drawback!
Hi, I'm always me.
Just to point out that the circuit sketched in the previuos post (http://www.diyaudio.com/forums/attachment.php?s=&postid=635800&stamp=1115325339 has a big drawback: the DC current shares in the two branches, so the class A maximum swing of the tube is reduced, because it works at (more or less) half the current it can works in a conventional SE topology.
Still a confirmation of the 'no free lunch' concept.
Hi, I'm always me.
Just to point out that the circuit sketched in the previuos post (http://www.diyaudio.com/forums/attachment.php?s=&postid=635800&stamp=1115325339 has a big drawback: the DC current shares in the two branches, so the class A maximum swing of the tube is reduced, because it works at (more or less) half the current it can works in a conventional SE topology.
Still a confirmation of the 'no free lunch' concept.
I have to correct myself again. Finally running out the simulation on the only working embodiment of the three proposed by Farber, I discovered that the attacchet circuit has no Idc/2 limitation on class A dinamic feature, because the other half of current that is needed by the primary winding is provided by the L inductor (that being that try to keep constant the current flowing in it, so if the valve does't sink int, it forces current into the primary giving Idc full swing).
So power is the same of a conventional SE or parafeed stage at the same DC current thru THE TUBE.
Performance seems good, especially with just a 800+800 turn winded on 5cmq core and a 25H/40mA inductor that should be quite small. I don't know if trust these simulation, but a simple check I made, decreasing the turn shows distortion going effectively higher.
I think that the model for nonlinear transformer does not take into account the real GAP one have when assembling lamination. In this case the simulation should be corrected with toroidal transformer.
The output tranformer plus inductor can also be comprised of a two toroidal cores, one gapped and the other not. Over the non-gapped toroid will be wind the primary plus secondary.
The gapped core of identical dimesion is stacked on this first trafo and the the aux (compensating or tertiary) coil is wind on BOTH cores. So we can obtain with just one coil compensating coils and the inductor, saving labor.
The low number of turns required will give better high frequency response and low loss.
So power is the same of a conventional SE or parafeed stage at the same DC current thru THE TUBE.
Performance seems good, especially with just a 800+800 turn winded on 5cmq core and a 25H/40mA inductor that should be quite small. I don't know if trust these simulation, but a simple check I made, decreasing the turn shows distortion going effectively higher.
I think that the model for nonlinear transformer does not take into account the real GAP one have when assembling lamination. In this case the simulation should be corrected with toroidal transformer.
The output tranformer plus inductor can also be comprised of a two toroidal cores, one gapped and the other not. Over the non-gapped toroid will be wind the primary plus secondary.
The gapped core of identical dimesion is stacked on this first trafo and the the aux (compensating or tertiary) coil is wind on BOTH cores. So we can obtain with just one coil compensating coils and the inductor, saving labor.
The low number of turns required will give better high frequency response and low loss.
Attachments
I found this thread when searching for discussion of using PP OPT in SE mode. The reason is that I had been musing over a simple way to use a PP OPT in a SE amp. My idea probably fails to take everything into consideration but here goes anyway.
My thinking was that the bias current in the second output tube of a PP amp provides the cancellation of magnetic saturation of the core. But why would it have to be a tube doing that. Why not just replace the second tube with an appropriate power resistor to draw the needed DC current?. Not efficient, but cheap.
I suppose that the hitch would be interaction between the two halves of the primary caused by the alternating component of the magnetic field generated by the AC signal. So is this a stupid (unworkable) idea?
mike
My thinking was that the bias current in the second output tube of a PP amp provides the cancellation of magnetic saturation of the core. But why would it have to be a tube doing that. Why not just replace the second tube with an appropriate power resistor to draw the needed DC current?. Not efficient, but cheap.
I suppose that the hitch would be interaction between the two halves of the primary caused by the alternating component of the magnetic field generated by the AC signal. So is this a stupid (unworkable) idea?
mike
Hello Mike,
The resistor idea will work, but it will really waste power big-time.
The resistor not only wastes DC power but is a load on the audio signal too.(12 % max eff. and half power output lost) Better to just use the second P-P pentode as a current source by removing its AC drive. Then it only wastes DC power. (25% max eff.) One can save the DC power as well if you replace the current source with an inductor, but is more expensive.(50% max eff.)
Another alternative is to use assymetric drive on the second P-P pentode (actually its best to replace it with a HV Mosfet or Bipolar trans.) to cancel the DC and double AC power output, while still preserving the SET sound. (50% max eff. 2X or 4X audio power output) Have a look here:
http://www.diyaudio.com/forums/showthread.php?threadid=56916&highlight=
Don
The resistor idea will work, but it will really waste power big-time.
The resistor not only wastes DC power but is a load on the audio signal too.(12 % max eff. and half power output lost) Better to just use the second P-P pentode as a current source by removing its AC drive. Then it only wastes DC power. (25% max eff.) One can save the DC power as well if you replace the current source with an inductor, but is more expensive.(50% max eff.)
Another alternative is to use assymetric drive on the second P-P pentode (actually its best to replace it with a HV Mosfet or Bipolar trans.) to cancel the DC and double AC power output, while still preserving the SET sound. (50% max eff. 2X or 4X audio power output) Have a look here:
http://www.diyaudio.com/forums/showthread.php?threadid=56916&highlight=
Don
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.