fdegrove said:
I dunno. Working with temperamental creatures that eat MOSFETs for breakfast seems pretty jagged if you ask me.
Lays? Lays is your bag, man. Those of us on the jagged edge eat jagged chips. I much prefer those jagged Tostios tortilla chips (which go well with the Tostitos salsa Jam was so kind to offer up). You could put an eye out with one of those things if you're not careful.
se
JAGGED POTATOES ASIDE...
Hi,
Uh? What's eating what here?
As far as I'm concerned FETs would only enter the picture as CCSs not as piranhas...
Me being me it would be an all tube exercise...as it evolves I do realise it won't be easy though.
Right, I don't even eat Belgian fries more than a few times a
year...you can keep the bag though.
Hi,
Uh? What's eating what here?
As far as I'm concerned FETs would only enter the picture as CCSs not as piranhas...
Me being me it would be an all tube exercise...as it evolves I do realise it won't be easy though.
Right, I don't even eat Belgian fries more than a few times a
year...you can keep the bag though.
Re: JAGGED POTATOES ASIDE...
It's dog eat dog, dog. Yo! Wassup?
Ok. How 'bout I see if I can talk Victor out of the schematics for a VK-P5?
Why, that's mighty generous for a Frenchman.
se
fdegrove said:
It's dog eat dog, dog. Yo! Wassup?
Ok. How 'bout I see if I can talk Victor out of the schematics for a VK-P5?
Right, I don't even eat Belgian fries more than a few times a year...you can keep the bag though.
Why, that's mighty generous for a Frenchman.
se
Hi,
Thinks...oh, yes....that Vic...I won't even think of stopping you.
We should be man enough to come up with at least as good a solution as what he spits out though...knowing how he tackles it won't hurt either...
As Hugh correctly suggested, the hard part is the summing.
Grrrr....I think I'll keep the bag after all....
Cheers,
Thinks...oh, yes....that Vic...I won't even think of stopping you.
We should be man enough to come up with at least as good a solution as what he spits out though...knowing how he tackles it won't hurt either...
As Hugh correctly suggested, the hard part is the summing.
Grrrr....I think I'll keep the bag after all....
Cheers,
fdegrove said:
Well you know, it's rather late in the game, the weekend's coming up and I figured it wouldn't hurt to take a little shortcut.
Actually I got to thinking about that comment. If the goal is to go balanced from input to output, what's there to sum? Did I miss something along the way?
Oh come on. No need to turn selfish. Just a little Hercule Poirot humor. You know, the famous "French" detective.
se
Yes, indeed, one must not design to ignore convention. It polarizes the appeal of the design and fences out a lot of people who otherwise would buy the product. Of course, for the radicals, this is very appealing, but there aren't many radicals out there.
While the idea of balanced all the way to the output is good, the only way to effectively derive the necessary voltage gain is to have independent (viz NOT differential) voltage stages. This assumes a tube power amplifier with a step down transformer of at least 20 to 1, and this means for 50W into 8R we need 56Vpp x 20, or 1080Vpp at the output tube plate, viz 560Vpp at each plate. In this instance, summing takes place without gain loss at the transformer, and God is in His heaven.
If each output tube has a voltage gain of 3, a fair figure (845/811-3), then we need a grid input of 560Vpp/3 = 187Vpp. So the split signals must each have at least four cascaded voltage amplifiers, since we start with 0.3mV, remember? This is mighty small.....
This entire calculation quickly gets out of hand unless we deal with the gain requirement in blocks. Hence, a head amp (hybrid cascode), followed by a preamp (high mu tube like the 6GK5), followed by a voltage block for the power amp (medium mu tubes for gain consistency and lowish impedance since voltage swings will be high and parasitics should not interfere).
I suggest we gun for 2Vpp sensitivity for the power amp section, and it won't hurt to regard the pre and power amp as separate. I have no problem with a fully balanced topology all the way; after all, if we are in a noisy environment the high CMRR is a strong advantage.
So, start at 2Vpp, so we need a power amp voltage gain of around 94. This can be done with just two gain stages, and medium mu tubes are the go. I'd suggest two cascaded 6SN7s, a la Herb Reichert's 'Flesh and Blood', followed by a Marantz Type 7 preamp cathode follower. This CF drives the grid of the output 845/811-3s. Of course, if you choose a higher mu output tube, you can eschew one of the gain blocks, just use a 6SL7 (mu is 70, around 45 with 100K plate load, and use it to drive a CF direct. This will reduce phase shift even further, and tighten up gain differences in the two balanced halves.
This way the entire signal is handled in balanced mode, and there is no conversion at any point along the way to SE. However, there is an issue with gain control in the two halves. Any asymmetry in gain will create uneven amplitudes in the two half cycles, and this will promote H2. The question is, will this adversely affect the sound? I say it won't if the differences are kept small, and if not, well, you essentially promote the distortion spectrum of an SET, and that ain't half bad.......
Cheers,
Hugh
While the idea of balanced all the way to the output is good, the only way to effectively derive the necessary voltage gain is to have independent (viz NOT differential) voltage stages. This assumes a tube power amplifier with a step down transformer of at least 20 to 1, and this means for 50W into 8R we need 56Vpp x 20, or 1080Vpp at the output tube plate, viz 560Vpp at each plate. In this instance, summing takes place without gain loss at the transformer, and God is in His heaven.
If each output tube has a voltage gain of 3, a fair figure (845/811-3), then we need a grid input of 560Vpp/3 = 187Vpp. So the split signals must each have at least four cascaded voltage amplifiers, since we start with 0.3mV, remember? This is mighty small.....
This entire calculation quickly gets out of hand unless we deal with the gain requirement in blocks. Hence, a head amp (hybrid cascode), followed by a preamp (high mu tube like the 6GK5), followed by a voltage block for the power amp (medium mu tubes for gain consistency and lowish impedance since voltage swings will be high and parasitics should not interfere).
I suggest we gun for 2Vpp sensitivity for the power amp section, and it won't hurt to regard the pre and power amp as separate. I have no problem with a fully balanced topology all the way; after all, if we are in a noisy environment the high CMRR is a strong advantage.
So, start at 2Vpp, so we need a power amp voltage gain of around 94. This can be done with just two gain stages, and medium mu tubes are the go. I'd suggest two cascaded 6SN7s, a la Herb Reichert's 'Flesh and Blood', followed by a Marantz Type 7 preamp cathode follower. This CF drives the grid of the output 845/811-3s. Of course, if you choose a higher mu output tube, you can eschew one of the gain blocks, just use a 6SL7 (mu is 70, around 45 with 100K plate load, and use it to drive a CF direct. This will reduce phase shift even further, and tighten up gain differences in the two balanced halves.
This way the entire signal is handled in balanced mode, and there is no conversion at any point along the way to SE. However, there is an issue with gain control in the two halves. Any asymmetry in gain will create uneven amplitudes in the two half cycles, and this will promote H2. The question is, will this adversely affect the sound? I say it won't if the differences are kept small, and if not, well, you essentially promote the distortion spectrum of an SET, and that ain't half bad.......
Cheers,
Hugh
AKSA said:
Since when did this discussion become about a commercial product? One of the great appealing qualities of DIY is that you don't have to stick with the same dozen or so tube types that manufacturers like ARC are forced to use. So stuff convention, that's just going to get us another cj.
When did the 50W requirement come into the discussion?
I've had 20W easily from 4 triode stages and a 300uV Denon. Add some pentodes in the right place and you're even better off. From the gist of your post, I'm assuming that all you're considering is CC or CF triodes.
Also, there are other tubes. A GM70 has a u=6 so half the grid drive requirements of an 845, which is very lazy at the grid, and about the same plate impedance, as well as a higher plate power. Ditto the trioded 813, with a u=8 and arguably better linearity than almost any other tube. And then there's the 211.......
Besides IMO, if you really need 50 tube watts, then your speakers are far too inefficient.
Are you deliberately taking the worst possible sounding topologies and adding them together for a reason? The F&B front end is so thick and creamy it can put you into a diabetic coma. And the Marantz 7 is the worst "highly rated" preamp ever made, and I speak from a lot of experience in trying to make it even better than mediocre.
OTTOMH for a gain of 90 roughly, how about a choke loaded trioded 6550/6L6 being driven by a CCS loaded 6H30. Two stages, FAR better drive for the 845 grid, lower distortion and more linear headroom than a CF or a common cathode 6SN7. Given a few minutes I could probably come up with another two dozen gain/driver topologies.
Oh, and on Thorsten's page, there's an EL84, in pentode, that gives a gain of around 100 in a very good sounding single stage, especially if you use it to drive a GM70.
fdegrove said:
Well, just that the subject was "Going balanced all the way."
Hugh's being too kind. I say, let 'em eat cake! The more you cater to these louts, the more they start to expect from you. Until one day, you've nothing left to give. And God help you then. They'll rip you to shreds.
Seriously though, a transformer (an ouput transformer in this case though you could just as well use an input transformer at the amp) will let you have your cake and eat it too. It's just as happy driving a balanced input as an unbalanced input. So the design can truly go "balanced all the way" and you don't have to give up any voltage gain in the process.
So, our famous "French" detective will have to investigate this thorny case.
Hmmm. I wonder if Suchet's French or Belgian?
se
Ah, yes, a little more technobabble!
Brett, Yes. I have deliberately chosen the worst sounding topologies because I'm actually trying to sabotage all this effort, no question. And my knowledge of electronics is primitive, forgive my appalling choices! You really should do something about that aggressive badinage; it gives us Aussies a bad name.
Further, 50W was a selected power rating which is moderate and practical for almost any speaker. I do not agree with your categoric statement about 'using the wrong speakers'. Truth be known, I do not like low power speakers, particularly full range drivers. I think they suck. The name of the game is realism, a little impact and slam, and I don't like that peaky, twee sound of the Lowther and Fostex. Yup, mainstream it should be. Personal taste and prejudice, man, we all have it....
However, some of your ideas are great! I very much like the idea of an EL84 in pentode as a single voltage amplifier in a power amp and it could be used to drive a big tube grid directly. My thanks to the ubiquitous Thorsten! Even in pentode, the EL84 in SE is a superb sounding tube, and yes, I also like the choice of 813 in triode as output, much more than a GM70 (I have one, BTW, don't like it much.) I'm confident Steve Bench would agree with you. Paul Le Clerq in the UK has done a lot of work on the 813 and believes it is a sleeper. I agree from what I have seen of the curves.
Lastly, I put it to Steve that this exercise should not be a test of design contrariness. It should have some mainstream standards, and they should be aligned with what many people would buy. This is not shameless commercialisation, this is just practicality and discipline. We are not writing 'Ullyses' here. To me this is not a Citroen DS23, this is a Renault 16. (Forgive my ancient model choices, this is tubes after all). We should not forget that the truly unforgettable designs are accessible, and their IP is every bit as outstanding as the Masers and Ferraris. (I admire the small block Chevy and the A12 Nissan, which I still believe are masterpieces of design economy and performance). I have no intention of building it for a commercial product, as I am fully committed to hybrids, and I'm doing this to improve my education and hopefully contribute for others.
Frank, you posted a very interesting conversion circuit, balanced to single. Ironically, I doodled just such a circuit yesterday. I first saw the idea from Fred Forsell, from Colorado I believe, about five years ago. One of the shortcomings was excessive H2 generation if the gain of the lower tube was not adjusted to match the cathode output of the top tube. The problem was that it was a push-PULL circuit, instead of a push-pull, if you get my drift.
This circuit seems to eliminate this problem very cleverly. The lower tube is a plate follower, with 100% feedback to ensure that both the output at the lower plate has voltage gain of 1 AND the Zout is commensurately reduced by local feedback to a point almost matching the 100R Zout of the upper cathode. This is very clever, could be the basis of an OTL actually, but it does require a 100K resistor grid stopper for the feedback divider, and this is never a good idea from either a noise or a control POV. I'd be tempted to either remove or reduce the upper tube 100K grid stop, too.
But it solves the issues beautifully with only a single bottle. I like it.
I'll make the point again that we should cater to both balanced and SE output. In this exercise, we should perhaps stick to the preamp only, and leave the power amp section to another thread.
Cheers,
Hugh
Brett, Yes. I have deliberately chosen the worst sounding topologies because I'm actually trying to sabotage all this effort, no question. And my knowledge of electronics is primitive, forgive my appalling choices! You really should do something about that aggressive badinage; it gives us Aussies a bad name.
Further, 50W was a selected power rating which is moderate and practical for almost any speaker. I do not agree with your categoric statement about 'using the wrong speakers'. Truth be known, I do not like low power speakers, particularly full range drivers. I think they suck. The name of the game is realism, a little impact and slam, and I don't like that peaky, twee sound of the Lowther and Fostex. Yup, mainstream it should be. Personal taste and prejudice, man, we all have it....
However, some of your ideas are great! I very much like the idea of an EL84 in pentode as a single voltage amplifier in a power amp and it could be used to drive a big tube grid directly. My thanks to the ubiquitous Thorsten! Even in pentode, the EL84 in SE is a superb sounding tube, and yes, I also like the choice of 813 in triode as output, much more than a GM70 (I have one, BTW, don't like it much.) I'm confident Steve Bench would agree with you. Paul Le Clerq in the UK has done a lot of work on the 813 and believes it is a sleeper. I agree from what I have seen of the curves.
Lastly, I put it to Steve that this exercise should not be a test of design contrariness. It should have some mainstream standards, and they should be aligned with what many people would buy. This is not shameless commercialisation, this is just practicality and discipline. We are not writing 'Ullyses' here. To me this is not a Citroen DS23, this is a Renault 16. (Forgive my ancient model choices, this is tubes after all). We should not forget that the truly unforgettable designs are accessible, and their IP is every bit as outstanding as the Masers and Ferraris. (I admire the small block Chevy and the A12 Nissan, which I still believe are masterpieces of design economy and performance). I have no intention of building it for a commercial product, as I am fully committed to hybrids, and I'm doing this to improve my education and hopefully contribute for others.
Frank, you posted a very interesting conversion circuit, balanced to single. Ironically, I doodled just such a circuit yesterday. I first saw the idea from Fred Forsell, from Colorado I believe, about five years ago. One of the shortcomings was excessive H2 generation if the gain of the lower tube was not adjusted to match the cathode output of the top tube. The problem was that it was a push-PULL circuit, instead of a push-pull, if you get my drift.
This circuit seems to eliminate this problem very cleverly. The lower tube is a plate follower, with 100% feedback to ensure that both the output at the lower plate has voltage gain of 1 AND the Zout is commensurately reduced by local feedback to a point almost matching the 100R Zout of the upper cathode. This is very clever, could be the basis of an OTL actually, but it does require a 100K resistor grid stopper for the feedback divider, and this is never a good idea from either a noise or a control POV. I'd be tempted to either remove or reduce the upper tube 100K grid stop, too.
But it solves the issues beautifully with only a single bottle. I like it.
I'll make the point again that we should cater to both balanced and SE output. In this exercise, we should perhaps stick to the preamp only, and leave the power amp section to another thread.
Cheers,
Hugh
Hi,
Read the title, read the book...
Nah, as you must have noticed this thread is a corporate spin-off.
Look at it this way; somewhere along the line ( no pun intended) you'll want to hook up a SE source ( I know I'm abusing your initials frequently) so a decent balun comes in handy.
Sure enough you CAN do this with yet another xformer, but that not what we're striving for...we still want to be able to lift this gasfactory without calling on Hercules Poirot.
Dear Hugh can have all the cale he wants, it was yours truly realising the importance of a balun and presenting it.
I realise you mean this in jest, from experience I can only say that it'll be the customers pushing you to your limits.
From which you ultimately benefit more than they will...
You have to stay ahead of the crowd, Steve...hard work, yet well worth it IMHO.
See above and beyond...we decided NOT to take the easy way out ...I may have to adopt your solution on my OTL amps for real estate reasons but let's not get ahead of ourselves.
Personally, I wouldn't loose any sleep over that if it weren't for the fact that Hugh sees it as a challenge to design a circuit that doesn't present this disadvantage.
After all 6dB isn't exactly peanuts.
Stop wondering. The good chap is as British as fish & chips...
Are you building a library of Agatha Christies or just wondering how on earth such a small country can play such good tennis?
Time to get your hands dirty, Steve and build some tube stuff...
Cheers,
Read the title, read the book...
Nah, as you must have noticed this thread is a corporate spin-off.
Look at it this way; somewhere along the line ( no pun intended) you'll want to hook up a SE source ( I know I'm abusing your initials frequently) so a decent balun comes in handy.
Sure enough you CAN do this with yet another xformer, but that not what we're striving for...we still want to be able to lift this gasfactory without calling on Hercules Poirot.
Dear Hugh can have all the cale he wants, it was yours truly realising the importance of a balun and presenting it.
I realise you mean this in jest, from experience I can only say that it'll be the customers pushing you to your limits.
From which you ultimately benefit more than they will...
You have to stay ahead of the crowd, Steve...hard work, yet well worth it IMHO.
See above and beyond...we decided NOT to take the easy way out ...I may have to adopt your solution on my OTL amps for real estate reasons but let's not get ahead of ourselves.
Personally, I wouldn't loose any sleep over that if it weren't for the fact that Hugh sees it as a challenge to design a circuit that doesn't present this disadvantage.
After all 6dB isn't exactly peanuts.
Stop wondering. The good chap is as British as fish & chips...
Are you building a library of Agatha Christies or just wondering how on earth such a small country can play such good tennis?
Time to get your hands dirty, Steve and build some tube stuff...
Cheers,
AKSA said:
Thanks. I thought you must have because they're hounds. Have you ever built used, modified or compared the circuits you mentioned with a range of similar topologies that do the same job?
I don't feel I'm aggressive, just blunt. I'm sorry you don't like it, but I'm not changing the way I express myself because it "might give Aussies a bad name". I don't speak for all Australians, just me, so don't throw that bollocks in my face again please.
Thanks for badinage; I had to look that one up.
Anything that needs 50W is probably in the 80's in sensitivity. I've never heard a realistic sounding speaker that low, with the Ariel at 92 being the lowest sensitivity speaker I'd ever consider again. It's been my experience that they bring more problems to the overall system performance than the amplifier.
I think they stink too, and I certainly wouldn't use a Fostex or a Lowther (except perhaps an Oris/Azura; interesting design, but I've yet to hear them). My system is hardly low power either; the lowest continuous RMS rating on any of the drivers in it is 200W.
Thanks, but I'd rather have high performance. Mediocre is easy to buy and doesn't need to be expensive, and that's what almost everything mainstream I've ever heard is.
Besides, who are you to railroad the thread into the direction and parameters you choose to be the most appropriate?
Thanks, but credit where it's due. I saw Thorsten's comments on the EL84, whom he got from Verdier, and decided to run with them. They work really well. Then Pete Millett mentioned something he tried and I investigated that and it sounded even better. Which led to further trials etc.
Agree re the 813, and it was Paul's amp that reminded me of the 813PP amp that the engineer I first trained with after school built. The EL84 is a peach too, but again, there are better, and they don't even need to be exotic, rare or expensive.
For some more ideas, try Stephen's 813PP. Very nice.
I don't agree with a single word of that. Who cares what people would buy? Because most of them buy crap. This is a DIY forum, and those that have been around this game a while know that whilst it's possible to do it cheaper, for similar money you can do it a whole lot better, and if you're going to put the effort in, why not do it better?
I can buy mainstream, how about high performance? I can afford to buy a Porsche if I want, same as I can own a Krell or a ML, but to my ears, there's higher perfomance to be had, much cheaper elsewhere. Mainstream is a Williamson.
Define what you mean by practicality. My amps are engineered to run for a looooong time, because I hate fixing things. I was trained that if it fails in service, you don't know how to design and/or build. Tubes chosen are quite readily available, modestly priced and don't need stupid B+ voltages etc etc.
I do know I've never seen anyone lust after a Renault 16 unless they wear an anorak and know the number and route of every train from memory.
If we're talking accessible and unforgettable, how about a Volkswagen or a 2CV? Italian exotics are nice, but I'd prefer a De Tomaso. At least it can be fixed easily if need be (I drive and ride hard), handles well and with some SVO bits goes like **** off a chrome shovel. I like American iron too, but my turbo rotary was my daily driver for years (in Sydney traffic) and it's best pass in street trim was about 11.2/136mph. It also won me a lot of money, as in those days the "mainstream" V8's couldn't touch it and it handle like on rails and with Honda VFR levels of compliance. Total cost was maybe $5k and it never let me down once.
I do agree with one thing though: lets keep this to a preamp for the time being, with a poweramp later.
fdegrove said:
Someday perhaps. I've still got this thing about high voltages after several nasty experiences as a kid. And since it'd take a fair amount of time to get up to speed with them, there's not much I could contribute to this thread as far as tubes go so I think I'll just watch the rest of this one from the grandstands.
se
CULTURE SHOCK...
Hi,
I am sorry to hear that...as a kid I had nasty experiences too, no shocks involved though...
Can live with that...which is probably the exact same answer I'd give Jam if he'd feel like twisting my arm again...
It may come as no surpise to you, but one can come close to the same satisfaction with sandswitches + xformers as one does with tubes...
Musical satisfaction should be the goal, so to paraphrase yourself:
Whatever makes you happy...
Hi,
I am sorry to hear that...as a kid I had nasty experiences too, no shocks involved though...
Can live with that...which is probably the exact same answer I'd give Jam if he'd feel like twisting my arm again...
It may come as no surpise to you, but one can come close to the same satisfaction with sandswitches + xformers as one does with tubes...
Musical satisfaction should be the goal, so to paraphrase yourself:
Whatever makes you happy...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.