What kind of sound quality to expect from Andrew Radford's DIY electrostatic amp?

[altie]

Man, I should've known these DIY ES phones would be the never ending project. Going from the plans on Headwize my pops and I built a pair of ESL drivers and got to work on the simple receiver-driven step-up amplifier. That should be the easy part, right? Not so much.

I've tried two sets of 20v-70v transformers. Both are enough of a step up to generate quite a bit of sound out of the drivers, but with extreme distortion. There's lots of crackling, weird harmonics and such to the point where vocals and some instruments are unintelligible. Is this amp really only intended as a quick test for the drivers? Note, I haven't hooked up exactly the step up he indicated. The highest step up my transformers will support is 8ohm-1K. I wouldn't think the higher current on the output side would be the source of the distortion with the receiver turned low, but it's still there. Any ideas? If my drivers put out sound at all, is it worth it to build a real amp? Please keep in mind that I'll sink plenty of time into this, but money I'm more attached to.

 

[cetoole]

Assuming your sensitivity isnt too horrible, you should be able to use the receiver to directly drive the headphones, but they wont go very loudly. I have measured the voltage needs of my Lambda Pros and ESP/950, and while they do need a good bit of voltage, you can probably run around 60-70dB within the voltage swing most receivers are capable of. Looks like D-S spacing on this headphone is about the same as Pro bias Stax and the ESP/950, so I would assume sensitivity isnt too far off. I would try this, at least eliminate the transformers as the source of your problems. Just apply the bias separately.

 

[altie]

Wow. That response was about 100 times faster than expected, and helpful. The sound is very faint but I think it has the same distortion. I should also note that if I have the transformers wired up for high step up and crank it up, the transformers themselves will "sing", putting out the same distorted sound as the drivers.

 

[cetoole]

Actually, it was also a semi-dumb answer, since your receiver is almost certainly single ended, so yeah, you will get very distorted sound from it driving an electrostatic headphone. I think I have heard numbers in the 10-30% THD range for SE drive electrostats. I am a bit surprised it is so quiet, unless your coating isnt nearly resistive enough, D-S spacing is too large, or bias is too low. I only average about 10vrms measured from the differential outputs of my SRM-1/mk2 amplifier driving Lambda Pros. It is amazing though how the volume drops if I plug into the normal bias jack, I would guess it is a good 6dB difference.

What is your bias voltage, and have you any way to measure the capacitance of the drivers? Any pics? Also, can you try a different receiver driving the transformers?

 

[altie]

I'll bring them into work tomorrow and see if someone has a camera. I'm around a bunch of EEs all day so they all want to see the phones too

 

[spritzer]

If you are using a D/S gap of 0.5mm and a 600v bias then you only need 100v RMS to make them very loud. The ratio on your transformers is only 1:11 (about a third of what Stax used) so you aren't getting enough voltage swing.

As for them sounding distorted, there can be quite a few reasons for that. If the stators aren't perfectly flat then they will introduce a whole lot of non-linear effects. Now if you are running them SE then there will be a whole lot of distortion to deal with as there is no suspension of any kind to move the diaphragm back to a neutral position.

 

[ericj]

If these are transformer coupled, if the bias is referenced to the center tap on the transformers, i don't see how they would be single-ended - even if the amplifier driving them is.

 

[spritzer]

If there is a CT then it wouldn't be an issue so the problem could be elsewhere.

 

[cetoole]

Quote:

Originally Posted by ericj /img/forum/go_quote.gif If these are transformer coupled, if the bias is referenced to the center tap on the transformers, i don't see how they would be single-ended - even if the amplifier driving them is.

No, that was just my stupid suggestion to try direct driving them from the receiver, which would have the needed voltage swing to at least get an impression of if the drivers are the issue or if it is the transformer setup, but forgot to think about the need for 'stats to be driven differentially.

You could have one side of the secondary grounded when transformer coupled, and that would give SE output.