(AK) classA HA/Preamplifier.

[Will]

Hi Alex,

Yep that was on the tick list. Before the serious listen, I'll be resetting/checking the voltages, bias and the front end.

[Will]

Hi Phil,

Hopefully this picture will help



The front panel was counter-bored (at a local machine shop, £10-ish) to allow the knob to sit 'in' the front panel. I then drilled two holes through the back of front panel, behind the knob, in line with the headphone socket and power led. By changing the resistor feeding the leds, which sit in the drilled holes, you can change the brightness. More holes, more leds, brighter!

I'll let her out later, in time to make tea! ;o)

[Spirit]

You bad man you

That, is exactly what I'll do on my next build. [This will of course require a front of case that's deeper than 1.5mm and not plastic ].

Cheers for the clarification, and helpful diagram

[Deleted]

Will,
that's looking really neat! I was tempted at first but really couldn't put you guys through it

Maybe the DAC if it happens....

[Will]

Guys,

Thanks for the compliments. I have a set of LDR's on the way from the Groupbuy at DIYA, and will be trying them out in this amp.

I do have one set of the LS devices left, which make the build of this amp very easy, so if you want them, or even if you want help in aquiring one of these amps for your own, please PM me. As I've said before, I can organise a group buy of the LS devices again, and help with the PCB's if needed.

Turn the Want into a Have

[insomniac]

Great job Will... very tidy!
I like your backlit volume approach too.. cool idea.

[Will]

Thanks Dave!

Now, especially for Phil,



During daylight, you can hardly see it, but in the evening, when we normally have just table lamps on, it's a nice subtle glow behind the knob.

I know it's just for show, but I wanted to make something that looked good in the living room, and not too clunky.

[Spirit]

Love it.

It *does* look good

Cheers mate.

[Will]

Well, I've set the voltages to +/-15V at the amp board, set the bias to 100mA, LTP test points to 0.02mV, cleaned my teeth and changed my socks, all ready for a serious listen.

I don't think that you'll be surprised to find that I'm very happy with the way the amp deliver music. Everything from my earlier post stands, but is now qualified by having all the components well and truly burnt in.

Out of curiosity, I de-tuned the LTP mod, so that they measured 50-60mV, as they were when first powered up when built. The comparison between this state and the the tuned state is sort of like not having you headphones on properly, It is music and it's good , but it's feels slightly 'out of focus', when compared to the tuned in/headphones on properly state, if those are the words. This LTP mod is very worthwhile.

Well recorded music can almost knock you off your feet. A Fleetwood Mac track that Alex recently uploaded for a test really took me aback. I've never heard the track before, but it's quite beautiful in it's simplicity, the whole timing of the piece keeping you waiting for each note to fall. The same can be said for the analytical powers of the amp. Comparing two versions of Love over Gold (both ripped from the same PC, but the later rip being after more extensive dampening had been performed) was straightforward, with the newer version sounding more fluid and effortless to listen too. As a point, the two versions of LOG had the same checksum.

Very pleased, but I still haven't screwed the top on ;D

[Deleted]

Do any of you guys with the SC Class A HA, either own,or have access to a pair of
ATH-W1000 ? If anybody does , I woud be interested in hearing how you find their performance after replacing the 68 ohm output resistors with 120 ohm 1% MF types.
Alex

[Will]

If anybody wishes to donate a pair, I'm happy to try the above!

[Deleted]

After a couple of days listening to the ATH W1000 with the SC HA with JLH (Jaycar HA) after replacing the 68 ohm output resistors with 120 ohms MF resistors, I found the W1000 performed much better with 120 ohm resistors as per the IEC specifications. However, the majority of the other popular headphones, including the K701, should still sound better with the original 68 ohm output resistors.
As a result of this, this morning I removed the 68 ohm output resistors in the Class A HA , and fitted PCB pins ,after enlarging the holes with a .9mm drill in a hand held drill chuck , which is commonly used to allow larger drills to accept smaller drill bits.
I then fitted 120 ohm MF resistors to the PCB pins. Initial listening tests reveal a large improvement in the performance of the W1000 right across the board. I would recommend fitting these PCB pins to enable different types of headphones to be used on a regular basis, just by replacing the series resistor with a more suitable value.This can then be very easily done without removing the PCBs. As with the SC HA with JLH, 68 ohms is still the recommended series resistor value for most of the other commonly used headphones.
People really do need to know the manufacturer'srecommended driving impedance for their preferred/ favourite headphones.
Alex

P.S.
Perhaps someone who uses the K701 with their Jaycar HA can verify that the existing 68 ohm output resistors are a good match ?

[Deleted]

Perhaps someone who uses the K701 with their Jaycar HA can verify that the existing 68 ohm output resistors are a good match ?

Definitely! If you remember I originally built mine with an ohm selector for comparing this on the 701s and 600s.

The 68ohm setting offered the best sound balance and clarity on the 701s. From memory, lower introduced a slight vagueness, while higher made for brashness.

[Deleted]

Sept 17, 2009 14:47:10 GMT @cjarchez said:

Perhaps someone who uses the K701 with their Jaycar HA can verify that the existing 68 ohm output resistors are a good match ?

Definitely! If you remember I originally built mine with an ohm selector for comparing this on the 701s and 600s.

The 68ohm setting offered the best sound balance and clarity on the 701s. From memory, lower introduced a slight vagueness, while higher made for brashness.

Chris
Thanks. I did remember your efforts with the selector switch, but not which headphones you used.
Alex

[Deleted]

MODULE SET UP TESTING PROCEDURE.
I was reminded today that I hadn't given any initial module check procedures.
1. Fit a 100uF capacitor in the feedback circuit where we normally have a link fitted.
2. Solder a 5.1Kohm MF resistor (or 4.7K) across the input terminals.
If the module is functioning correctly, and has a well balanced input pair, the Output DC voltage should be within a couple of mV of Zero.
If all O.K. the FB capacitor may be replaced by a link, and the offset corrector fitted.
If no Offset Corrector is fitted, leave the FB capacitor in situ. At normal listening levels, the DC offset should be reasonably low, and will normally show much less deviation with volume control setting when a DC coupled source such as a CD player is connected. The LF411 is a good alternative to the AD744 in the offset corrector, although TL071 may be used, but with not as good a correction due to variation between samples.
Alex

[Deleted]

A long time between drinks !
First off, I will give credit to Robert for suggesting that 18-0-18VAC toroidals may do even better. I did state some time back that I would eventually try that, and today I did. I had a few little headaches to overcome though, as my HA is inside a 1unit rack case. First off, the higher rectified DC voltage was a little over 25.4 Volts. This meant that I had to change to the much physically bigger 2,200uF 35 V electros on the Jaycar PSU board for long term reliability. Because of their size, a couple of component mounting holes were obscured,and I had to mount a few diodes under the PCB, but flat against the PCB. Even then ,the damn electros were so tall, that I had to make do with single M3 nuts and spacers for mounting the PCB, to give me clearance from the top cover. I used standard resistor values, so I had to settle for +-20V before the CL JLH, which has a fairly small voltage drop.Next, I had to make minor surgery on Will's Offset Corrector PCB board design, as I needed to make sure that the voltage ratings of the offset corrector I.C.s were not exceeded. To do this, I removed the series dropping resistors, and drilled a hole in between them, into the earth area. I then mounted 78L15 and 79L15 voltage regulators in the 3 holes. (remember that the 79L15 has a different pin out !)
I then fitted a 100nF capacitor between each voltage input and earth to ensure stability of the regulators.
Even though the PSU electros may not be optimum yet, the improvement in dimensionality and low end dynamics has to be heard to be believed.

Leo and Will, do not stuff around, fit 18-0-18 30VA toroidals to yours and be amazed ! "California Project-Papa Doo Run Run" sounds so much better again, when using the ATH W1000 and 120 ohm series output resistors .Who said the W1000 doesn't have a bottom end ? ;D
Miguel
It is no longer just a little better than the SC HA versions we already have with the JLH and LM4562HA, it blows them right out of the water !!!
Alex

P.S.
The front end balance will need to be readjusted, but the bias didn't change much from 100mA.

[Will]

Good stuff, Alex, I'm up for trying this out. All nice and straight forward to do as well. I gather the bits together, and when normal service resumes (Ha!) I'll give it a go.

[robertkd]

[Deleted]

Setting the dual regulator PSU voltages.

Using 120 ohms and 1.8Kohm 1% MF resistors for the voltage setting resistors will give very close to +-20V into the Current limited JLH, which will drop a few hundred millivolts. If 130 ohms and 2.0K ohm 1% MF are used instead, there will be approx. 20.45V into the JLH, and a little over
+-20V out.
I intend changing to 130 ohm and 2.0Kohm. This will still give a more than adequate IN/OUT voltage difference of approximately 5v. (The regulation dropout voltage of the LM317/LM337 is around 2V.)
Alex

[Will]

I've now implemented the 20V supply on my Class A, and have also found an improvement.
You certainly get more 'slam' now!

I also checked out how the amp compares when running with one traffo/reg board/JLH vs. having individual traffos/reg boards/JLHs per channel.
To me, there is a subtle improvement in the way the music is presented, slightly more at ease when going dual over single. Whether it is worth matching components over two PSU and JLH's is debatable, but I'm happy I did it.

[Deleted]

Dec 19, 2009 19:37:23 GMT Will said:

I've now implemented the 20V supply on my Class A, and have also found an improvement.
You certainly get more 'slam' now!

I also checked out how the amp compares when running with one traffo/reg board/JLH vs. having individual traffos/reg boards/JLHs per channel.
To me, there is a subtle improvement in the way the music is presented, slightly more at ease when going dual over single. Whether it is worth matching components over two PSU and JLH's is debatable, but I'm happy I did it.

Hi Will
Results are pretty much as I expected. I am not sure that it worth the trouble of an additional 18-0-18 30VA toroidal, although I could fit another one in the big rack case supply for my Class A preamp and 15W/Ch Class A power amp.
After hearing the differences with the upgrade to the Class A HA, I have almost finished doing the same to my main Class A preamp. Still got the Offset Correctors to fit the little 78L15 and 79L15 voltage regs in, to protect their I.C. from too too high supply rails.
Alex
P.S.
Are you using the CL in the JLHs ? I am using 1 ohm emitter resistors, and 1.8K base resistors for the Current Limiters. I am also getting more 3D sound with good ripped material, when playing from the 32GB Corsair Voyager GT USB 2.0 pen.

[Deleted]

Will
I had quite a fright when powering up the Class A preamp again after increasing the supply rails at the regulators to a little over +-20V into the JLH, and at the last moment remembering that I had to also drop extra voltage before the coils of the Delay Circuit. I had almost full supply voltage present at the outputs of the preamp !
After tearing out most of my remaining hair, I found that the JLH had the correct -Ve rail supply of -20V going in, but only -.8V out !
Eventually, I found that I must have pushed down too hard on the top of the Current Limiter device (2SC5171) when reinstalling,and caused a track to fracture.
I had used my original larger JLH PCB , because it has multiple output pins, which makes for tidier wiring.
Anyway, so far I have only tested with DTV audio, but the extra depth of image is very obvious, It's more 3D sounding too.
Alex

[Will]

Alex,
With two JLH's etc, I would still do it again, if going for a 'best of the best' build, as there is an improvement to be had. If you get to the stage where you have a spare JLH/psu and 18V traffo (I seem to have plenty of 15V traffos for some reason....), I'd certainly recommend it. As for the CL, yes they are in place and using the same value as yourself.

I've had more of a listen this morning (left on overnight) and presentation is definitely more vivid vs.15V. Having the luxury of a variable PSU, I've reduced voltage a couple of times, re-balanced, listened and then gone back up to 20V. Yes, much better!
Listening to the Fleetwood Mac version of 'Need Your Love So Bad' and Peter Green is in the house! Fantastic!

[Will]

One thing to mention is that before I replaced the post-rectification caps in the PSU's and placed the 78/79l15 before the dc offset boards, I used to get a 'fart' through the headphones (equates to 1.6Vdc ) Now it's gone, which means I no longer have to disconnect the headphones before power off.

[FritzS]

Dec 20, 2009 11:40:39 GMT Will said:

One thing to mention is that before I replaced the post-rectification caps in the PSU's and placed the 78/79l15 before the dc offset boards, I used to get a 'fart' through the headphones (equates to 1.6Vdc ) Now it's gone, which means I no longer have to disconnect the headphones before power off.

I wonder how nobody had made a DC offset protector and power on delay for an headphone amp - the once I found is
ε12 ("epsilon 12") is a turn-on delay muting and DC offset protection
amb.org/audio/epsilon12/

wikipedia about HA
en.wikipedia.org/wiki/Headphone_amplifier