SR45 amplifier
- Thread starter debk
- Start date
Resurecting this discussion to run some ideas past the group.
I'm considering using my new PT2s to build a whole second SR-45 amp, but with a few changes...
First, I'm considering building both amps in one chassis -- like John did, and going with a slightly larger plate -- maybe something like 12" x 18", all film cap psu, and also looking into the possibility of using separate filament transformers and the FC1 to supply the 45 filaments. Aside from the obvious reason, I'm also thinking this will take a bit of the load off the PT2 so I can run a 6dn7 slightly more comfortably.
I have a pair of Dowdys on order with Mike, and those come with endbells so I'll most likely mount them above the TFA2004 jrs, which should free up a slot for placing the fc1 in a magnetic null formerly occupied by the plate choke. Also, the orientation will be rotated so that the opt and plate choke are at the rear of each half of the amp, the 6dn7s near the back and the 45s up front and keeping the pt more or less in the center of each side of the amp.
So, I guess what I'm asking is if my thinking on the lightnening of the pt's load for the 6dn7 makes any sense, and two if a separate filament tx and the fc1 have any issues associated with them -- so sort of like a hybrid sr-45 and paramount, and using the paramount driver/shunt reg board with the 6dn7.
This will allow me to leave the current sr-45s pretty much as-is and only steal the tfa-2004s when the time comes.
Thoughts?
-- Jim
I'm considering using my new PT2s to build a whole second SR-45 amp, but with a few changes...
First, I'm considering building both amps in one chassis -- like John did, and going with a slightly larger plate -- maybe something like 12" x 18", all film cap psu, and also looking into the possibility of using separate filament transformers and the FC1 to supply the 45 filaments. Aside from the obvious reason, I'm also thinking this will take a bit of the load off the PT2 so I can run a 6dn7 slightly more comfortably.
I have a pair of Dowdys on order with Mike, and those come with endbells so I'll most likely mount them above the TFA2004 jrs, which should free up a slot for placing the fc1 in a magnetic null formerly occupied by the plate choke. Also, the orientation will be rotated so that the opt and plate choke are at the rear of each half of the amp, the 6dn7s near the back and the 45s up front and keeping the pt more or less in the center of each side of the amp.
So, I guess what I'm asking is if my thinking on the lightnening of the pt's load for the 6dn7 makes any sense, and two if a separate filament tx and the fc1 have any issues associated with them -- so sort of like a hybrid sr-45 and paramount, and using the paramount driver/shunt reg board with the 6dn7.
This will allow me to leave the current sr-45s pretty much as-is and only steal the tfa-2004s when the time comes.
Thoughts?
-- Jim
Paul Joppa
Moderator
One of the unique features of the PT-2 is that the two halves of the 6.3v winding surround the 2.5v winding, shielding it from line noise and rectifier spikes of the high voltage and primary windings. If you use another source of filament power, you might want to try it both ways to see if you hear a difference - I never did that experiment, just designed in the shielding from the first.
The PT-2 runs at 10-11kGauss, much lower than the usual transformers (16kG; 13.5 for 50Hz rated but operated at 60Hz). This reduces vibration as well as radiated magnetic field. Consider getting transformers with dual primaries so you can wire for 240v, run at 120 - if you want to preserve this advantage. Be aware that split-bobbin transformers, which are very common these days, do have large external electrical fields unless you have a grounded shell or end bells. The split bobbin does reduce capacitive coupling between primary and secondary, providing a similar benefit to a shielded winding at a lower cost.
The PT-2 will run a bit cooler with a lighter load, about 10% less than a standard Paramour II.
The PT-2 runs at 10-11kGauss, much lower than the usual transformers (16kG; 13.5 for 50Hz rated but operated at 60Hz). This reduces vibration as well as radiated magnetic field. Consider getting transformers with dual primaries so you can wire for 240v, run at 120 - if you want to preserve this advantage. Be aware that split-bobbin transformers, which are very common these days, do have large external electrical fields unless you have a grounded shell or end bells. The split bobbin does reduce capacitive coupling between primary and secondary, providing a similar benefit to a shielded winding at a lower cost.
The PT-2 will run a bit cooler with a lighter load, about 10% less than a standard Paramour II.
Hi Paul,
thanks much. I'm not sure I completely follow the part about 240 v primaries, but that's probably due to my lack of transformer winding schedule knowledge.
I also wondered if I may need to load the 2.5v secondary some to keep voltages from going too high -- or maybe that doesn't matter so much with the shunt regulation -- except for the 6.3v filament, but that can be adjusted other ways.
Still, most of this will depend on what I experience once the sr-45s are in the system -- which of course means the system has to get put together first. I'm working on it...
-- Jim
thanks much. I'm not sure I completely follow the part about 240 v primaries, but that's probably due to my lack of transformer winding schedule knowledge.
I also wondered if I may need to load the 2.5v secondary some to keep voltages from going too high -- or maybe that doesn't matter so much with the shunt regulation -- except for the 6.3v filament, but that can be adjusted other ways.
Still, most of this will depend on what I experience once the sr-45s are in the system -- which of course means the system has to get put together first. I'm working on it...
-- Jim
Question for PJ -- are any of the new shunt reg/c4s borads going to be useful in the SR-45? A related question would also be if there is any progress on a new driver board for the sr-45s, or would y9ou recommend that people use the paramount soft-start board?
Getting ready to start planning the new layout for the stretch model sr-45 -- 11-1/2 deep instead of 10" but still 8 wide, portrait orientation, and monoblocks is what I've settled on -- just works much better with the rest of my system to keep them as monoblocks.
Dowdys will go on top over the OPT, grid chokes will go where the plate chokes used to be, and the space between the driver and 45 will probably be mostly eaten up with a 27-33 uF PF cap (the downside of that huge 125H Dowdy.
I'm also going to go with AC filaments and most likely 6dn7 driver/shunt reg tubes.
Thanks,
Jim
Getting ready to start planning the new layout for the stretch model sr-45 -- 11-1/2 deep instead of 10" but still 8 wide, portrait orientation, and monoblocks is what I've settled on -- just works much better with the rest of my system to keep them as monoblocks.
Dowdys will go on top over the OPT, grid chokes will go where the plate chokes used to be, and the space between the driver and 45 will probably be mostly eaten up with a 27-33 uF PF cap (the downside of that huge 125H Dowdy.
I'm also going to go with AC filaments and most likely 6dn7 driver/shunt reg tubes.
Thanks,
Jim
Paul Joppa
Moderator
Jim-
Sorry I missed your Sept 4 post; the forum software is not perfect (!) To answer (sort of) - because all transformer secondary voltages depend on the primary voltage and to a lesser extent on the other voltages and currents, there is no answer beyond "try it and see what you get" for filament voltages. Resistive losses in the wire from transformer to 45 filament are not going to be zero, as well! Best bet is to make a "true rms" measurement if you can, or can borrow a suitable meter; line voltage is far from a pure sine wave and the rectified high voltage makes the other windings even worse. Plan on possibly slipping in a suitable resistor...
As for the C4S boards, the latest version (v.4.4) is still what's in the Paramount "soft-start" kit. It has an adjustable bias trimpot for the driver section. As I think I've said before, this board does not have a really heavy duty heat sink for the C4S feeding the shunt reg; if you line voltage is over 120v that can be a problem. It's one I still plan to address in a new design, but that has not progressed in the last few months - the BeePre has taken most of our time lately.
I hope this is helpful; I've kind of lost track of this thread and don't remember all the things going on....
Sorry I missed your Sept 4 post; the forum software is not perfect (!) To answer (sort of) - because all transformer secondary voltages depend on the primary voltage and to a lesser extent on the other voltages and currents, there is no answer beyond "try it and see what you get" for filament voltages. Resistive losses in the wire from transformer to 45 filament are not going to be zero, as well! Best bet is to make a "true rms" measurement if you can, or can borrow a suitable meter; line voltage is far from a pure sine wave and the rectified high voltage makes the other windings even worse. Plan on possibly slipping in a suitable resistor...
As for the C4S boards, the latest version (v.4.4) is still what's in the Paramount "soft-start" kit. It has an adjustable bias trimpot for the driver section. As I think I've said before, this board does not have a really heavy duty heat sink for the C4S feeding the shunt reg; if you line voltage is over 120v that can be a problem. It's one I still plan to address in a new design, but that has not progressed in the last few months - the BeePre has taken most of our time lately.
I hope this is helpful; I've kind of lost track of this thread and don't remember all the things going on....
Hi Paul,
Yes, I too have lost track of all the details -- and yes you did say that the SS board in the paramount is not really up to the job -- I just forgot about that.
I haven't measured the line voltage in the new place yet, so will do that tomorrow.
Now I'll have to look at what the range of options are for that monster pf cap
Do let us know when you're ready to look into a new driver board though.
Thanks,
Jim
Yes, I too have lost track of all the details -- and yes you did say that the SS board in the paramount is not really up to the job -- I just forgot about that.
I haven't measured the line voltage in the new place yet, so will do that tomorrow.
Now I'll have to look at what the range of options are for that monster pf cap
Do let us know when you're ready to look into a new driver board though.
Thanks,
Jim
Paul Joppa
Moderator
I will have a new design - it may incorporate some of the Crack Speedball and/or BeePre boards, not decided yet. The old board has some issues I am not at all happy with.
It will be based on the PT-7 and include DC filament power. The circuit is pretty close to done, but the parts and boards are not selected yet. Right now the BeePre is still taking up a lot of time, getting all the parts in-house and tweaking the layout for the manual.
It will be based on the PT-7 and include DC filament power. The circuit is pretty close to done, but the parts and boards are not selected yet. Right now the BeePre is still taking up a lot of time, getting all the parts in-house and tweaking the layout for the manual.
Paul,
So, will any of these new boards work in the old sr-45? I already know the answer regarding the soft-start board, but interested in backwards compatibility to the existing design. In the case that that won't work, can I transplant all the nice iron (nickel tfa-2004 jr, dowdys, and bpc-16ni to the new design? I'm assuming you're designing for the s.e.x./stereomour output transformers and of course I'm wondering if there is enouugh wiggle room in the operating points to work with the 3k of the TFA-2004.
Best for me, of course would be to be able to use one of the new driver boards in the old existing design.
-- Jim
So, will any of these new boards work in the old sr-45? I already know the answer regarding the soft-start board
, but interested in backwards compatibility to the existing design. In the case that that won't work, can I transplant all the nice iron (nickel tfa-2004 jr, dowdys, and bpc-16ni to the new design? I'm assuming you're designing for the s.e.x./stereomour output transformers and of course I'm wondering if there is enouugh wiggle room in the operating points to work with the 3k of the TFA-2004.Best for me, of course would be to be able to use one of the new driver boards in the old existing design.
-- Jim
Paul Joppa
Moderator
Jim-
I had to spend several hours on ferries today, and used the time to look at alternatives.
Right now, what appeals to me the most is a completely fresh design. It always bothered me that we made so many compromises to get away with existing parts and layouts, in what was (IMHO) our best-sounding amp - the topology was way ahead of the rest of the design (AC heater, no time delay, etc.) I kind of hate to do this, I always liked the idea of growth paths, but the original WAS a growth path, compromised to work with the original Paramour II, and it just didn't sell enough to make economic sense.
The original was an experiment; only a few were made and the manual was nearly non-existent. I learned quite a bit from that experiment; among other things I learned that this is never going to be a big seller. Even if I come up with a prototype that sounds great, I have doubts that Doc B will want to develop a full-scale manual in the traditional format - that's a LOT of work and has to be spread over many units to make sense. It may only makes economic sense as an already-built amp.
So I am thinking of a much larger circuit board to house the power supplies - including regulated DC on the filament and heater, as well as the shunt-regulated high-voltage power. The manual for PC boards should be simpler to do. I'd like to do it with all film caps (except for the heater power, there are no polypropylene caps at 10,000uF!), and include time delay relays for high voltage and for shorting the OPT to prevent core magnetization. And impedance switches and an input level adjust for channel balance - the whole nine yards.
The MQ iron would still be a reasonable choice for the operating point, and the chassis plate may or may not be the same - that partly depends on whether or when we update the Paramount chassis. Of course, the Dowdy choke won't fit on any of our chassis plates anyhow ...
Of course, this is not set in stone yet. If there's an outcry for something more Bottlehead-y and less high-end, I have a few ideas. I'm just thinking that only a relatively few people will be into it so deep as to want a spendy 2-watt amp.
I had to spend several hours on ferries today, and used the time to look at alternatives.
Right now, what appeals to me the most is a completely fresh design. It always bothered me that we made so many compromises to get away with existing parts and layouts, in what was (IMHO) our best-sounding amp - the topology was way ahead of the rest of the design (AC heater, no time delay, etc.) I kind of hate to do this, I always liked the idea of growth paths, but the original WAS a growth path, compromised to work with the original Paramour II, and it just didn't sell enough to make economic sense.
The original was an experiment; only a few were made and the manual was nearly non-existent. I learned quite a bit from that experiment; among other things I learned that this is never going to be a big seller. Even if I come up with a prototype that sounds great, I have doubts that Doc B will want to develop a full-scale manual in the traditional format - that's a LOT of work and has to be spread over many units to make sense. It may only makes economic sense as an already-built amp.
So I am thinking of a much larger circuit board to house the power supplies - including regulated DC on the filament and heater, as well as the shunt-regulated high-voltage power. The manual for PC boards should be simpler to do. I'd like to do it with all film caps (except for the heater power, there are no polypropylene caps at 10,000uF!), and include time delay relays for high voltage and for shorting the OPT to prevent core magnetization. And impedance switches and an input level adjust for channel balance - the whole nine yards.
The MQ iron would still be a reasonable choice for the operating point, and the chassis plate may or may not be the same - that partly depends on whether or when we update the Paramount chassis. Of course, the Dowdy choke won't fit on any of our chassis plates anyhow ...
Of course, this is not set in stone yet. If there's an outcry for something more Bottlehead-y and less high-end, I have a few ideas. I'm just thinking that only a relatively few people will be into it so deep as to want a spendy 2-watt amp.
Paul Joppa
Moderator
I'll use the SEX/Stereomour output iron at 4K primary, so the load impedance will be 2-4-8-16 ohms. That's been central to the plan all along. It has a generous magnetic headroom at this operating point, and most listeners - including myself - have been very happy with it. The Magnequest BH-5 at 3K is still suitable, as is the BH-2 plate choke, and I'll try to leave room and mounting holes for them.
Paul,
Thanks for the detailed reply and explanations. I think, given what I have and have invested in, it makes most sense for me to just continue on the path I'm on, perhaps design and build a more substantial heatsink for the existing driver arrangement/shunt reg board, and my larger chassis.
I will however possibly look into the opt shorting switch/relay, but realizing that would require an additional power supply that may or may not impact noise levels. Also, given my speakers' sensitivity and the existing reports on the amps noise performance, I'm not sure how much, in reality, I'd gain in sonics over a fully maxxed out version of the existing design -- vcaps, copper chassis plate, dowdy, foam FEP OCC wire, etc.
After all, it's still going to be 2 glorious watts!
-- Jim
Thanks for the detailed reply and explanations. I think, given what I have and have invested in, it makes most sense for me to just continue on the path I'm on, perhaps design and build a more substantial heatsink for the existing driver arrangement/shunt reg board, and my larger chassis.
I will however possibly look into the opt shorting switch/relay, but realizing that would require an additional power supply that may or may not impact noise levels. Also, given my speakers' sensitivity and the existing reports on the amps noise performance, I'm not sure how much, in reality, I'd gain in sonics over a fully maxxed out version of the existing design -- vcaps, copper chassis plate, dowdy, foam FEP OCC wire, etc.
After all, it's still going to be 2 glorious watts!
-- Jim
johnsonad
New member
Jim R. said:Xavier,
I want those cobalt tfa-2004 jrs!
-- Jim
I have first dibs Jim! Sorry my friend ;D
