blessed by the pope) to reinstate the offset corrector when required.

Although the op-amp could be removed to remove the corrector from circuit, using the pads will allow the offset corrector to be tested itself, in isolation.

Is this useful to you guys, or overdoing it a bit?

The suggestions in the list I posted have been implemented, and going from Shaun's comment on hearing an improvement when making changes in the offset corrector, I've powered it from the front end supply, not the power stage, which believe me, would have been much easier (you all owe me a beer, by the way

I'm at a stage now of tidying up, and checking for crossed/touching traces. It's a bit tight around the OC opamp, so might need to re-jig.

Anyhow, here's where we are now...

[Deleted]

Is this useful to you guys, or overdoing it a bit?

Hi Will
Did you see my previous reply about fitting load resistors across the decoupling for the Correctors Voltage regulators, and just removing the I.C. ? (some VRs like to see a minimum load of several mA)
There is nothing left that is able to upset the basic operation with the I.C. removed.The resistances are too high.
Regards
Alex

[Will]

Hi Alex,

Yep, the resistors are in now, providing a small load.

The offset isolation was mentioned in an email, by someone who'd built the amp (that narrows it down a bit!) that being able to disconnect the offset corrector from the circuit would have allowed for easier fault finding when they had a problem. This to test the operation of the corrector itself, away from the main circuit.

We should not really have any problems now, given that that the regs are the to220 ones as suggested by Frans.

[Deleted]

Although the op-amp could be removed to remove the corrector from circuit, using the pads will allow the offset corrector to be tested itself, in isolation.

Hi Will
You would still need to leave the input to the Offset Corrector intact, or the output is likely to go to very close to the supply rail voltage.
Regards
Alex

[Will]

Morning,

Based on the numbers indicated in Shaun's list of interested parties, and discounting Chong's 10 PSU boards (just to make life easy, the coffee hasn't kicked in yet) A set of two amp pcbs and two supply pcb's is looking like c£17 a set. This for thick copper, 1.6mm pcb, nice colour, standard manufacturing time and delivery. If we want expedited manufacturing and delivery, the cost goes up to c£20 a set.

Amp pcb is 135mm x 90mm.

[Deleted]

Sounds like good value Will. I'll go with whichever delivery speed suits everyone else, it will probably take me a few years to get it built, as usual, anyway.

Syd

[Deleted]

Hi Will
Sounds like great value.I just hope that the rest of you can duplicate the results that Shaun and myself are getting.
Kind Regards
Alex

[jeffc]

Hi Will,

Well done, PCB looks very nice and good value for so few.

cheers.. jeffc

[Deleted]

Hi WillYou would still need to leave the input to the Offset Corrector intact, or the output is likely to go to very close to the supply rail voltage.RegardsAlex

Hi Alex

i may have this wrong but with the Feedback cap in place the offset corrector IC could be removed to allow for fault finding.

that's what I'd suggested to Will because it helped to isolate my front end problem with the offset corrector removed and the feedback cap in place.

or whats the point in allowing for the feedback cap?

but as you say it may not be necessary.

but

going on the Pre/HA build it may have helped with some of the initial problems to take the OS corrector out of the mix and still be able to test the rest of the circuit.

in my post i listed a few things that i felt may have been useful in the hope of being corrected. not really as the definitive list of things that must be done.

Your amp is still sounding just great with the FB1+

Hi Will

that board looks fantastic and the price well it's less than it cost me in materials to etch my own.

great work

now all you need to do is build up a 20 or so ;D

take care

[Deleted]

Hi Shaun
The 100/220uF FB cap is a must for initial testing.
I don't like the idea of cutting tracks, and believe that simply not fitting the I.C. is all that is necessary at this stage, and removing it again if there are problems when the Offset Corrector is installed. The components at the input and output to the Offset Corrector won't hurt if the I.C. is removed, unless you have made a major error in that area, or a solder bridge or something.
If you are trying to test the Offset Corrector in isolation, you would need to connect the input to 0 volts and check for the DC out.
If the amp has reached the stage of testing where it has low DC out from before the I.C. is plugged in, then you can use that into the Corrector's Input instead of trying to connect 0 volts to it.
If there is low DC out before plugging in the I.C. and much higher afterwards, then you have a fault in the Offset Corrector area, such as a crook VR etc.
I hope that I haven't thoroughly confused you by now, as I am starting to feel a little that way myself!
Kind Regards
Alex

[Deleted]

May 5, 2012 0:30:58 GMT @sandyk said:

Hi Shaun
The 100/220uF FB cap is a must for initial testing.
I don't like the idea of cutting tracks, and believe that simply not fitting the I.C. is all that is necessary at this stage, and removing it again if there are problems when the Offset Corrector is installed. The components at the input and output to the Offset Corrector won't hurt if the I.C. is removed, unless you have made a major error in that area, or a solder bridge or something.
If you are trying to test the Offset Corrector in isolation, you would need to connect the input to 0 volts and check for the DC out.
If the amp has reached the stage of testing where it has low DC out from before the I.C. is plugged in, then you can use that into the Corrector's Input instead of trying to connect 0 volts to it.
If there is low DC out before plugging in the I.C. and much higher afterwards, then you have a fault in the Offset Corrector area, such as a crook VR etc.
I hope that I haven't thoroughly confused you by now, as I am starting to feel a little that way myself!
Kind Regards
Alex

Hi Alex

no confusion and a really good explanation of how the testing procedure works out in practice.

seems to make perfect sense to me and maybe using an IC socket would be the way to go. i know that some don't like them too much but I've used them on all of my builds without a problem.

I'd much rather take a chance on the socket than try to de-solder that 8 legged beast if it should go pop.

it would be quite easy to remove for fault finding purposes and the offset regs could be tested @ the IC input pins before committing the expensive IC. it also gives an opportunity to experiment with different chips.

for my part having the option of fitting the feedback cap was just a real bonus when it came to testing the main board circuit. it also allowed me to compare the SQ with AC and DC feedback which was interesting and showed a noticeable difference between the two (as i would expect it was an eye opener).

I'm pretty confident that people are going to get consistently good results with this amp SQ wise and build wise also.

the good work that Will has put into the board layout will help that along nicely.

I'm still soooo happy with mine.

take care

[Deleted]

Hi Shaun.
I would definitely recommend fitting a high quality gold flashed, turned pin DIP 8 I.C. socket. Unlike the HA PCB there will be no problems there, as the PCB isn't as thick this time around.It also gives the opportunity to see which of a batch, or type of I.C. permits the lowest DC offset at the output.
Kind Regards
Alex

[Deleted]

May 5, 2012 9:48:20 GMT @sandyk said:

Hi Shaun.
I would definitely recommend fitting a high quality gold flashed, turned pin DIP 8 I.C. socket. Unlike the HA PCB there will be no problems there, as the PCB isn't as thick this time around.It also gives the opportunity to see which of a batch, or type of I.C. permits the lowest DC offset at the output.
Kind Regards
Alex

Hi Alex

yes i totally agree with the use of nice quality sockets which will leave plenty of room for experimentation.

i have them fitted in your PRE/HA and Power and not had any problems but it did allow me to try a few Op Amps which was useful.

take care

[Will]

May 4, 2012 10:08:23 GMT @sandyk said:

Hi Will
Sounds like great value.I just hope that the rest of you can duplicate the results that Shaun and myself are getting.
Kind Regards
Alex

Can't see why we wouldn't

With the offset corrector, perhaps to much is being made on the past problems encountered with the thick pcb's? As mentioned, this should be mitigated with the use of normal thickness pcb's. Offset isolators removed. By the way, the track cutting was not a big thing, as the trace was less than a mm thick, and only needed if you wanted to completely remove the corrector from circuit.