Count me in with this Addams Family build:
An externally hosted image should be here but it was not working when we last tested it.
Reproduction is very fine, one can pick up a mouse dropping from the soundstage with eyes closed. It does all it promises
Great, I see what you are doing now. FYI, the 6SL7 model from Ice Amplifiers is pretty far off the datasheet, but not enough to make a big difference as far as the AC analysis is concerned...
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Let's see:
Amp with no wrong headed type output cathode resistor: approx. heat generated under operating conditions is anywhere between 27 and 40 Watts, assuming the use of wrong headed tube rectifier.
With wrong headed cathode resistor: approx. 30 to 43 watts
For both cases heat generated by magnetic component losses is ignored. So yes, we are generating up to 11% more heat.
I can see where you are headed and I agree. The only right headed amp is class D.
Sheldon
Count me in with this Addams Family build:
An externally hosted image should be here but it was not working when we last tested it.
Reproduction is very fine, one can pick up a mouse dropping from the soundstage with eyes closed. It does all it promises
Hadn't seen that one. Nice work.
Sheldon
Yes, well... there's another on its way but that's not final yet.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Reaping souls again in Holland, are you?
Looking forward for your 45 plans,
I've got beautiful AVVTs collecting dust.
Hi, it was a cut and paste from Darius I must admit
There are not that great many triodes suited for this topology.
Visit the inventors' pages and look for Der Triode Faktor.
Actually, the only reason to try to acurately measure THD is to investigate if the impressive character of this amp is at all due to 2nd hardmonic distortion.
I should move over to a digital ossillioscope anyway, so this is hopefully a step in the right direction.
If you do build it, you won't be dissapointed. The e83cc provides ample gain. There are not many choices... 12at7 might be worth investigating.
Last edited:
Actually, the only reason to try to acurately measure THD is to investigate if the impressive character of this amp is at all due to 2nd hardmonic distortion.
I should move over to a digital ossillioscope anyway, so this is hopefully a step in the right direction.
If you do build it, you won't be dissapointed. The e83cc provides ample gain. There are not many choices... 12at7 might be worth investigating.
Oscilloscope is a good thing, very useful indeed!
Soundwise the ECC83 is not my cup of tea, thin and nervous IMO. Too low the current also. I have tried 12AT7's & ECC81's = Worst sounding triode ever heard.
Member
Joined 2009
Paid Member
I can see where you are headed and I agree. The only right headed amp is class D.
If you choose to run an extra 12W of heat or so under your chasis along with the need for higher voltage capacitors, for no real gain, that's your choice.
I know you feel I'm attacking all your hard work on the amplifier you built - but if we look past this then perhaps my suggested approach would make the L-W Darius style topology more popular, less costly and potentially safer (lower voltage) for those that follow after us in this hobby ? I've already got the parts for my L-W so I'll be sticking with the 500V h.t.
I see your concern. No, there won't be any phase shift to worry about. There's no current flow to speak of being drawn by the grid of the CF. The coupling cap will have negligible ac voltage across it until there's grid current - and this isn't going to happen until the output tube is also hitting grid current and pinning the voltage of the cathode of the CF. There are caps already in the circuit anyhow, the one which feeds the hum to the cathode of the input tube and the one that feeds the bootstrap to the anode load of the input tube - all without issue. I think the hum cancellation will work very well.
Last edited:
I didn't say it wasn't a smart decision to do it this way, but just that it *is* positive feedback and that it *does* come at the expense of something. It is an excellent system optimization to make, no doubt.To put things into perspective, the linearity of the input stage is vastly improved, while any loss in linearity of the CF is totally negligibe.
The two real downsides of this arrangement are a little phase shift because of the feedback path thru the bootstrap cap, and reduced output impedance of the CF. With the loads presented here however, these are also not noticeable in this arrangement, so yes we kind of get a free lunch
At full output? I don't believe it. The 6SL7 doesn't have enough standing current to drive the negative swing of the CF's AC load at 45kHz. It will clip. Too much miller capacitance to drive and it also has to drive the bootstrapped resistor in the first stage along with the other one. The load line doesn't lie. It will run out of current at those frequencies
I mean, it is probably not such a big deal since nobody would want full amplitude 20kHz coming out of the thing anyway, but don't make things up about your amplifier to make it seem more impressive. It was obviously carefully optimized to make the most out of the 6SL7 but it is what it is.
Have you measured full-power bandwidth on an actual built amp?
I'm not talking about the BW of the CF. I'm talking about what frequency the CF will clip at with the 2A3 driven to full amplitude.
Maybe I've made a huge mistake somewhere but I'm not seeing it. The miller capacitance of the 2A3 represents an impedance of 100kOhm at 20kHz. At 45kHz it is less than 50kOhm and you still have to drive the resistances in the first stage plate load. If I've made a mistake, I'd love to know what it is. That is a really steep load line for something that is starting with a standing current of 1.2 mA.
ETA: I just noticed that the schematic in your simulation is different than the one I was talking about earlier (the one posted by Sheldon). Your simulation has much more standing current in the CF stage. It won't have the same problem.
Last edited:
Actually, for the design goldenbeer posted, the extra heat is about 7W. True, it's not nothing, but it's not hard to deal with - especially since there are no heat sensitive components in the design. Yes 500V is more than 300V. I would think that both require due caution. I don't think this design will ever become that popular. Because stages can't be developed independently, it limits tube choice to a greater degree. And to use a different tube set requires some iteration. The follower mitigates this issue to some degree, and a SS follower even a bit more - much more if you cap couple the input tube to the follower (as you suggested). That would also affect the way output currents circulate. So it's becoming a different design (not better or worse, necessarily).
Don't worry. My feelings aren't hurt. It's not even my design. I've built all SS amps (including one with output transformers, based on Susan's Zeus), amps with both SS and tubes (including a phono pre based on 3A5's), and only tubes (well I cheated a bit and used a hybrid bridge in the LW). They were all fun and added to my knowledge. And that's why I build them.
The LW was a new challenge, based on DC coupling, but adding some interesting ways of integrating the power supply, using a somewhat different distribution of current paths. And I was also intrigued with the LW version because I found it an elegant solution to problems of its time (which can certainly be solved by many other means today, and which I have not argued against). I introduced the LW concept here because it seemed apropos to the thread topic.
I have no problem with discussing the relative merits of doing things in various ways. But I do raise my hand when broad assertions like good or bad, better, worse, are offered without at least some level of quantitative support or guess. Without some testable basis for making such assertions the discussion can't lead to anything useful. All designs have compromises. These compromises affect the outcome in different ways to different degrees. Speaking of degrees, I may be fine with an extra 7W of heat, you may not. That doesn't mean one way is right, and one wrong. It depends on the design goal. But we can determine the actual amount in question.
I'm not just talking about measurements. I'm fine with builders stating their preference for sound. But it's usually hard to figure what that means without a lot of context (system, etc.), and without being able to hear ourselves. I liked my amp. Would others find it satisfying? Can't say. I can say that it has no glaring measurement faults compared to those I've seen of other single ended designs.
Sheldon
Member
Joined 2009
Paid Member
Oh I do have an oscilloscope, but it just isn't working well enough for any decent measurement.. Its ok for some troubleshooting.
I found in every instance of ecc83 "sounding bad" it isn't implemented well. When implemented well in the correct circuit it performs exceptionally well - even cheap examples are very very good. I can say similar things about the 12at7 and the 6SL7 which all too often are not implemented well.
The ecc81 is not a tube I can comment on, but the plate curves do not look terribly promising.
Member
Joined 2009
Paid Member
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.