Reviving the AK Class A HA/Preamp Build

[jeffc]

Guys,

I've been populating the PCBs too and am almost done, just the power transistors/heat sinks and a few terminal blocks to go. Using the nice AK Black Label JLH PCBs gifted to me by Alex plus a few other hard-to-find bits. Measured and matched BC337/BC327 and 2SA5171/2SA1930 hfe as close as I could get them this arvo and have just fitted the BC337/BC327 now. A few pics of progress with the PSU transformers the PCBs below but still plenty to do with fun wiring and casing stuff.

cheers.. jeffc

18-0-18 50VA Transformers input terminal block

18-0-18 50VA Transformers output terminal blocks

18-0-18 50VA Transformers output PCB type fuses

18-0-18 50VA Transformers plastic box - tight fit

Will W PSU - top

Will W PSU - bottom - ESID diodes

AK Class A PCB - Top 1

AK Class A PCB - Top 2

AK Class A PCB - Bottom - ES1C diode fitted under

AK Class A PCBs as they'll be positioned in the case

[Deleted]

Nice work Jeff !
I am having trouble getting a decent photo of my completed Preamp PCBs. I may have to try outdoors when sunny.

Kind regards
Alex

[jc]

Yes X2, very neat. I'm just about to pop in my order for the parts I don't have. I do, however, have a choice of two already built power supplies to choose from, one from the original cobbled Jaycar boards and another Avondale unit I picked up as an non made project and completed it.

BUT

I'm still toying with the idea of making an integrated amp as I have a pair of power amp boards tucked away. That would mean a total rethink on the PSU! Currently I'm on the lines of two TXs with low power and high power coils and then the four PSU boards etc. Of course I'd need a taller case to accommadate all the gubbins too! A Class integrated, hmmmmmmmm, I like the sound of that...


Jeffc you have email on your gmail account

[jeffc]

Thanks Alex/Chris,

Hopefully I've used parts well enough within spec for the amp to operate up to spec, but have tried to check these carefully with my new little MEGA tester before fitting so fingers crossed. As for parts selection for audio finesse, bit of a lucky dip here but again fingers crossed.

Very ambitious with an integrated Class A amp build Chris, it's tricky enough building one of these, but go for it. I hate interconnect wires hanging everywhere so integrated appeals to me... as you'll as this build progresses...
Yes just checked my gmail... good to see 'more' JLHs are on there way.. Although I've got a few already, they'll get used sometime for somewhere useful for sure...

cheers.. jeffc

[Deleted]

Just out of curiosity, this morning I checked DC Out etc. in my existing HA which is virtually identical to what you guys are constructing here.

 One channel had 0.0mV DC Out, and the other was only 0.3mV out with the attenuator in the typical listening level position.

This is with OPA134 used in the Offset Corrector section.

I am still waiting on some BZX55 C3V3 Zener Diodes from the U.K. to complete the new PCBs. These Zeners have typically 3.3V at only 5mA current, where most other 3.3V types need closer to 20mA.

Alex

[jc]

In the first round of building we experimented with using Charcroft naked foil resistors in some of the more critical audio signal path positions. IMHO they made a small but notable improvement. The prices of these, however, have gone through the roof since then at £8 to £16 a pop, roughly £120 per stereo amp for 2 off of R positions 1,2,11,12,31 . I think we paid nearer to £50 back then in our moment of madness!

A cheaper alternative, if you wanted to go down this path, would now be the Caddock range at approx. £3 each, roughly £30 in total which are claimed to be close in performance. There are of course the standard ones and cheaper "better" ones available as well.

A source

[Deleted]

Sept 23, 2015 13:23:41 GMT jc said:

In the first round of building we experimented with using Charcroft naked foil resistors in some of the more critical audio signal path positions. IMHO they made a small but notable improvement. The prices of these, however, have gone through the roof since then at £8 to £16 a pop, roughly £120 per stereo amp for 2 off of R positions 1,2,11,12,31 . I think we paid nearer to £50 back then in our moment of madness!

A cheaper alternative, if you wanted to go down this path, would now be the Caddock range at approx. £3 each, roughly £30 in total which are claimed to be close in performance. There are of course the standard ones and cheaper "better" ones available as well.

A source

Hi Chris

 I only bothered using them with the 4.7K and 10K gain setting resistors in my Preamp. I didn't bother with the 56K input resistor which is shunted by the typically much lower value of a typical attenuator setting. I don't feel there is much value in using them with much lower value resistors which have a lower Johnson noise. I  didn't bother using them in my HA as I felt the very minor gain wasn't worth the inflated cost of the resistors.At a later date I obtained used naked foil 4.7K and 10K resistors from ebay for the HA.

I used pairs of the physically larger 0.1% resistors for the emitters of the differential pair and the current mirror that I had left over from the original group buy. If you have a good DMM, it's also easy enough to get a few extra 1% types and match the pairs.

Kind Regards

Alex

[Deleted]

Now for the reassuring part.

I have just tested both of my completed PCBs using +20V  and -20V from my bench PSU.

With both PCBs having a 3.3Kohm resistor across the Input socket, the DC out for both was a little over 12mV.

With the OPA134 Offset Corrector I.C.s plugged in, both then had 0.3mV or less, DC out.

 Before adjusting the Bias Current, I set the Bias trimpots to around half way using my DMM (250 ohms)

This ensures that you can adjust the bias smoothly to the intended Bias of 100mA.  This is measured by reading the voltage across one of the groups of 4 x parallel 10 ohm resistors . A bias of 100 mA will result in a voltage drop of 250mV across these resistors .

This voltage will increase a little as the PCB warms up.

 With the front end balance adjustment, I didn't pre-set their trimpots, and I initially read a difference of around 130mV with both PCBs.

This was easily adjusted down to 0.0mV, and when I further adjusted the trimpot, I was able to achieve a reversed polarity difference of >70mV.(I didn't try to see how far I could go !)

This means that unless the diodes you use at that location are well out of specification, you will not have any problems achieving full front end balance. Try setting the Balance trimpots to around 130 ohms initially.

I will try replacing the existing PCBs in my Preamp with the new PCBs another day, for an extended listening comparison test.

Alex

[pingu]

Most informative Alex, thanks for the info. I will need some guidance thro this build I am sure and found this really helpful, thanks
So all this activity has spurred me on. I am doing the BOMs for the HA and the PSU today .
Hopefully start next month when the bits arrive and I have a few days off.

Martin

[jc]

Finally made a start! (poor photo though)

[Deleted]

Chris
If you have a couple of 47pF polystyrene laying around, try replacing the 100pF at the input with them.
Presently, I have my 2 new Class A PCBs replacing the existing in my preamp.
They sound almost identical, but I feel the original has a very slight edge.
I don’t think it is due to the offset corrector having double the C and half the R in it’s input which would lower the self generated noise of the OPA134.
I did notice something with the original that I had forgotten. I had replaced the input 100pF polyprops with 47pF polystyrene.
I know it shouldn’t matter, but I must have heard a small improvement originally. They have thinner leads, so shouldn't be too hard to remove again if you decide to.

Kind Regards
Alex

[jc]

Hi Alex,

Right you are, I think I ordered some last time around intending to change them, I'll have to search the vault  

[jc]

I stuffed a few more components in today, I think that's my lot until the new supplies come in.

[Deleted]

Sept 27, 2015 1:14:46 GMT jc said:

Hi Alex,

Right you are, I think I ordered some last time around intending to change them, I'll have to search the vault  

Hi Chris

If you have a few extra 68pF for the compensation capacitors, you could try those instead of the 100pF.

Regards

Alex

[jeffc]

Hi Alex,
Good to know that the front end balancing and DC offset are all working fine on your new boards. My amp, JLH and PSU PCBs are all now done, so time to power up the transformers when I get a sec and start testing each. On the PSU, I seem to remember that placing a small value film capacitor across the AC in as calculated based on transformer VA rating and type wasn't helpful, so best not to bother. This correct?
Chris, I also had a 10K foil resistor not used in my AK Class A HA build so I used it here simply because I had it. As for other resistors, I was a bit miffed that quite a few that I used were TE LR1 0.6 W types that the 0.25W versions of didn't impress too much in this review.
www.sasaudiolabs.com/resistor.htm
For this reason I've been looking at potential alternatives to test sometime in some of those more important spots, and have actually been looking at 1W and 2W SMD wire wound types that are relatively cheap. Might get to typing some after the amp has been up and running for a while. Looks like your using 1W types based on their size in your pics, any particular brand you've found to perform well, or just larger physical size = lower noise?

cheers.. jeffc

[Deleted]

Sept 27, 2015 11:20:31 GMT jeffc said:

Hi Alex,
Good to know that the front end balancing and DC offset are all working fine on your new boards. My amp, JLH and PSU PCBs are all now done, so time to power up the transformers when I get a sec and start testing each. On the PSU, I seem to remember that placing a small value film capacitor across the AC in as calculated based on transformer VA rating and type wasn't helpful, so best not to bother. This correct?
Chris, I also had a 10K foil resistor not used in my AK Class A HA build so I used it here simply because I had it. As for other resistors, I was a bit miffed that quite a few that I used were TE LR1 0.6 W types that the 0.25W versions of didn't impress too much in this review.
www.sasaudiolabs.com/resistor.htm
For this reason I've been looking at potential alternatives to test sometime in some of those more important spots, and have actually been looking at 1W and 2W SMD wire wound types that are relatively cheap. Might get to typing some after the amp has been up and running for a while. Looks like your using 1W types based on their size in your pics, any particular brand you've found to perform well, or just larger physical size = lower noise?

cheers.. jeffc

Hi Jeff

 The small value cap didn't seem to accomplish much in the Preamp/HA and caused a blow out in the soundstage with the X-DAC V3 .

John Swenson's typical snubber is 22nF in series with 330 ohms across the outers of the secondary windings, but again didn't seem to make much difference, possibly due to the JLHs.  

Just remember that due to the very small copper lands on this PCB that it's not a good idea to try and make changes afterwards!

The 1W and 2W resistors are likely to have much thicker leads, and may not be too easy to get out again without damaging or destroying the tiny copper lands, even though this PCB is thinner than the original.

Kind Regards

Alex 

[jc]

Alex, I think I have extra 68pf as well which may be a better option (in my illogical mind at least!) than halving the value.

Jeff, the larger resistors are 0.6w 5% tightly screened to under 1% which I picked up from an American seller very reasonably (if I remember who I'll divulge) and I also had several values in 0.4w 0.1% (smaller physically). As Alex said the leads on the larger resistors could cause problems, the 0.6w I've used only just fit the through-holes and I certainly wouldn't want to entertain removing them!

[jc]

BTW Jeff, I liked your solution for connecting the ES1 diode under the board. For my part I used a tidied up version of my previous effort on the first boards.

[pingu]

Help please,

I seem to have lost several pictures from here for some reason. Not all but just the ones I need.
Alex's diagrams from page one of this thread have disappeared, amongst others .

So has any one got or can they point me towards the diagrams for HA / Pre and Gregs new PSU or a BOM for them

A wee bit frustrating as I was going to order the bits a day or so ago

Thanks
Martin

[jc]

I hve the preamp stuff Martin but I didin't download Gregs PSU details and they're not on his site either. Maybe Alex is the only one that can provide those. Preamp stuff on the way....

Alex, have you been doing some photo cleanout on on your provider?

[jc]

Sept 29, 2015 23:26:22 GMT jc said:

I hve the preamp stuff Martin but I didin't download Gregs PSU details and they're not on his site either. Maybe Alex is the only one that can provide those. Preamp stuff on the way....

Alex, have you been doing some photo cleanout on on your provider?

I lied!!!

I've sent all I have to your email.

Happy building

[Deleted]

Sept 29, 2015 23:26:22 GMT jc said:

I hve the preamp stuff Martin but I didin't download Gregs PSU details and they're not on his site either. Maybe Alex is the only one that can provide those. Preamp stuff on the way....

Alex, have you been doing some photo cleanout on on your provider?

Hi Chris

Yes, I have been doing some photo cleanout, but not any of those, just mainly old cartoons etc.

If anyone needs further info just ask in the thread, and if I have it I will post it in the thread as well as email it directly .

Kind Regards

Alex

[jc]

 I liked Jeff's method for mounting the SMD diodes enough to steal it! Mine stayed top-side.

jeffc



cj

[pingu]

Hi Chris,
Thanks for the mail I now have all the details.
Now I only need all my new p/w's that I changed after my mail was hacked. I know I put them away very, very safely ;(

Thanks
Martin

[jc]

Some progress but no cigar yet...now awaiting parts.