1626 HPA

[Cressy Snr]

So now I have half a dozen 1626, courtesy of Mr I and Paul B (thanks a lot chaps) and I’m on with building a headphone amp with them.

This means a complete strip-down of the existing 6V6 HPA, and a re-think of the whole power supply setup, as the mains transformer produces far too high an HT value for use with the 1626. I’m using a 12.5V mains toroid for the heater supplies to the input and output stages and a 4V CT frame transformer for the rectifier heater (UU10.) The HT is provided by 230V-230V toroidal isolation transformer.

A lot of careful mechanical work on the chassis was done today as I didn’t want it to become a dog’s dinner.
I have heaters glowing on the front row of valves plus heat to the rectifier.

I’m leaving the PSE DC coupled Darling for a later project as I want to hear the valve as a headphone amp first. The influence comes from the Ampsandsound “Kenzie” headphone amp, which has had good reviews.

[Cressy Snr]

The chassis now:

IMG_0490.jpeg (114.07 KiB) Viewed 9760 times

I had to kludge a piece of aluminium over the carnage at the rear of the chassis after the removal of the Hammond mains transformer and choke. It looks reasonable I think.

There was a Darling amp with James output transformers at the Darling Fest 1999 but not the 6123HS I have on mine; they were the smaller, cylindrical models.

My 6123HS are of course total overkill for the minuscule amount of power available, but they are the best I can do and it didn’t stop me using them to good effect, with AC/P power valves years ago.

[Cressy Snr]

It’s all working and free of hum:

IMG_0492.jpeg (155.54 KiB) Viewed 9641 times

I had to remove the cover from the mains toroidal transformer as it was causing mechanical noise.

[Cressy Snr]

I followed the Darling style build philosophy, including shared cathode resistors on the input and output stages and both stages fed from the same HT node without any decoupling.
Input valve is the ECC81 which has a similar distortion pattern to the 1626, so I was hoping for some cancellation, though this is always a hit and miss situation:

1626HPA.png (71.36 KiB) Viewed 9624 times

[Cressy Snr]

Power supply for HPA:

HPAPSU.png (102.83 KiB) Viewed 9582 times

[Cressy Snr]

OK,
Listening to the headphone amp last night in the dark I noticed a subtle red glow from the anodes of both 1626 valves. Now, both Epstein and Danielak reckoned the 5W dissipation limit could be exceeded and the valves would run happily at 6WPd. Mine were dissipating 5.6W so I'd thought that would be OK. However, it was not.

I revised the power supply by doubling the R part of the final RC filter to 1K5 and adding 30R to the cathodes of the output stages to give a 500R cathode resistor. Those mods have taken a Watt off the output stage dissipation, which is down to 4.6WPd and all appears to be well with the circuit now.

Revised Schematics:

1626HPANewVoltages.png (71.52 KiB) Viewed 9459 times

1626HPAPSUV2.0.png (102.75 KiB) Viewed 9459 times

[Cressy Snr]

Inside:

1626HPAInside.jpeg (149.54 KiB) Viewed 9448 times

Rectifier heater transformer on the left.
12V heater toroidal transformer on the right.

[Cressy Snr]

It sounds good. Lovely valves 1626s.
As I said earlier, these flea powered SETs make perfect headphone amps.

[andrew Ivimey]

Neat.

[Cressy Snr]

Cheers Andrew,
More movement on the headphone amp front. Only because I can, there are no other reasons other than I like to mess about.

Anyway, by carrying out these mods, I've put the fume extraction through its paces over the past couple of days and I've had no further health problems.

I've given up making mods to equipment, then claiming everything sounds better/fantastic/fab etc For me, it's a bit of harmless fun, rather than a quest for perfection and valve headphone amps are the latest avenue I've found to explore. It's quite a rewarding little niche.

So...the drivers are now a pair of GE 6AU6 in triode mode and the input sockets have been moved to the front:

HPA_6AU6.jpeg (136.06 KiB) Viewed 8761 times

It still sounds fab though.

[Cressy Snr]

Yet more schematics:
LED bias doubles as the on/off indicator light

1626HPA_6AU6Driver.png (65.94 KiB) Viewed 8760 times

PSU has an additional 8uF as the first cap, so that the maximum 'condenser' value for the UU4 rectifier is not exceeded. The supply is therefore CRCLCRC. The 1k5 resistance has been equally split over the supply, so that the output voltage is roughly the same as before.
The background is silent which is not bad at all for an AC heated valve lineup right next to the old lugholes.

1626HPAPSU_V4.png (94.71 KiB) Viewed 8760 times

[Nick]

It would be interesting (well, fun), to do some crosstalk measurements, maybe crosstalk against frequency to see what those shared cathode components are actually doing. I have always wondered about them (though not enough to actually try them and measure myself evidently).

[Cressy Snr]

Yeah,
It might very well act as a cross-feed setup. I remember cross-feeds being quite popular in the 1970s. At that time, cross-feed manufacturers seemed to be selling the idea of a more natural soundstage when listening to stereo via headphones.

As you say Nick, there will probably be a frequency dependent cross-feed action going on and also a frequency dependent phase shift happening.

There seems to be an effect which places extreme left and right images outside the head so that the soundstage is quite wide, which is what you want with headphone listening. That'd be a phase phenomenon I think.

I'm certainly getting a a more convincing soundstage than the kind of thing I used to get in the late 70s/early 80s when I last used headphones (Sennheiser HD400.) On speakers however, the presentation might come over as if the signal were going through an effects box named 'Hall' or 'Stadium.' That might be just what you want, or it might not depending on your point of view re fidelity to the original signal.

There's a bloke on DIYAudio who has designed an op-amp based cross-feed setup grafted into a Musical Fidelity X-Cans HPA. That probably works more predictably and with far more user control than the simple shared cathode job I'm using.

Link below, which is actually quite interesting, as he did some work in LT Spice before building the device:
https://www.diyaudio.com/community/thre ... st-7721272

[Mike H]

Anyway, by carrying out these mods, I've put the fume extraction through its paces over the past couple of days and I've had no further health problems.

Nice to know.

[Cressy Snr]

The HPA continues its evolution.
I've attached a little stand at the rear to keep the headphones on. The stand used to hold a set of three chopping boards but had not been used in years, so I thought I'd change its use.
The base is glued to the wooden rear top-plate but the vertical part that holds the headphones can be pulled out, so that access to the underside is not impeded.

Stand.jpeg (129.46 KiB) Viewed 4600 times

Also round the back we can see a Black Gate WKz, 220uF + 220uF, 350V continuous, 400V surge capacitor.

IMG_0510BG.jpeg (96.88 KiB) Viewed 4600 times