[mulveling]

I'm kind of morbidly curious about what happened to the KGDT -- seems like it fell right off the map. It appears to have very, very few builds out there, and as far as I can recall nobody has commented on the sound. It looked to be the most rolling-friendly of all the tube electrostatic amps (well, except for the ES-1, though that amp would also make its owner's life a living hell). Is it just that the KGST was better?

 

[Jones Bob]

A recent KGDT for sale sighting with photos...

http://www.head-fi.org/t/822311/kgdt-kevin-gilmore-and-mjolnir-audio-balanced-electrostatic-headphone-amplifier-sr-007

 

[JimL11]

 
Am I reading this correctly i.e.  a higher voltage rail will not affect headroom?

As Jones Bob has stated, even 100 VRMS produces a very loud level (100 dB) in Stax headphones - equivalent to listening to an air hammer at 1 yard.  Most people won't listen at levels near this, and those who do will damage their hearing.  Basically, nobody will, or should, listen at levels near the output capacity of the amp.  So, if you can jump 3 feet in the air, it doesn't matter whether you are in a house with a 10 foot ceiling or a gym with a 50 foot ceiling - your head won't get close to either.  

 

[JimL11]

  I'm kind of morbidly curious about what happened to the KGDT -- seems like it fell right off the map. It appears to have very, very few builds out there, and as far as I can recall nobody has commented on the sound. It looked to be the most rolling-friendly of all the tube electrostatic amps (well, except for the ES-1, though that amp would also make its owner's life a living hell). Is it just that the KGST was better?

Actually, the ES-S is NOT a very good candidate for tube rolling.  I've done a simulation of the basic SRX circuit and  the only suitable tubes for the input cascode amp section are the 12AT7 (best), and 12AX7, 5751 and 6SL7, which are not quite as good - you lose a few dB of gain compared to the 12AT7, and the circuit has little enough gain as designed.  The EL34 is OK for the output tube but its gain is marginal, so the 6S4A and 6SN7GTA/B are the best.  Nothing else will allow the circuit to work correctly, i.e. with flat frequency response and reasonable distortion.

 

[Rossliew]

As Jones Bob has stated, even 100 VRMS produces a very loud level (100 dB) in Stax headphones - equivalent to listening to an air hammer at 1 yard.  Most people won't listen at levels near this, and those who do will damage their hearing.  Basically, nobody will, or should, listen at levels near the output capacity of the amp.  So, if you can jump 3 feet in the air, it doesn't matter whether you are in a house with a 10 foot ceiling or a gym with a 50 foot ceiling - your head won't get close to either.  

Got it

 

[Rossliew]

I'm kind of morbidly curious about what happened to the KGDT -- seems like it fell right off the map. It appears to have very, very few builds out there, and as far as I can recall nobody has commented on the sound. It looked to be the most rolling-friendly of all the tube electrostatic amps (well, except for the ES-1, though that amp would also make its owner's life a living hell). Is it just that the KGST was better?

Was wondering the same thing when I saw the for sale post of the kgdt. Tempting at those prices which includes the 007 sz1 versions...yay or nay?

 

[sarora]

I am the one selling my KGDT because I am moving and don't have the space, but it is a great amplifier with the quality you expect from the KG & Mjolnir-Audio designs. It is very close to the KGST in terms of sound. Here's some information right from Birgir @ Mjolnir-Audio:
 

The Octave is the warmest of the bunch.  Traditional tube amp in every sense of the word but with some clever stuff thrown in the mix.  The KGDT and KGST are similar as they are based on the same design.  The DT is a bit warmer due to lower voltage rails and the tubes used.  The ST has this slight warmth to it which suits the 009's to a T.  Helps with some of the upper midrange nastiness they are prone to bring. 

 

Here is a new amp design from us, the KGDT.  As the name might suggest this is a sister amp to the KGST.  The goal here was to allow to tune the sound more by tube rolling.  On the KGST you can just use 6S4A’s and only them, on the KGDT you can use many of the larger 9 pin dual triodes.  The amp was designed around the ECC99/12BH7A and the 6N6pi/6CG7 but others will work assuming they have similar specs and the same pinout.  The circuit board has a simple toggle switch to select single 6.3V heater or a center tapped 12.6V setup.  Now some slight compromises had to be made to accommodate  these tubes so the rail voltages are a bit lower than on the KGST and there is a bit lower current running through the output stage.  This is done so that the tubes can’t be run way too hard and thus not last long.
The sound of this amp is very similar to the KGST and so is the performance.  Great control of even tricky phones like the SR-007 Mk1 plus the ability to alter the tone of the amp with different tubes.  117V or 230V are no problem as the amp has a switch inside like all my new amps.

 

[wuwhere]

  I am the one selling my KGDT because I am moving and don't have the space, but it is a great amplifier with the quality you expect from the KG & Mjolnir-Audio designs. It is very close to the KGST in terms of sound. Here's some information right from Birgir @ Mjolnir-Audio:

 
If I had seen that a week ago I would have bought it. That's a great deal.

 

[JimL11]

I have not heard the KGDT but its design is very close to the KGST, which I have heard, and is a very good sounding amp.  The KGST was designed as the ultimate expression of the overall circuit topology of the SRM-007 but with a better choice of tubes and a constant current loaded output stage.  Similarly, the KGDT circuit topology it is very close to the SRM-T1 but the use of constant current loads in the output stage plus the more powerful ECC99 output tube make it much, much better in terms of its ability to drive any Stax headphone.  The price seems more than reasonable IMHO.

 

[sarora]

   
If I had seen that a week ago I would have bought it. That's a great deal.

 
Yeah, I used the ECC99 Gold Pin and they are pushed almost to the maximum (safely of course) at +300V with -300V bias, so 600V peak-to-peak which is quite a lot from the tubes, but the amplifier is quiet and there is absolutely no noise floor. And unlike some of the Stax amplifiers, that voltage swing is actually fully usable

 

[wuwhere]

I just bought an 007 and an SRM-717 last Saturday. But I'm thinking about your 007 though.

 

[sarora]

Back when I purchased the amp, I did ask Birgir what the swing voltage was via. e-mail and this was his reply - 
 

  The KDT swings about 370Vrms so about par with the Stax amps.  The difference here though is how much of that is actually usable.  The constant current sources make it usable all the way which is not the case with the Stax amps. 
 
They are pretty much the same in terms of load.  The Mk2 maybe a smidge easier due to the more present top end but the change is tiny.

(I was asking about his opinion on the 007MK1 vs MK2 with the KGDT)
 
 
as well as the B+ voltage of 300V and a bias of -300V for 600Vpp.

 

[JimL11]

   
Yeah, I used the ECC99 Gold Pin and they are pushed almost to the maximum (safely of course) at +300V with -300V bias, so 600V peak-to-peak which is quite a lot from the tubes, but the amplifier is quiet and there is absolutely no noise floor. And unlike some of the Stax amplifiers, that voltage swing is actually fully usable

So the real issue with the Stax tube amps is the output stage design, which uses load resistors.  For example, with the SRM-T1, which has an output stage standing current of 9.7 mA/channel, and a specified maximum output voltage of 300 volts RMS, the load resistors and feedback resistors suck up about 9.2 mA peak, leaving only 0.5 mA for the headphones.   Now, this is a very high level, roughly 110 dB SPL at the ears, but as you may imagine, with this little bit of current available the sound is going to be a tad strained.  The use of constant current sources in the KGDT (and the other KG amps) considerably decreases the current sucked up by the amp, leaving a much greater amount of current available to the headphones.

 

[JimL11]

   
Yeah, I used the ECC99 Gold Pin and they are pushed almost to the maximum (safely of course) at +300V with -300V bias, so 600V peak-to-peak which is quite a lot from the tubes, but the amplifier is quiet and there is absolutely no noise floor. And unlike some of the Stax amplifiers, that voltage swing is actually fully usable

Just to make it clear, the ECC99 is specified for a maximum standing plate voltage of 400V (that is, a voltage it will tolerate indefinitely), so a design voltage of 300 volts (75%) on the plates isn't a big deal.  The T1 actually pushes significantly harder, as the 6CG76FQ7 tubes are specified for a max plate voltage of 330, and they are run at about 300 volts (90%).

 

[sarora]

 
Just to make it clear, the ECC99 is specified for a maximum standing plate voltage of 400V (that is, a voltage it will tolerate indefinitely), so a design voltage of 300 volts (75%) on the plates isn't a big deal.  The T1 actually pushes significantly harder, as the 6CG76FQ7 tubes are specified for a max plate voltage of 330, and they are run at about 300 volts (90%).

 
Yes, I asked Birgir about this specifically when looking at the datasheet (the 400V on the ECC99) and when designing the KGDT they decided that this was the best/maximum voltage (300V on the B+ with -300V bias for 600Vpp and ~370Vrms) to go with in order to keep the amplifier smoothly running without the tubes dying quickly or causing other issues with the topology that allowed it to leave a much larger amount of current for the headphones, especially the 007Mk1's which are notoriously hard to drive.