Building an OTL headphone amp - advice?

Jump to Latest
#23
dubadub said:
Greetings friends. I'm putting together a headphone amp based on the 6N1P Headphone Amp which is in turn based on the Morgan Jones 6DJ8 Amp and I had a couple questions about proper usage.

View attachment 1386147

As you can see, resistor R7 runs from output to ground. Does R7 dissipate the load if headphones are not connected? I've read it's best not to have headphones plugged in during power-up and shutdown. I have a switching jack; should additional resistance be added to increase the load when nothing is plugged in?
Any other OTL amp do's and don'ts?

I have HD600 but will do initial testing with AKG k240 (55ohm)

thanks for taking a look!
Click to expand...
I've actually done a complete diy build of this same exact circuit, in my build I used a 2W/5% 22k Metal Film resistor for R7 in both the left and right channel amplifiers, the amp itself only puts-out a few tens of milliwatts, so you can get-away with using 1W resistors, in the original HeadWize website article I used for my build, it says that the amp can drive headphones down to 32 Ohms impedance, I'm using a pair of Audio Technica ATH-M50X rated at 38 Ohms with my 6N1P OTL Headphone Amp build and it has no problems driving the Audio Technica with plenty of volume, to my ears, the sound is crystal clear at comfortable listening volume, nice bass, mids, and highs, very pleasant to listen to for extended listening sessions, I used my multimeter to measure the main +HT supply in my build, it was +306V DC on the + terminal of the last 47uF/500V supply filter cap in the power supply, this doesn't appear to have any detrimental effect on the sound quality.

So far my 6N1P OTL Headphone Amp build has been working perfect ever since I built it, with practically zero-issues, with no audio source connected to the inputs, there is some mains hum, which is to be expected since the inputs are very high impedance, with an audio source connected to the two inputs the amp is dead-silent with practically zero-hum even at full volume, I'm very happy with the results I'm getting with my build, the amp is a joy to use, it works with any audio source that has a stereo line-level output, I'm using three new Svetlana 6N1P tubes I ordered online from Evatco, in my build, they seem to sound good to my ears, note that I'm yet to try any new old-stock 6N1P tubes in my build.
I had a lot of fun building it and I would highly recommend giving it a go as a diy build project, I did have a bit of a hard time sourcing some 200 Ohm/5%/2W resistors and some 42k/5%/2W resistors, for the 200 Ohm resistors I ended up using two 100 Ohm/5%/2W Metal Film resistors in series, and for the 42k/2W resistors, I ended up using two 82k/5%/1W resistors in parallel, that gave me a 40.7k/5%/2W resistor which was "close enough for government work" and seems to work fine.
On pins 3 and 6 of the two halves of the 6N1P tube that's connected as a White Cathode-Follower output stage, I measured about 160V DC with my multimeter, incidentally, I had a bit of a hard time sourcing some 220uF/250VDC caps, and also, because space in and on the chassis was a bit limited, I elected to substitute a 470uF/450V DC can cap, for the two parallel-connected 220uF/250V DC caps, there seems to be no adverse effects on sound quality that I can hear, using a 470uF/450V DC can cap may extend the bass rolloff a bit.

I'll post a link to my 6N1P OTL Headphone amp build shortly.

Here we go: https://www.diyaudio.com/community/...ifier-using-three-svetlana-6n1p-tubes.416152/
 
  • Like
Reactions: dubadub
#24
dubadub said:
Thank you for taking a look. The HD600 would be the primary headphones for this amp, the AKG k240 would just be for testing purposes, but with the large difference in impedance maybe that's not a good idea.

My main question has to do with connecting the load to the output. Powering up a transformer-coupled amp with no load connected can quickly destroy the OPT - is there such a threat here? I'm worried about the Pop at power-up damaging the headphones, or a tube failure at power-up damaging them, so I'd rather connect the 'phones after a few minutes of warming up.

I already have a PT in hand, the Hammond 269AX. In fact, this thing is 95% built. Here's the power supply:

View attachment 1386210

I'm no good at PSUD2 but I expect 320-350v from this supply.

I'll try elevated AC heaters at first and build a DC supply if hum is an issue, but I'm concerned that the 2A supply is insufficient to provide 1.8A of rectified power.

I've also considered a variant of this amp with 6N6P as outputs, powered by the Antek AS-05T280, 280v@90mA 6.3v@4A, but I don't know what I'd drive it with. Seems this is a nice amp for the 300ohm cans.

w
Click to expand...
I've had practically zero issues with my 6N1P OTL Headphone Amp build, with regards to powering it up with a pair of headphones plugged into it, or with nothing plugged into the headphone out socket, the 6N1P tubes do take a little bit of time to fully warm-up from the instant the amp is switched on, this is an OTL, or Output Transformerless amplifier, it does not use an output transformer like most conventional tube based audio amplifiers do, so, since there's no Output Transformer, the risk of destroying an expensive Output Transformer due to no speaker load connected is virtually non-existent, so, therefore you don't necessarily have to have the headphones connected up to the amp prior to powering it up, the OTL Headphone amp will happily work without the headphones connected, and no damage to the 6N1P tubes should occur, unless for some reason they go fully conductive due to a fault in the amp itself, which is highly unlikely if the amp has been built properly with no short-circuits.
 
#25
Mark Tillotson said:
The problem with valve powered headphone amp without an output transformer is the lack of galvanic isolation from lethal high-voltages. Some fool proof protection circuitry to prevent lethal voltages appearing on the headphone cable is surely obligatory? Otherwise you could be one shorted capacitor away from a very nasty experience.
Click to expand...
To be honest, the only time I can see when either, or both of the output caps can go short circuit is either if their DC working-voltage is exceeded, or if the output caps end up leaking their electrolyte and dry up, in my build of the 6N1P OTL Headphone amp, which uses the same exact circuit, I measured about +160V DC on pins 3 and 6 of the 6N1P tubes connected as White Cathode-Follower output stages, +160V DC is less than the working-voltage of the output coupling caps, and I would recommend using caps rated to at least 450V DC caps, I used two 470uF/450V DC can caps in my build, which I recommend, 470uF/500V DC caps will also work perfectly well as output coupling caps, the working voltage needs to be higher than the highest expected voltage on pins 3 and 6 in the 6N1P White Cathode-Follower output stages.
 
#28
Every time I see this schematic, I think the same thing: "Three tubes for a stereo amp?" (symmetry people)

Just spitballing here. I have an impulse to take the second triode from V1 and use it in the same channel. You could double up V1 and put the two triodes in parallel. Anodes, grid and cathodes tied together one for one. [more spitballing] Is there an advantage to double up one of the two triodes that make the output?

Maybe I have an irrational fear on this one. But man, I do not like Mr Right Channel sharing components with Mr Left Channel when I can.
 
#29
audiomot The only real advantage as far as I understand, with regards to connecting triodes in parallel, is lowered output impedance, Fender connect the two triodes of a 12AT7 tube in parallel for use as a driver stage for a spring reverb tank, the plate/anode load is a small transformer that isolates the +HT from the transducers in the reverb tank, and it also lowers the output impedance, in the OTL Headphone amp circuit one triode is used as a standard triode voltage amplifier, and then the other two triodes in the circuit form a White Cathode-Follower output stage which provides a low impedance output for the triode voltage amplifier stage.
 
  • Like
Reactions: audiomot
#30

DrNomis44,​

"The only real advantage as far as I understand, with regards to connecting triodes in parallel, is lowered output impedance,"

And what is wrong with lowered output impedance? Especially when we are talking headphone amplifiers. Seems like a lot of discussion about design and topology for headphones amps is about output impedance and how to lower it.
 
#31
Tubes are inherently high-impedance devices, in the case of small twin-triode tubes, like for example the 12AX7, or ECC83, when the triodes are connected as standard common-cathode voltage amplifier stages, they have a very high input impedance and a high output impedance, for a 12AX7 the plate/anode resistance is about 62,000 Ohms (62k Ohms), tubes do not like operating directly into low-impedance loads because they can't directly supply enough current to drive a low impedance load without excessive clipping distortion, in order to overcome that, the output of the common-cathode voltage amplifier needs to be connected to an output stage that converts the common-cathode voltage amplifier's high output impedance to a low output impedance to better match with the low impedance load it is driving, we can use a cathode-follower, or common-anode triode amplifier stage to provide a low output impedance for our common-cathode triode voltage amplifier stage, a cathode-follower/common anode amplifier has a high impedance input and a low output impedance, but it only has a voltage gain of slightly less than one.
 
Last edited:
You must log in or register to reply here.
Top Bottom