so i found this obscure buffer while back https://www.diyaudio.com/forums/analog-line-level/26846-simple-discrete-unity-gain-buffer-post316511.html
And have been using it where ever i can with great success. In comparison i find it to sound even better than a jlh buffer and in fact its the best sounding low part count buffer i have tried.
Now id like to scale it up to drive headphones much like the way the dao buffer is used and would like to know how id do that.
Simply tweak the resistor values and use medium power bjts on heatsink? Hmm
And have been using it where ever i can with great success. In comparison i find it to sound even better than a jlh buffer and in fact its the best sounding low part count buffer i have tried.
Now id like to scale it up to drive headphones much like the way the dao buffer is used and would like to know how id do that.
Simply tweak the resistor values and use medium power bjts on heatsink? Hmm
Besides the things you mentioned, C3 has to be increased, of course, and maybe you will need two stages or a Darlington for T1. With one stage, the input impedance at the base of T1 won't get higher than (hfe_T1 + 1) Z_load. For example, with an hfe_T1 of 100 and a Z_load of 32 ohm, that's only 3232 ohm.
Those transistors have a maximum voltage of 25 V, so less than the difference between the positive and negative supplies. If you apply very large input signals, greater than 12 V peak, you might just drive them into collector-emitter breakdown.
Assuming that you keep the momentary vCE below 25 V somehow, looking at this datasheet https://www.onsemi.com/pub/Collateral/MJE200-D.PDF , they can handle 200 mA at any VCE up to 25 V without running into second breakdown, see figure 5. At a VCE of 13 V, roughly the VCE in the circuit when there is no signal, they can handle 700 mA without second breakdown.
So all in all, a 200 mA bias current should be safe for second breakdown. The branch T2-T1 may carry larger currents during positive peaks, but then the voltage across T2 is below 13 V, so then it can handle much more current. The voltage across T3 increases during positive peaks while the current through it remains constant, so T3 will be the second breakdown bottleneck.
Regarding cooling, at a 200 mA bias current, the average dissipation in T2 and T3 will be about 2.6 W. They can handle that up to a case temperature of 128 degrees C (see figure 1). T1 and T4 will dissipate much less because there is not much voltage across them.
The one thing I'm not sure about is what happens during the current peaks you might get when the headphone connector temporarily gets shorted during plugging or unplugging the headphone. Although the average current is limited by the current source, there is no circuit limiting the momentary current through T1 and T2. This will only be a potential problem when plugging or unplugging during positive signal peaks, so T2 then has a collector-emitter voltage below 13 V and can then handle 700 mA or even 1.6 A during 5 ms. Maybe a 10 ohm power resistor in the collector lead of T2 could work as a crude current limiter.
With an hFE of at least 70 and assuming hfe to be equal to hFE, the input impedance at the base of T1 will be 2272 ohm at 32 ohm load.
Assuming that you keep the momentary vCE below 25 V somehow, looking at this datasheet https://www.onsemi.com/pub/Collateral/MJE200-D.PDF , they can handle 200 mA at any VCE up to 25 V without running into second breakdown, see figure 5. At a VCE of 13 V, roughly the VCE in the circuit when there is no signal, they can handle 700 mA without second breakdown.
So all in all, a 200 mA bias current should be safe for second breakdown. The branch T2-T1 may carry larger currents during positive peaks, but then the voltage across T2 is below 13 V, so then it can handle much more current. The voltage across T3 increases during positive peaks while the current through it remains constant, so T3 will be the second breakdown bottleneck.
Regarding cooling, at a 200 mA bias current, the average dissipation in T2 and T3 will be about 2.6 W. They can handle that up to a case temperature of 128 degrees C (see figure 1). T1 and T4 will dissipate much less because there is not much voltage across them.
The one thing I'm not sure about is what happens during the current peaks you might get when the headphone connector temporarily gets shorted during plugging or unplugging the headphone. Although the average current is limited by the current source, there is no circuit limiting the momentary current through T1 and T2. This will only be a potential problem when plugging or unplugging during positive signal peaks, so T2 then has a collector-emitter voltage below 13 V and can then handle 700 mA or even 1.6 A during 5 ms. Maybe a 10 ohm power resistor in the collector lead of T2 could work as a crude current limiter.
With an hFE of at least 70 and assuming hfe to be equal to hFE, the input impedance at the base of T1 will be 2272 ohm at 32 ohm load.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.