I can't resist cheap, Chinese tat from eBay, so I went and bought this https://www.ebay.co.uk/itm/TU-2-WCF-Modified-6N2-6N6-Tube-headphone-amplifier-DIY-Kit-No-tube/. Turns out to be a bit of a dud, lots of distortion using a standard, low impedance (18 Ohm) pair of headphones. I can see it mangling a sine wave on my (also cheap Chinese) oscilloscope. Driving a 600 Ohm resistive load shows no distortion but I don't have a pair of pro, high impedance headphones to test with. Can the circuit (attached) be made to work with low impedance headphones? Would an output transformer help, or should I just chuck it in the bin?

 

Personally, I wouldn't use an OTL amplifier on headphones, but it looks like this one really needs the transformer anyway. There is nothing but a capacitor between the high voltage tube supply and your headphones.

An output transformer to match the high impedance tube output to the low impedance drivers in the phones should do it, but it will introduce group delay, etc, so it needs to be a good one or what is the point in using tubes?

 

Thanks for that. What should I be searching for to get something half decent of the correct spec? I've been happy with the Lundahl step up transformers that I got for my MC cartridge but don't know what to look for on their website that looks like a suitable transformer.

 

Thanks for that. What should I be searching for to get something half decent of the correct spec? I've been happy with the Lundahl step up transformers that I got for my MC cartridge but don't know what to look for on their website that looks like a suitable transformer.

Unfortunately, nothing inexpensive appears easily in searches but this is the idea: https://www.ebay.com/itm/STANCOR-12...MP-SINGLE-END-SE-11-WATT-6L6-6V6/323735628366

 

So, I've now stared at datasheets for audio transformers for long enough to make my head seriously ache. If my test with a resistive load of 500 - 600 Ohms gives me zero distortion, will a transformer of 600 Ohms (impedance) on the primary work for me? I've been looking at the Hammond range (https://www.hammfg.com/electronics/transformers/audio/143-146.pdf) which has a whole bunch of options, but none of their lower output impedances, 8 or 50 Ohms, matches my 18 Ohm headphones. And what's the difference with 500 and 500ct for the input impedance?

 

8Ω should work.

 

500 ohm ct is for a different type of output stage topology. If you wind up with such a transformer, ignore the center tap. Or not. you can drive the transformer with one end and the center tap, and ignore the other end. The impedance in this case will be 150 ohms. I agree with Y; if your only options are 8 and 50, go with 8.

ak

 

Ah, CT = Centre Tap, got it.

I've gone for a pair these Hammond "classic" 600->8/4 ohm audio transformers: http://www.farnell.com/datasheets/2239256.pdf. If they don't work in this application then I can probably use them in some other (non-Chinese) project.

 

Note that typically, the primary impedance of a tube circuit output transformer is much higher than 600 ohms. But 600 is way better than 8, so it will be interesting to see if this works well.

ak

 

Note that typically, the primary impedance of a tube circuit output transformer is much higher than 600 ohms. But 600 is way better than 8, so it will be interesting to see if this works well.

ak

Yes, I suspect that output is at about 5K, but he said his resistive 600Ω load wasn’t overdriving, so...

Yes, it will be interesting.

The OTL design of that amp is a little concerning, even if you did use high impedance phones on it.

 

I don't know much about circuit design to say for sure, but I'm hopefully moving in the right direction. Although it's an expensive way to spend £60 - fixing a £20 problem ;-( If it doesn't work, I'll be back, offering to swap two expensive audio output transformers for a £3 OPA134.

 

The OTL design of that amp is a little concerning, even if you did use high impedance phones on it.

Seen it before. Before Ray Dolby created Dolby, he was a circuit designer at AMPEX, and was a critical member of the design team for the first broadcast quality video recorder. He had a similar true-push-pull circuit as the output stage in his video processor. All vacuum tubes (early 1950's), ran on +/-250 V (!), DC coupled throughout; the output front/back porch was at GND and sat there like a stone. It was a stunner.

ak

 

Seen it before. Before Ray Dolby created Dolby, he was a circuit designer at AMPEX, and was a critical member of the design team for the first broadcast quality video recorder. He had a similar true-push-pull circuit as the output stage in his video processor. All vacuum tubes, ran on +/-250 V (!), DC coupled throughout. It was a stunner.

ak

Literally!

 

Seen it before. Before Ray Dolby created Dolby, he was a circuit designer at AMPEX, and was a critical member of the design team for the first broadcast quality video recorder. He had a similar true-push-pull circuit as the output stage in his video processor. All vacuum tubes, ran on +/-250 V (!), DC coupled throughout. It was a stunner.

ak

The Futerman OTL power amp uses video tubes to directly drive low impedance loads. It's a great idea from a fidelity perspective but it eats finals, as is no surprise. At least it is not wired to something on your head, though.

OTL3.pdf

 

Note that typically, the primary impedance of a tube circuit output transformer is much higher than 600 ohms. But 600 is way better than 8, so it will be interesting to see if this works well.

The transformers arrived today, and I hacked up a quick test set up. It sounds like the obvious distortion has gone, but there is a lack in volume such that it isn't much of an amplifier, more like just a buffer stage. And possibly one that has an odd sound quality in its overall frequency response. I need to do some more comparisons to see if the distortion/frequency response without the transformers was there at the same listening level.

 

The transformers arrived today, and I hacked up a quick test set up. It sounds like the obvious distortion has gone, but there is a lack in volume such that it isn't much of an amplifier, more like just a buffer stage. And possibly one that has an odd sound quality in its overall frequency response. I need to do some more comparisons to see if the distortion/frequency response without the transformers was there at the same listening level.

Not very promising...

 

It's been a while since I've been able to get back to this project, but some good news. My first quick 'n' dirty test was flawed in some way. When properly hooked up, and using a decent source, those Hammond transformers work really well on the 4ohm tap. At first I thought it was a little light in the bass, but that was my old Nakamichi 'phones. Switching to the Sennheiser M2s, that the amp is destined for, sorted that out.

Although way over budget for a "simple" headphone amp, it was an interesting project. Now I've just got to build a case for it. My original design was to repurpose the type of box that fancy bottles of port get shipped in. With an extra pair of transformers to take account of, as the Jaws quote goes, I'm going to need a bigger box.

 

Glad to hear that it worked out. At least you have a thrermionic device that you will use regularly, putting you in a tiny minority of the human population.

 

I've been in that minority for nearly thirty years! I bought my VTL300 monoblocs (16x 6550s between the pair) in 1990.