Silicon Chip Headphone Amp Tweaks

[Will]

Thanks for the feedback guys, much appreciated. I was thinking of getting a pair of closed back after Christmas, so I'll bare this in mind.

[Will]

Oct 31, 2009 4:50:46 GMT Spirit said:

...I believe Will is (was? dunno what the B32S is driving) running his 701s with his Class A and loving it.

Spot on Phil, PC - B32s - AKClassA - K701, and indeed loving it.

I'm going to rebuild my SCHA in the next month or so, so that it incorporates my cheapo dac, allowing either normal audio in or via spdif. Might go USB input as well, if that damn musiland dac doesn't stop winking at me!!

[dean0]

Has anyone compared, OPA637 to LM4562HA op-amps in this amp?
Is the OPA637 compatible? I have 2 and a adapter.
Thanks
dean0

[robertkd]

Dec 5, 2009 12:34:28 GMT dean0 said:

Has anyone compared, OPA637 to LM4562HA op-amps in this amp?
Is the OPA637 compatible? I have 2 and a adapter.
Thanks
dean0

Sorry Dean0 missed this, looking at the data sheet for the OPA637 says

The OPA637 is stable in gains equal to or greater than five.

So may not be a good idea with stock gain of 2, to be honest I also don't like the stock layout either.

Changing the gain to 5 isn't hard but may also impact a little on performance but doable by changing the feedback resistor to 4KR or more and you might want to reduce the capacitor in parallel to something like 49pF.

<edit>
PS the OPA627 may be an alternative without modifications to the circuit

[punkjr]

Could somebody please tell me whether or not I should make the mods to the PSU without having the JLH fitted? It looks like the JLH may take a while.

[Deleted]

Apr 29, 2010 21:53:31 GMT punkjr said:

Could somebody please tell me whether or not I should make the mods to the PSU without having the JLH fitted? It looks like the JLH may take a while.

Tom
Rermoving the big 5W resistors will reduce the load on the power supply which was designed for use with the full SC Studio Series preamplifier.
The 5V regulator and it's related components are not required for this use either.The diagram in the Tweaks Thread should make clear what is required for best results.
Removing the 330 ohm 5W resistor will make available extra current from the -VE supply rail for loud passages of music.
Alex

[punkjr]

Okay thanks Alex. But where abouts is the following device fitted on the PSU..."A 470pF 100V (minimum voltage ) film capacitor is connected across the 2 outer terminals of the 3 pin AC Input block. This appears to further reduce diode switching "noise"."?

[Deleted]

Apr 30, 2010 6:23:30 GMT punkjr said:

Okay thanks Alex. But where abouts is the following device fitted on the PSU..."A 470pF 100V (minimum voltage ) film capacitor is connected across the 2 outer terminals of the 3 pin AC Input block. This appears to further reduce diode switching "noise"."?

Tom
The capacitor is soldered across the outer 2 terminals under the PCB, where the 3 wires from the transformer secondary are screwed in.
Alex

[punkjr]

Okay, so I'm running my amp in conjunction with a rotel rsp-1068 pre amp (running in stereo bypass, just really using it as a switch box and volume control). I have removed R3 and R6, and also the two 10uP NF's near the RCA inputs. I thought I read to do this somewhere, but can't seem to find where I read it now, not sure I've done the right thing.

[dean0]

Is it possible to use the transformer (MT2086. Jaycar) with 110v input (USA volatge)?
Thanks
dean0

[Deleted]

deanO,

If you have a potential buyer from a 110v country they will have to either swap out the toroidal for one made for 110v or use an external step-up transformer.

Some transformers do wire up differently for different supply voltages, this one does not.

[dean0]

thanks for the info

[Deleted]

I've just put my SCHA back inside inside its case. It was dismantled for a while to use its PSU and JLH to aid the testing and checking of other kit. Then only to be packed away, still dismantled. :-(

The only changes I made before replacing the lid was properly screened internal audio lines (only twisted previously) and the capacitor change on the JLH that Alex reported a wee while ago.

I let it run for several days before plugging myself in.
I've swapped directly from the MF XcanV2, I really do not remember there being such a huge difference before!

The extra authority and clout, along with deeper reaching bass, speed and clarity.

Maybe I had just forgotton?

There is a big cheesey grin on my face now

[Deleted]

Hi Chris
The change to 4,700uFs in the JLH does give you a little more of what you described.
They fit nicely laying down on Jon and Will's PCB too .
Presently I am running in a JLH in a new +5V Linear PSU for the Corsair Voyager.
Alex

[sp]

I've built a couple of these headamp boards. I have only modded as far as discarding components, or whatever I can pickup at Jaycar. I have a Lehmann Black cube linear amp that retails here at well over 1k, and I gotta say - this amp holds its own.

I bought the LM4562 from Altronics and it never actually "synergised" with the Jaycar amp - I found I preferred the stock unit. Well, I just put the fast switching diodes in, removed the PSU resistors and deactivated the 5 v rails - and the LM4562 now sounds very open and less restrained like it was before - how odd. The diode swap made quite a difference. Still burning in, it might surpass the Lehmann amp - I hope so.

I have a second board ready for bridging to run it balanced with the Bryston BDA-1 DAC - this will be my first balanced amplifier, but recent experiences suggest that balanced circuitry makes quite a difference. Currently using a borrowed pair of HD800 headphones and I own a pair of LCD2s as well.

It turns out that I have a bit of a clue with regards to reading an electrical diagram - its the same as reading the wiring diagram of modern vehicle electronics. I was even able to diagnose why half my LEDs didn't work - after checking voltage drop of every component to ground I found a brocken track on the PSU so -15V didn't work.

Being encouraged by this...I bought and built the Altronics DAC kit without a hitch. I am enjoying this new found interest and thought I would sign up to say hello.

Sam.

[Will]

Hi Sam, and welcome to the RG!

The description of the sound of the LM4562 is pretty much what most people think, compared to the original op-amp. PSU improvements are a very worthwhile move, but I'd suggest having a look at Chris's (cjarchez) post two posts back. That description is what the JLH ripple eater does for the LM4562, it's a huge step forward!

[sp]

I need to look into these JLH boards.

Meanwhile, I finally got around to removing the input capacitor...this was the change to make this amp on par with the Black cube linear in objective raw speed and transient reponse. With bigger low ESR electrolytics the subbass extends lower than the BCL with a corresponding increase in lower midrange details which makes for a subjective improvement over the BCL. So the input capacitor makes a massive difference.

I fitted the LM4562 the wrong way around the other night and blew both the fuses...I knew I left the fuses there for a reason, everything still works and I don't think anything was damaged...it sounds the same.

At this stage I am very content to know that a 50 dollar kit amp can objectively match a $1000 retail German made professional amplifier, I also subjectively prefer it, because of the lower subbass extension which adds to the midrange body and hence the presentation is less analytical.

Both amps share the same voltage regulators and transistor to boost the opamp input. The only drastic difference in topology is that the BCL has no feedback into the the opamp where the jaycar unit does.

[lark]

Thanks Alex and crew for your assistance - The last few components arrived today to complete and test my new SCHA with Alex's mods and JLH Shunt.

I powered it all up - no bang (this is good).
Plugged in a source - no bang.
Plugged in some (cheap) phones - still no bang.
Turned up the attenuator - no bang, but sound from phones.
Swap out the cheap phones for my 702's, whack them on my head and enjoy the sound..

I'm using a 50k attenuator and find I need to wind it up to 3 or 4 o'clock to get normal listing levels.. Is there something wrong on my board or is this normal? - It will get loud enough if I crank it the rest of the way though so I don't *need* more volume.

Also, I've left the SC power supply circuit as per the original design and the 5w resistor is cooking hot such that I can't touch it comfortably. I have nothing attached to 5v. Is this normal?

Cheers.

[Deleted]

Hi Lark,

As Alex is probably asleep by now;

The physical volume level is normal, it is part of the design for the gain to be set this way. It can be increased if you decide you need it (I did).

I would highly recommend modding the power supply as per Alex's instructions. The 5v o/p can be used for a power-on LED if you like.

I'll leave more detail to Alex when Sydney wakes up

Oh, and congrats on a successful build

[Deleted]

May 18, 2011 14:15:32 GMT @cjarchez said:

Hi Lark,

As Alex is probably asleep by now;

The physical volume level is normal, it is part of the design for the gain to be set this way. It can be increased if you decide you need it (I did).

I would highly recommend modding the power supply as per Alex's instructions. The 5v o/p can be used for a power-on LED if you like.

I'll leave more detail to Alex when Sydney wakes up

Oh, and congrats on a successful build

Hi Lark
As Chris says.
I suggest you remove those surplus PSU components as per the Tweaks thread. It will run cooler, and more current will also be available for program peaks, instead of being wasted as heat.
This PSU was designed originally for the Studio Series Preamp, where the +VE rails are loaded much more heavily than the -VE rails due to a remote control PCB, switching relays etc.

Too many people seem to think that normal listening volume at 9 or 10AM on the volume control is a good thing.The thing to aim for is MAXIMUM USEABLE volume level just before full rotation.
This will ensure a better signal to noise ratio etc.
Some heaphones are less sensitive than others, in which case the gain may be increased a little if necessary.
Regards
Alex

[lark]

Ok, so I've removed:

330R 5w resistor
47uf 25v cap
7805 (Reg 3)
10uF cap (near Reg 3)

Placed a wire link between the two outside holes for Reg 3

My PSU had a 1N5338B (ZD1) where Alex's mod page has 100R 5W (R1) which is suggested to be changed to a 10K .5w resistor. I found some images of other peoples modded PSU and saw that they replaced the diode with a 10K resistor - so I went ahead and replaced ZD1 with the 10K .5w resistor..

BUT.. Now I power it back up and it get nothing (0 volts) from the -15 & +15 rails and 22v DC from the 5v Rail..... Hmmm what's gone amiss?

Photo of the component side


Photo of the solder side (excuse my flux mess)


See any obvious problems?

[Deleted]

lark
I can't see anything obvious.
+22V out from the original +5V terminals is correct, as it now available to use for connecting a front panel LED from those terminals.The PSU should now look similar to the attached schematic,other than the actual voltage setting resistor values around the LM317 and LM337.
First off, carefully check for around +22V to the input pin of the LM317 , and -22V to the input pin of the LM337.
Don't forget that they have different pin assignments.
Try meauring from th earth screw of the +5V terminal block to the + and -15V terminals. Sometimes the screw terminals don't make contact properly until they are screwed up.
Alex

P.S.
If you have some Isopropyl Alcohol, turps etc, try cleaning under the PCB with an old toothbrush, then have a good look for any possible dry joints or solder bridges, especially around the 3 pins of the voltage regulators.

Attachments:

[lark]

Ok, I've checked the voltages (337 middle pin = -22v) (317 pin closest to the 47uF cap location (removed) = -22v)...

Still 0 volts at the output terminals...

[Deleted]

lark
As well as the added bits to my original reply, also measure the resistance from +VE out, and -VE Out, to the centre screw. (earth) You should read 1.2K for each, which is the series resistance of the 1.1K and the 100 ohms resistors that sets the output voltage.
Alex

[lark]

Thanks again Alex,

I figured it's most probably a dry joint, I took the brute force approach and ran my iron over all the joints again and hey-presto it's all good again.

This is just a reminder that my soldering skills still have some improving to do!

Cheers.