My adventure with the HT SMPS modules continued. Did further testing with higher power modules which worked ok.
The headphone stage (HP) has a 1MHz bandwidth, impressive. Here is the circuit breadboarded after the modification of the 2P29L HP amp:
The cap multiplier isn’t mandatory as the gyrator PCB has a high PSR so a good well-filtered HT supply will work. I added the cap-multiplier for 2 reasons. I have the PCBs made and also I was looking to improve the supply noise of the SMPS modules. My intension was to power this up with an LT supply module I have based out of 2 variable SMPSs.
The Chinese 40W HT SMPS modules that can do 450V work fine. They need a good HF filter as well as a proper LC stage to decouple and filter the ripple to make it absolutely dead quiet.
Here is the variable HT supply I build some time ago which I use for multiple circuits:
With this stage I have the choice of swapping the output caps. Either a massive (aka expensive) Mundorf EVO Oil 100uF/350V for direct coupling of the 300R headphones, or via the Sowter HP output transformer with a smaller cap.
I personally prefer the OT coupling in this circuit, but both sound extremely well.
More work and measurements to follow.
Hi Ale, Don’t talk about the rugby!
Did you ever do a production run of your cap multiplier board?
cheers
tim
Yes, no rugby talk then!
Not yet, I need to find the time to do so. I’ve been tweaking the board a bit but think it’s on its final level now. I may add an HT indicator (e.g. Neon bulb like the INS-1 as I have plenty of them)
Will let you know when I run a production batch. I have quite a few other PCBs which I developed and are being very useful at this stage:
1. HT supply board – flexible use of hybrid rectifier (e.g. SiC diodes and damper valves) as well as chokes
2. LT supply board for DHT – with external chokes and on board CMR choke
3. Fixed bias circuit for grids with on-board trimpots (or external pots) and grid coupling circuit. It relies on Rod Coleman’s fixed regulator or any other fixed bias regulator
4. Fixed bias protection circuit and anode/cathode current meter. This is the one I used on this headphone amplifier. Only because I had the prototype PCBs available!
Cheers
Ale
Thanks Ale,
I have used a ST901T darlington in a simple CM to good effect.
best
tim
HI Tim,
I’ve never tried the ST901T but seems like a very good option up to 350V. It has higher gain than the MOSFETs so should perform better. Will need to buy a few and trace them first. I think I could modify the PCB to cater for a darlington based BJT pass device instead of the MOSFET. For higher voltages, the MOSFET is cheaper and better suited
Ale
there is a 2kV NPN BJT Darlington part in TO-220, but it’s a bit pricey: STP03D200
https://www.mouser.co.uk/datasheet/2/389/stp03d200-956465.pdf
Hi Ale,
which chinese modules did you use in this project?
Regards
Here’s a recent post in AliExpress
￡8.29 28％ Off | DC12V to DC 200-450V 70W High Voltage Converter Boost Step Up Power Supply Breadboard
https://s.click.aliexpress.com/e/T9reiXqOr
Dear Ale, always happy to read your posts, very informative ! A question about the PSU filter : what EMI filters did you use to put 450V DC in ? Anything standard ? Or self built ? A part # would be very appreciated !
Kind regards, Charles
Hi Charles,
I used pre-built AC mains EMI filters, they do the job perfectly well here.
Ale
Now am listening to the cap-couple setup. The Mundorf EVO Oil are very good caps. The sound is very detailed, the dynamics are there and the overall frequency response is superb.
The 6e6p-DR are running at 22mA. I have jumpers across the array of 6 SiC bias diodes in the PCB. Easy to change as needed depending on the bias requirements.
When I get the time I will match a few pairs with the eTracer. Russian valves are famous for having their parameters all over the place. I found a large stash of 6e6p and 6e5p in my storage, however I can’t locate the golden pin 6e6p-DR stash, argh! I’m playing with just a pair I had at hand. This is what happens when you move places twice and packed your workshop. I don’t know where half of my stuff is now.
You can hear the “dings” when the valves warm up, it take a few minutes and after that it’s all good. The stage is dead quiet. Despite I could see some HF noise due to the filaments being fed by an unfiltered SMPS Meanwell variable module. I do this because is an IHT, not a DHT. You can’t get the DHT working with the SMPS in this application. It’s very hard to get the noise that low. At least, I failed after a few attempts. Do-able with the 2P29L as is low current and you can easily build a good CLC filter to get rid of the noise over 10kHz.
I need to take this back to the bench and do a proper FFT to see where the harmonics and noise are. Either way I can say that I like very much what I hear, the amp is dead quiet. I’m not getting listening fatigue.
A 32R headphone will need the step down transformer in parafeed mode for sure. The hybrid mu-follower will have too much distortion with it and will struggle to supply the current needed getting into slew rate distortion.
If you want to build this, please go ahead. A great HP amplifier to enjoy. I’d recommend to go with a normal PS supply, not SMPS. Not worth the hassle in my view. I did because I had the PS already built for my experiments.
Hi Ale,
I have no experience with 6E6 (or 6E5) tube, to my shame. I’m not moving from german postal tubes yet 🙂
Seriously, many 6E tube users say 6E5 is a “better sounding” tube than 6E6.
Definitely, your opinion is … ???
Hi Rajko,
It’s been a long time since I’ve used the 6e5p as switched to the 6e6p given pinout being similar to D3a, etc. 6e5p has an odd pinout. I will need to re-wire the socket to plug in the 6e5p and compare. That will have to wait so I could do a proper AB comparison. My previous experience with the 6e5p was very good as well and can’t recall any materia difference I can call out am afraid.
I will likely test D3a, E280F, E810F and similar triode-strapped pentodes before making any changes to the HP amp
Thanks for the suggestion!
Ale