zenkit1 with akg714pro or sennheiser hd650

raduschka

Member
2011-01-10 5:31 am
Hello, I apologize if I'm posting in the wrong place. I built a PTP 2watt/channel SET amp ( Steve Deckert's zen1, or SE84) with tube rectification (5Ug4). I have a Minitorii so I would like to use the zen as a headphone amp, however it produces an unpleasant hum, even after drastic improvement through increased C2 capacitance (110uf). The part where I am lost is the resistance that I assume i have to add in series with my headphones. I have tried various resistors in series (headphone out is tapped from speaker posts). 100 ohm -still too much hum, 330 ohm and certain input tubes - I can live with the level of hum. My headphones (AKG k712pro [IMGDEAD]https://flic.kr/p/FBLZPu[/IMGDEAD]) have a sensitivity of 105 dB SPL/V
Max. Input Power 200 mW
Rated Impedance 62 Ohms.
Does this sound feasible? Am I doing something wrong? Can I add resistance? Am I frying the amp?
So far sound and headroom are excellent.

Thank you very much for taking the time.

Radu Christian

pics date from before I installed the headphone jack

[IMGDEAD]https://flic.kr/p/nQS8KX[/IMGDEAD]
[IMGDEAD]https://www.flickr.com/photos/[email protected]/14316270855/in/dateposted-public/[/IMGDEAD]
 
I found the schematic for this amp but it does not have any part designators, so which one is C2? The filter cap on the A+ supply?

Also, please attach images.

If it depends on input tube, chances are you need to clean up your B+ supply further, and/or get in electrolytics in parallel to the 3.3 µF poly caps, preferably both. Might also have something to do with the AC heating though, DC heating would be preferred for the input tube at least. Possibility number 3, shielding and grounding.

If you do need to drop output level resistively, use a loaded voltage divider (of several hundred ohms and adequate power ratings) rather than a simple series resistor. As it stands, the amp only emits about 3 Vrms tops though, so if you get anoying levels of hum like that, dynamic range is not very good at all absolutely speaking. Hum and noise spec is <1.0 mV, which in solid state would be kinda average for a 45-50 dB amp. This one has much lower net voltage gain, maybe around 20 - 30 dB tops depending on input tube mu. But of course, doing it all on two tubes with common cathode circuitry means that PSRR is pretty bad, so you need to clean up your supplies extremely thoroughly. Solid state rather invests in circuit complexity.

But speaking of circuit complexity, you could try replacing the 47k carbon plate resistors by either adjustable current sources or gyrators (typically solid state). That should improve PSRR on the first stage considerably. Theoretically you could also use a plate choke, but finding or making one that provides hundreds of kOhms of impedance at audio frequencies is not at all easy.
 

raduschka

Member
2011-01-10 5:31 am
Thank you! I feel relieved to have feedback on this amp that has been giving me headaches for so many years now in various incarnations. I guard the kids today, hopefully later I can post pics. If i remember correctly I did try heating the tubes with an external DC source (It must have been an ordinary 12v printer adapter )and it did not have any impact on the level of hum. Through the sensitive akg702 the hum is overwhelming. Also through 100db or so 10inch Nirvana fullrange speakers. I,ll take out the voltmeter ans measure, the speaker posts do not have resistors in series.
 

raduschka

Member
2011-01-10 5:31 am
...er, by guarding the kids i mean I'm alone at home with them, gf and I are happily together :)

This is hopefully the schematic you found:

http://www.decware.com/newsite/ZKIT1schematicrev2011.pdf

filter cap right after 5u4g is 40 uf, high voltage (500 i believe, I'll look). C2 is indeed following right after and is by now experimentally around 110uf, by piling up parallel caps i have lying around.
What value should I use on the 3.3caps? I suppose a low one. I only have film caps in those values ( 0.1, 0.22 uf).
Thank you again!!!!
 

raduschka

Member
2011-01-10 5:31 am
...tried to measure the hum, i cannot get a useful reading with my radioshack voltmeter.
i will use my ears. This amp is reputably dead quiet with sensitive speakers and at least one user reports it is dead quiet with 100 ohm resistors on the output, but he does not specify headphone sensitivity/impedance. Could it be that i experience normal hum? I thought normal hum was no hum :)
 
C2 is indeed following right after and is by now experimentally around 110uf, by piling up parallel caps i have lying around.
Then tackle the two series 20µFs (on B+) next, that should have a major effect as the preceding resistor is a 6k8 rather than 1k. I suppose B+ is around 250 volts, so a 350V part would be good. (I would definitely do away with the two series caps. Should you decide not to, give each one a parallel resistor of, say, 100k to 1Meg. Otherwise voltage sharing depends on capacitor leakage current.)
What value should I use on the 3.3caps? I suppose a low one.
No, I was thinking more like 10-33 µF in parallel or so.

You may want to keep an eye on voltage on B+ and after the 10k resistors, assuming your meter will manage. There may be some overshoot or oscillation if you get unlucky and accidentally create a double pole in RCRC filter response. Right now we have time constants of 6k8 * 10µ vs. 10k * 3µ3, that's about a factor of 2 apart. This could also be more, just avoid having them become about equal. Don't forget about the A+ filter stage either, which now is at 1k * 110µ = 110 ms.

Aside from that, the only other side effect that may appear is very long turn-on time. Right now I wouldn't see a problem with going a factor of 3 or 10 higher though.

EDIT:
...tried to measure the hum, i cannot get a useful reading with my radioshack voltmeter.
I'm not surprised. That's what they used to have audio millivoltmeters for. These days, if you have a way to calibrate the input level on a laptop mic/line input (e.g. by doing a loopback recording of a test tone of up to a few hunded Hz while tapping output voltage with a multimeter), that may get the job done. Should be battery-operated to avoid ground loop issues.
 
Last edited:

raduschka

Member
2011-01-10 5:31 am
I should get an audio multi-meter AND an oscilloscope! Just for tweaking this one amp, it's been a real zen dilemma, always asking to me to let go, let go... and hope that one day it would work just fine. I'm about to think that I,ll get a pair of sennheisers hd650 and solve the issue. I have an older pair of akg, along with the k702 and k712, and their impedance is higher, sensitvity lower, and indeed, the hum is way less audible.