Hi Ian, have tried base resistors. It depends very much on Cob whether you get much control that way. With the low Cob types needed here, HF control is not very effective until higher values are used, which risks spoiling performance at 1.2A Ic.
I found the best approach was to use transistors with similar HF rolloff, or the same power transistor in both positions. A slow power transistor coupled to an HF-capable driver transistor is certain to lead to a 'tail-chase' as the slow transistor tries to keep up with rapidly changing command!
I found the best approach was to use transistors with similar HF rolloff, or the same power transistor in both positions. A slow power transistor coupled to an HF-capable driver transistor is certain to lead to a 'tail-chase' as the slow transistor tries to keep up with rapidly changing command!
Hi Rod, you call the positive leg of the circuit a Gyrator but it looks like a capacitor multiplier. Are they same? Thanks.
Mark II version
I have been working on a Mark II version of this circuit, mostly because Kurt (Zigzagflux) showed me that it measured poorly if the MJ150xx transistors (that I suggested) were used in place of the Japanese originals.
The 2SC5511 is now obsolete.
While looking through the alternatives (which must have unusually high Hfe at only 2 or 3V Vce, and LOW Cob) I have looked again at the schematic and found ways to reduce the output ripple voltage.
I'll post a schematic if it achieves the aims, and looks to have the same stability that the Mark I has - no notable drift in six years working life.
It's remarkable that I have had requests for PCBs to support this circuit all this time, and I feel that if the new circuit is a success, I'll finally do a batch of PCB.
And since the transistors are not easy to find, I'll buy a stock of those too, and supply it as a set (PCB + transistors) or assembled/tested. (any preferences?)
I assume that most builders will use them for 300B/2A3 power amps or preamp duty (26, 10Y, LS8, LP2 &c &c) with current of 1 to 2.5A
Is that a good idea? Post here or PM if you are interested in some hardware.
I have been working on a Mark II version of this circuit, mostly because Kurt (Zigzagflux) showed me that it measured poorly if the MJ150xx transistors (that I suggested) were used in place of the Japanese originals.
The 2SC5511 is now obsolete.
While looking through the alternatives (which must have unusually high Hfe at only 2 or 3V Vce, and LOW Cob) I have looked again at the schematic and found ways to reduce the output ripple voltage.
I'll post a schematic if it achieves the aims, and looks to have the same stability that the Mark I has - no notable drift in six years working life.
It's remarkable that I have had requests for PCBs to support this circuit all this time, and I feel that if the new circuit is a success, I'll finally do a batch of PCB.
And since the transistors are not easy to find, I'll buy a stock of those too, and supply it as a set (PCB + transistors) or assembled/tested. (any preferences?)
I assume that most builders will use them for 300B/2A3 power amps or preamp duty (26, 10Y, LS8, LP2 &c &c) with current of 1 to 2.5A
Is that a good idea? Post here or PM if you are interested in some hardware.
I would consider buying a stash if you can prove good performance with the Japanese xistors. Still would surprise me to see good HF performance and high gain at low Vce, but my experience with BJT's has been limited - just pump the base current to get them to perform.
For the time being, I've resorted to the simple LT1086 CCS, which works nicely, all things being equal. But I suspect your upgraded boards would fit in my existing space, so they could potentially be retrofitted with no changes.
For the time being, I've resorted to the simple LT1086 CCS, which works nicely, all things being equal. But I suspect your upgraded boards would fit in my existing space, so they could potentially be retrofitted with no changes.
Rod,
I would like nothing better then to be able to turn my brain off as regards DHT filament power supplies! Additionally, I am very impressed with your elegant solution.
Please count me in for 4-6 sets (PCBs & xistors).
I would like nothing better then to be able to turn my brain off as regards DHT filament power supplies! Additionally, I am very impressed with your elegant solution.
Please count me in for 4-6 sets (PCBs & xistors).
New design, with the aim of reducing ripple voltage. I've ordered components to build them, Japanese transistors with good performance at Vce = 2 to 3V, and I'll publish it when I have some measurements.
3A needs a slightly bigger transistor than the 1A version, but otherwise it should build the same. I've ordered some different parts for high current, and will give it a try. I believe it's worth looking at, since the Russian ГМ70 (GM-70) runs 20V 3A.
.
