I have SE with EL84, and soon another one with ECL82. They are great for full range.. There is hum and I cannot get rid of it (maybe because of all electronics in that space, TV, monitor, WiFi router, etc etc..). The Yamaha amps I use now have no hum at all. I don't stand hum (the mid and highs are also horn loaded.)
Hum is the price you pay with tubes. Try supplying direct current into the filament with a current source, I think you need 760mA at 6.3V, and you can do this with a CRC off a 6.3Vac winding, small full bridge, and two 6,800uF caps at 10-16VW rated with 1R 2W resistor between the electros, then into a LM317 set up for 6.2V to bring the voltage down to 6.3Vdc. Make two separate supplies for each tube but set it up for two tubes, four for two channels.
And yes, make the tube in a different metal box to keep away the EMI from digital electronics.
Cheers,
Hugh
And yes, make the tube in a different metal box to keep away the EMI from digital electronics.
Cheers,
Hugh
hell just froze over....
While the simple class-A circuits favoured by the SE DHT crowd do have lousy PSRR performance, there's neither a need nor an excuse for hum in tubes any more than there is in solid state electronics.
In my experience hum is usually caused by a combination of bad power supply design; poor wiring & layout and an assumption that "indirectly heated" means "no coupling".
Now, DHT tubes have their own charms but, again, there's no need to have hum. Either proper humbucking or a decent current source supply will result in zero hum.
As to "why not SE DHT" doing it properly is neither cheap nor efficient (damn'd heaters). Decent PP is possible but also not cheap (e.g. Olsen's Karma)
One should also consider the JLH'69, various Hiraga designs or look at Parker's Zero Feedback Impedance amp.
While the simple class-A circuits favoured by the SE DHT crowd do have lousy PSRR performance, there's neither a need nor an excuse for hum in tubes any more than there is in solid state electronics.
In my experience hum is usually caused by a combination of bad power supply design; poor wiring & layout and an assumption that "indirectly heated" means "no coupling".
Now, DHT tubes have their own charms but, again, there's no need to have hum. Either proper humbucking or a decent current source supply will result in zero hum.
As to "why not SE DHT" doing it properly is neither cheap nor efficient (damn'd heaters). Decent PP is possible but also not cheap (e.g. Olsen's Karma)
One should also consider the JLH'69, various Hiraga designs or look at Parker's Zero Feedback Impedance amp.
Last edited:
How to get rid of hum for $3:
Us a cheap DC to DC buck-boost converter that uses a 400kHz to 1.2MHz switching speed to take all the noise well above audio band. Apply a decent CRC filter after the built in LC filter. Power the step up from any SMPS wall wart.
You can actually use these to power the main rail in a Class A amp. I have done it for 1.4amp bias at 32v even. No 60/120/180 Hz content at all in FFT's.
Us a cheap DC to DC buck-boost converter that uses a 400kHz to 1.2MHz switching speed to take all the noise well above audio band. Apply a decent CRC filter after the built in LC filter. Power the step up from any SMPS wall wart.
You can actually use these to power the main rail in a Class A amp. I have done it for 1.4amp bias at 32v even. No 60/120/180 Hz content at all in FFT's.
Choke inputs rule
Afterall, they were heavy, expensive and produced less voltage.
But their technical performance is still way, way better than capacitor input power supplies.
on my bucket list is a (c)LC filter a,k,a, choke input for a solid state amp. They were always the n'est plus ultra for tube circuits but disappeared when solid state rectifiers and cheap electro caps turned up.
Afterall, they were heavy, expensive and produced less voltage.
But their technical performance is still way, way better than capacitor input power supplies.
Last edited:
Hi Thog,
I have tried CLC on a SS amp. But unless it is a Class A, you must use small inductors because otherwise the screw up with the voltage supply because of the wide current variations.
Class A is different and it works well, like a tube which operates at relatively constant current. You can use common mode chokes in Class AB SS; the P3717A Coilcraft works well on my Class AB amps. But they are tiny inductances, and thick wires to keep DCR low.
Fixing tube hums, particularly DHT, is akin to maintaining tune on a hopped up hog.......
Hugh
I have tried CLC on a SS amp. But unless it is a Class A, you must use small inductors because otherwise the screw up with the voltage supply because of the wide current variations.
Class A is different and it works well, like a tube which operates at relatively constant current. You can use common mode chokes in Class AB SS; the P3717A Coilcraft works well on my Class AB amps. But they are tiny inductances, and thick wires to keep DCR low.
Fixing tube hums, particularly DHT, is akin to maintaining tune on a hopped up hog.......
Hugh
Thanks - not many people seem to have tried this.
In theory a "swinging" choke with an "appropriate" airgap should give better regulation than a conventional cap input circuit.
But until I try it, it's all just idle speculation.
heh heh! Having been owned by (old) SUs on a BMC B-series donk for a while...
Last edited:
A new Quasi developed in partnership with Hugh Dean is born: the HyQu, a Hybrid Quasi headphone amplifier using smaller output devices and single rail power supply, with of course, cap coupling on the outputs to ensure safe operation with headphones. Here is my hand etched prototype running on 24v rail and 95mA bias current. Sounds great and measurements are not too shabby either. Here are FFT's for 1.41vrms into 270ohm load - about 0.003% THD with about 29ppm H2 and 9ppm H3 and descending higher harmonics with a -130dB flat noise floor. Suppression of the wall AC mains 60Hz worked out pretty well too. This will be a product line so sorry, can't share the details of schematic.
Left:
Right:
Left:
Right:
Attachments
Last edited:
Eric,
You are spoilt for choice on this forum!
Some guys, though, would like to exploit their skills for commercial products, and you might be interested to see that good circuits here really do make it in the world market. At least XRK is straight up about it; not like the silent Ebay copy cats who tell no one anything they are doing..... with other people's intellectual efforts.
This circuit is exceptional and you can build the quasi any time for any number of great pcbs given free by members here.
Ciao,
Hugh
You are spoilt for choice on this forum!
Some guys, though, would like to exploit their skills for commercial products, and you might be interested to see that good circuits here really do make it in the world market. At least XRK is straight up about it; not like the silent Ebay copy cats who tell no one anything they are doing..... with other people's intellectual efforts.
This circuit is exceptional and you can build the quasi any time for any number of great pcbs given free by members here.
Ciao,
Hugh
No harm, no foul. Thank you XRT for sharing the design concept, photo and plots. It gives us DIYers new ideas on how to use this circuit.
Have you considered selling this as a short form kit set consisting of pcb and board components, in the same vein as the original AKSA 55? Would seem to be a good candidate; low parts count, simple construction. If so I suggest that the final, debugged circuit be laid out on a single pcb comprising the stereo amplifier and power supply - all in one - just like Hugh's boards.
Some nice, compact, aluminium enclosures are available from China that would suit this nicely. Orient the output devices so they are underslung and bolt directly to the alloy chassis for heatsinking.
Have you considered selling this as a short form kit set consisting of pcb and board components, in the same vein as the original AKSA 55? Would seem to be a good candidate; low parts count, simple construction. If so I suggest that the final, debugged circuit be laid out on a single pcb comprising the stereo amplifier and power supply - all in one - just like Hugh's boards.
Some nice, compact, aluminium enclosures are available from China that would suit this nicely. Orient the output devices so they are underslung and bolt directly to the alloy chassis for heatsinking.
Hi Christian,
Thank you for your understanding and consideration in making this a kit. I have actually been thinking exactly along these lines. Layout would indeed be single board with integrated PSU and under-hung outputs to for the standard aluminum chassis that is available on eBay/Aliexpress. What's neat about this amp for headphone use is that, although relatively cool running (5w dissipation), it runs in Class A for 99.5% of most headphone use - that is up to about 500mW to 700mW info 55ohms depending on supply rail). That's actually threshold of pain in many headphones.
but it provides that dynamic headroom needed on some headphones such as inefficient planar magnetic ones that are only around 82dB sensitive. Those particular units are looking for up to 2w - which is achievable with higher voltage supply rails. All the while retaining this incredibly beautiful harmonic distortion profile. The above measured profile is in fact, one of the best examples I have measured in any amp, ever.
Last edited:
I can see you've given some good thought to this already. I've been thinking about buying a quality pair of headphones for early morning listening and this could be the excuse I've been looking for. Please keep me posted and consider me your first customer!
A word to the wise: listen to Hugh's advice in all matters relating to parts dimensioning and board layout. However.... if Hugh ever suggests a certain six legged part containing dual BJT transistors then POLITELY DECLINE for my sake and that of your other kit builder!
Hugh knows what I'm talking about
A word to the wise: listen to Hugh's advice in all matters relating to parts dimensioning and board layout. However.... if Hugh ever suggests a certain six legged part containing dual BJT transistors then POLITELY DECLINE for my sake and that of your other kit builder!
Hugh knows what I'm talking about