I was just playing with building a solid state front end for the T2. I was having some issues simulating the batteries because the LT1021 was not behaving well. It wants about 1.5mA of current, but the pass transistor was taking all the current in the models so it would never get to 10V.

I wanted to see how this behaved in real life so I built two batteries that I'll use in my prototype amp. They will need to be set at 444V for my purposes.

Here's the scematic...

and here's the prototype...

After adjustment, they come up at 444V and stay there. Very solid. I've had them on the scope and didn't see any issues yet.

I'll still need to see if it behaves with the K216 on top and current sink on bottom.

Edited September 2, 201114 yr by Kerry

I think I've posted the full schematic before, but here it is again.

I'd love comments

Edited September 2, 201114 yr by Kerry

From T2 Thread...

I have no idea that use semis for first stage is good or bad for the sounds.

Article from MJ 1995/5 (Page 37)

---------------------------------------------

“The first stage is a cascaded differential circuit that uses two 7308 tubes that are equivalent to the 6DJ8. A transistor constant current source was added to both the plate load and the cathode. The supply voltages for this differential circuit are +250V and -600V.

The reason tubes are used for the first stage is to avoid thermal related parameters. If semiconductors were used in this circuit, the junction temperature would be influenced by the current. It would be easy to unbalance the differential circuit when the current changes. However by using tubes, the temperature is already biased by heaters so regardless of changes in current, the operating parameters will be very stable. These tubes are covered by metal shielding caps.”

I'm not sure of the sound either. I've used this front end on a dynamic amp with a T2 topology. It sounds really good

I'm building this in a way that I can try different front ends. I'm not trying to replicate the cascode with semi's. It's closer to a BH front end, but using the T2 active batteries to translate to the bottom rail instead of the current tunnel. From there back it is very similar to the T2.

The idea was to experiment and come up with a poor man's T2. I've dropped the voltage down a bit too so I can use a simple current source (IXCP10M90S) on the output stage. I can probably go a bit higher on the voltage, but want to hear it first.

I've almost got the front end finished, but I've still got more work to do on the power supplies (still need to build 30V, -425V and 160V supplies) before I can build the output stage.

Got the input section, +/- 15V and 30V supplies done

Here's what I've got so far. I'm using the 30V supply to bottom of the 425V supply to create a -455V line for the negative rail.

The input section came up perfectly. I've tied the output (drain) of the K216 to some 10K resistors against the -15V rail just to test. This would represent the top of the batteries. The DC is sitting at 9.3V. Seems about right.

Awesome work!

I'd be a bit concerned over the red irradiated pills at the bottom of your scope there.

In the case of a plate CCS, is evaluating the performance of the CCS pretty much the same as evaluating the performance of the amplifier? All I can think it to measure distortion on the plate, and since that where the output is taken... Also, I similarly assume that a oscillating CCS isn't really discernible from a oscillating amplifier for the same reason?

In the T2/Blue Hawaii amps, the current source is tied to the anode of the output tubes which flows pretty much right to your ears. Seems like the CSS can definitely impact the final result.

If I were to speculate, the CSS's job is simply to provide a constant current against the fluctuating voltages produced by the amplifier. If it does this one thing well, then you are mostly hearing the amplifier. If it does it poorly, then I'm less sure.

nm

Edited December 30, 201312 yr by luvdunhill