Hi,
You can measure 200mA with a DMM either at the cathode or anode of the 6C33's. However it is more convenient to use the 0.1R cathode resistor (as per schematic) and adjust bias for 20mV across it. This way you will avoid the soldering-desoldering process each time you wish to confirm bias.
It doesn't really matter to have exactly 0.2A as long as you have minimum DC offset at the output (be patient here)
For substituting e182cc I do not think 12au7 can provide enough current to drive the 6c33 grids properly. For modern equivalent I would suggest the ecc99 or 6H30 although I wouldn't expect to find them cheaper than a good NOS e182cc.
For NOS I think that 12BH7 would be a good candidate here
Frank may have even better suggestions
Argiris
You can measure 200mA with a DMM either at the cathode or anode of the 6C33's. However it is more convenient to use the 0.1R cathode resistor (as per schematic) and adjust bias for 20mV across it. This way you will avoid the soldering-desoldering process each time you wish to confirm bias.
It doesn't really matter to have exactly 0.2A as long as you have minimum DC offset at the output (be patient here)
For substituting e182cc I do not think 12au7 can provide enough current to drive the 6c33 grids properly. For modern equivalent I would suggest the ecc99 or 6H30 although I wouldn't expect to find them cheaper than a good NOS e182cc.
For NOS I think that 12BH7 would be a good candidate here
Frank may have even better suggestions
Argiris
Efharisto Argiri:
Thanks indeed for your insight, though would any changes be necessary should I decide to use 6h30 or 12bh7 in this case in terms of resistors since I guess different operating conditions should be used for these tubes? What concerns me is not the cost but the lack of availability of e182cc.
Being a layman, can you please explain the concept of "dc offset"?
How do you measure this?
As far as using the 0.1 ohm resistor is concerned, that would measure the current on one of the 6c33cs. What about the other one?🙁
Thanks indeed for your insight, though would any changes be necessary should I decide to use 6h30 or 12bh7 in this case in terms of resistors since I guess different operating conditions should be used for these tubes? What concerns me is not the cost but the lack of availability of e182cc.
Being a layman, can you please explain the concept of "dc offset"?
How do you measure this?As far as using the 0.1 ohm resistor is concerned, that would measure the current on one of the 6c33cs. What about the other one?🙁
Hi,
The circuit was designed with the ECC82 in mind so you don't need to change anything there.
When you change to E1822CC, 12BH7A, ECC99 keep in mind the heater current requirements for these valves.
Other than that, I'd leave it as is.
For current sensing, you can put another 0.1 Ohm resistor on the anode lead of the other 6C33-C.
Also useful to have is a panel style ammeter for monitoring the current through the tubes.
Personally I have a pair of testpoints for measuring the bias voltage and a panelmeter to monitor the current of the output devices.
With a simple left right switch each tube can be checked.
A good example on how you can implement this is:
BRUCE ROZENBLIT 6AS7G OTL
As for any legal advice....I'll have to misbehave first now, won't I?
Maybe not...
Cheers,😉
The circuit was designed with the ECC82 in mind so you don't need to change anything there.
When you change to E1822CC, 12BH7A, ECC99 keep in mind the heater current requirements for these valves.
Other than that, I'd leave it as is.
For current sensing, you can put another 0.1 Ohm resistor on the anode lead of the other 6C33-C.
Also useful to have is a panel style ammeter for monitoring the current through the tubes.
Personally I have a pair of testpoints for measuring the bias voltage and a panelmeter to monitor the current of the output devices.
With a simple left right switch each tube can be checked.
A good example on how you can implement this is:
BRUCE ROZENBLIT 6AS7G OTL
As for any legal advice....I'll have to misbehave first now, won't I?
Maybe not...
Cheers,😉
e182cc
the 5687 should be a good sub for the182cc, easy to get and reasonably priced in comparison to the 6h30 and ecc99
DC offset, is the DC voltage that is present at the speaker terminals if the o/p stages are not ballanced up correctly, In practice, all directly coupled amps have some DC offset, so long as its below around 100mV its not generally a problem
the 5687 should be a good sub for the182cc, easy to get and reasonably priced in comparison to the 6h30 and ecc99
DC offset, is the DC voltage that is present at the speaker terminals if the o/p stages are not ballanced up correctly, In practice, all directly coupled amps have some DC offset, so long as its below around 100mV its not generally a problem
Gia sou Nico,
Some clarifications to your queries:
The first part was addressed by Alastair above. You measure it by running the amp idle (no input signal), using a dummy load (an 8 Ohm/10W resistor connected at the speaker terminals) and measuring with a DMM the DC value across it.
In my amp it is somewhat high, at 80mV, but it isn't a problem at all, even on very efficient Fostex horns I have tried it.
All recommendations above are meant to be drop-in replacements, I haven't looked into the datasheets in detail but I expect them to work correctly without any component changes. However you should be aware that the pinout is not the same and of the above only 5687 is pin to pin compatible. For the others you'd have to look up their pinout (second triode section is different).
In any event I wouldn't worry about the availability of e182cc, for instance I had a quick look on eBay and there are right now no less than 5 auctions on e182cc 🙂
You could use another one on the anode of the other, but in practice you could just measure the cathode, swap tubes (upper and lower) and measure again until you find a reasonably matched pair
If it's any comfort i am not an EE either 😉 but judging from your questions I get the impression this may be your first project. If this is the case, I wouldn't start with this one as it is somewhat demanding and with an OTL you have your speakers connected directly to the output tubes
In any event, should you decide to tackle it I'm sure you'll get a lot of help in this forum, in my case Frank and Hans (tubetvr) were very helpful
cheers
Argiris
Some clarifications to your queries:
The first part was addressed by Alastair above. You measure it by running the amp idle (no input signal), using a dummy load (an 8 Ohm/10W resistor connected at the speaker terminals) and measuring with a DMM the DC value across it.
In my amp it is somewhat high, at 80mV, but it isn't a problem at all, even on very efficient Fostex horns I have tried it.
All recommendations above are meant to be drop-in replacements, I haven't looked into the datasheets in detail but I expect them to work correctly without any component changes. However you should be aware that the pinout is not the same and of the above only 5687 is pin to pin compatible. For the others you'd have to look up their pinout (second triode section is different).
In any event I wouldn't worry about the availability of e182cc, for instance I had a quick look on eBay and there are right now no less than 5 auctions on e182cc 🙂
You could use another one on the anode of the other, but in practice you could just measure the cathode, swap tubes (upper and lower) and measure again until you find a reasonably matched pair
If it's any comfort i am not an EE either 😉 but judging from your questions I get the impression this may be your first project. If this is the case, I wouldn't start with this one as it is somewhat demanding and with an OTL you have your speakers connected directly to the output tubes
In any event, should you decide to tackle it I'm sure you'll get a lot of help in this forum, in my case Frank and Hans (tubetvr) were very helpful
cheers
Argiris
My OTL
My OTL is proceeding fine. I have chosen the driver section of Rozenblit with the output section of Ciuffoli using a 1kva power transformer and 3 huge chokes. Individual tube biasing for 6c33c is also to be used. I guess the amp would weigh more than 40 kg all housed in an aluminium amp casing with fan ventillation😀
Anyway I am at the point of using the output tubes and am considering 2 per channel 6c33c or 4 per channel EL 509 or 5-6 per channel 6aS7GA. I am more geared towards the 6c33c but have read so much about its unreliability from experienced heads. I have also heard about the unreliability and lack of "proper" bass control of 6AS7GA hence my plea for your vast experiences. Which tube would you consider as the best choice (million dollar question??😱
How can one manufacturer like Atma Shere use 6as7 exclusively while others find it unsuitable?
My OTL is proceeding fine. I have chosen the driver section of Rozenblit with the output section of Ciuffoli using a 1kva power transformer and 3 huge chokes. Individual tube biasing for 6c33c is also to be used. I guess the amp would weigh more than 40 kg all housed in an aluminium amp casing with fan ventillation😀
Anyway I am at the point of using the output tubes and am considering 2 per channel 6c33c or 4 per channel EL 509 or 5-6 per channel 6aS7GA. I am more geared towards the 6c33c but have read so much about its unreliability from experienced heads. I have also heard about the unreliability and lack of "proper" bass control of 6AS7GA hence my plea for your vast experiences. Which tube would you consider as the best choice (million dollar question??😱
How can one manufacturer like Atma Shere use 6as7 exclusively while others find it unsuitable?
Hi,
Regarding reliability, (or unreliability) of 6C33C or other tubes for that matter it really depends on who you believe in.
Viktor K who runs BAT has for one great trust in 6C33C, according to him there is an initial fallout of 1% in the first 100 hours and after that no special problems.
I also believe that 6C33C is a reliable tube which I base on my own daily used and several other OTL amps I have built for my friends using this tube, I have had no failure whatsoever for a total of 4 amplifiers with a total of 16 tubes, my own amp has now been working for ~4 years without any problem or even need to readjust bias.
Reliability of tubes in OTL amps has to do with how much you run them over spec which is the normal way in most amps, I run my 6C33C on 200mA idle current and 28W dissipation which is way below spec, maybe why that is why they last long.
I think that one of the biggest source of unreliabilty in OTL amps is parallelling of output tubes without proper derating and good circuits for current sharing, I have heard that many amps using 6AS7/6080 have had that problem where one failed tube start a chain reaction causing many tubes to fail in a short time, this is another reason to use 6C33C as one tube by itself can support peak currents up to 3.5A which make it possible to get significant output power with only 2 tubes per channel.
It seems that sockets has been reported as a problem with 6C33C, I never had that problem, as I reported here earlier I use original Russian sockets and I can not measure any significant voktage drop over the heater connections which must mean that it must be OK.
You write that you will be using a fan? How many W do you plan to get from each channel and how much idle current and anode voltage in each tube, I run a stereo amp in one chassis size ~450 x 300mm with a cage and even though it gets a bit hot I dont use a fan and did not had any problems.
It would be interesting to know what you mean by Bruce R input stage and Andreas output stage? I cant imagine how that will look like, can you show us a schematic?
Regards Hans
Regarding reliability, (or unreliability) of 6C33C or other tubes for that matter it really depends on who you believe in.
Viktor K who runs BAT has for one great trust in 6C33C, according to him there is an initial fallout of 1% in the first 100 hours and after that no special problems.
I also believe that 6C33C is a reliable tube which I base on my own daily used and several other OTL amps I have built for my friends using this tube, I have had no failure whatsoever for a total of 4 amplifiers with a total of 16 tubes, my own amp has now been working for ~4 years without any problem or even need to readjust bias.
Reliability of tubes in OTL amps has to do with how much you run them over spec which is the normal way in most amps, I run my 6C33C on 200mA idle current and 28W dissipation which is way below spec, maybe why that is why they last long.
I think that one of the biggest source of unreliabilty in OTL amps is parallelling of output tubes without proper derating and good circuits for current sharing, I have heard that many amps using 6AS7/6080 have had that problem where one failed tube start a chain reaction causing many tubes to fail in a short time, this is another reason to use 6C33C as one tube by itself can support peak currents up to 3.5A which make it possible to get significant output power with only 2 tubes per channel.
It seems that sockets has been reported as a problem with 6C33C, I never had that problem, as I reported here earlier I use original Russian sockets and I can not measure any significant voktage drop over the heater connections which must mean that it must be OK.
You write that you will be using a fan? How many W do you plan to get from each channel and how much idle current and anode voltage in each tube, I run a stereo amp in one chassis size ~450 x 300mm with a cage and even though it gets a bit hot I dont use a fan and did not had any problems.
It would be interesting to know what you mean by Bruce R input stage and Andreas output stage? I cant imagine how that will look like, can you show us a schematic?
Regards Hans
Hi,
Keep in mind that Ralf karsten uses USRR 6AS7Gs exclusively.
Personally I use 6080s but if I find time to rebuild the amp or start all over I'd go with the 6C33-Cs.
Me too, I am curious.
Cheers,😉
Keep in mind that Ralf karsten uses USRR 6AS7Gs exclusively.
Personally I use 6080s but if I find time to rebuild the amp or start all over I'd go with the 6C33-Cs.
Me too, I am curious.
Cheers,😉
I wonder if that 'bass control' perception is just a result of certain feedbackless OTL designs that have around a unity damping factor. I can say with conviction that this is not at all the case with my high damping factor dc coupled OTL. And I've had no reliability issues with the 6AS7G's I've used in my design over 16 years barring maybe a case or two of infant mortality. That being said, if I was building a new OTL today, I think I'd likely use 6C33C tubes for the output.
Hi,
While I could do that, I'm sure that if you're willing to read this thread you'll find plenty of information.
Enough, IMO, to make up your mind about what suits your needs best in terms of power and topology.
I'm sure that if you have any particular question you'll find plenty of people here willing to share their experience with you.
Cheers,😉
While I could do that, I'm sure that if you're willing to read this thread you'll find plenty of information.
Enough, IMO, to make up your mind about what suits your needs best in terms of power and topology.
I'm sure that if you have any particular question you'll find plenty of people here willing to share their experience with you.
Cheers,😉
I read the entire thread through, and thought it would be helpful to learn your perspective by comparing a design you favour to those of others.
On a different subject, has anyone thoughts on transformer coupling the driver to the output stage?
On a different subject, has anyone thoughts on transformer coupling the driver to the output stage?
Hi,
In a nutshell, the design that's closest to my own philosophy is the Andrea Ciuffoli OTL V 8.2.
It can easily be perfected by adding individual bias for the output tubes amongst other details.
If I'm not mistaken Tubetvr's OTL is similar in concept.
This is certainly not a new idea but most OPT coupled SEPP designs can be found in Japan.
As for use with a true OTL design,I have my doubts about it letting the magic through but I haven't tried it out either...
Cheers,😉
In a nutshell, the design that's closest to my own philosophy is the Andrea Ciuffoli OTL V 8.2.
It can easily be perfected by adding individual bias for the output tubes amongst other details.
If I'm not mistaken Tubetvr's OTL is similar in concept.
This is certainly not a new idea but most OPT coupled SEPP designs can be found in Japan.
As for use with a true OTL design,I have my doubts about it letting the magic through but I haven't tried it out either...
Cheers,😉
Why?
Why use a transformer in an OTL amp? The whole idea with OTL is to get rid of transformers and the problems they introduce as distorsion, frequency limitation and phase shift, transformers and inductors are the least ideal components used ín tube amps so it must be a general goal to get rid of these as much as possible.
Regards Hans
Hi,
Tubetvr,
Would you include PS chokes in that statement too?
I'm very much aware that these aren't the most linear devices but then what really is?
The main reason I ask is that it seems an economical way (to some extent) to regulate (again to some extent) the outputstage of an OTL as A. Ciuffoli does in his latest version of his design.
TIA,😉
Tubetvr,
Would you include PS chokes in that statement too?
I'm very much aware that these aren't the most linear devices but then what really is?
The main reason I ask is that it seems an economical way (to some extent) to regulate (again to some extent) the outputstage of an OTL as A. Ciuffoli does in his latest version of his design.
TIA,😉
It depends but what I really was refering to was chokes or transformers directly in the signal path.
I have used PS Chokes in the power supply for transformer coupled power amp as I dont really like the idea of building a high current tube regulated supply and never have achieved good enough results with SS based regulators for this purpose.
In my OTL amp I dont use any chokes and get very good results with reasonable size of capacitors and no active regulation.
Regards Hans
Hi, all -
In my dc coupled OTL design, none of the supplies are regulated. I rely partly on a balanced topology (except for supply connections to the totem pole output stage) to provide supply rejection. Since I use triode outputs, I find it helps to have as clean as possible output stage supply rails here particularly, hence a pi filter configuration with a series choke element for its supplies is used, as as is also done for the input and driver stage supplies. I also use a trick where I improve supply rejection through neutralization by summing small amounts of at least some of the supply voltages to the signal path in antiphase to their overall contributions through tube plates, cathodes and their associated resistors. I should reverse engineer the darn thing someday to get the exact details since I never drew an accurate or complete schematic of the dc coupled OTL. 🙄
In my dc coupled OTL design, none of the supplies are regulated. I rely partly on a balanced topology (except for supply connections to the totem pole output stage) to provide supply rejection. Since I use triode outputs, I find it helps to have as clean as possible output stage supply rails here particularly, hence a pi filter configuration with a series choke element for its supplies is used, as as is also done for the input and driver stage supplies. I also use a trick where I improve supply rejection through neutralization by summing small amounts of at least some of the supply voltages to the signal path in antiphase to their overall contributions through tube plates, cathodes and their associated resistors. I should reverse engineer the darn thing someday to get the exact details since I never drew an accurate or complete schematic of the dc coupled OTL. 🙄
Hi,
I should do the same really...
In my particular case the output stage is pretty straightforward as it is simply a bipolar supply taken from a bank of caps, +/- 160V
off 10.000µF electrolytics.
The bias is taken from those rails as well so it's floating with whatever moves ahead of it.
Since I'm in the process of improving (or trying) things I was wondering about adding a second bridge so each polarity would have it's own supply.
As mains voltage seems to go more up than down over the years I also noticed that in order to burn of some excess volts maybe a choke filter isn't such a bad idea.
The other stages being regulated by a tubed series reg anyway.
I also remember Julius Futterman and Harvey Rozenberg claiming that for the OTL-1 the regulated supply of the output made all the difference.
I can't help notice but see that Croft Acoustics have gone down the same path so something must make sense.
Many years ago I had the good fortune to be able to have a listen to an OTL-1 in Germany and I must say I'm still mightily impressed with it even though I wouldn't like to foot the electricity bill.
The amp (mine that is) still sounds better than most others I listened to over the past 15 years or so but you just don't know 'till you tried, right?
Anyway, opinions are welcomed.
Cheers,😉
I should do the same really...
In my particular case the output stage is pretty straightforward as it is simply a bipolar supply taken from a bank of caps, +/- 160V
off 10.000µF electrolytics.
The bias is taken from those rails as well so it's floating with whatever moves ahead of it.
Since I'm in the process of improving (or trying) things I was wondering about adding a second bridge so each polarity would have it's own supply.
As mains voltage seems to go more up than down over the years I also noticed that in order to burn of some excess volts maybe a choke filter isn't such a bad idea.
The other stages being regulated by a tubed series reg anyway.
I also remember Julius Futterman and Harvey Rozenberg claiming that for the OTL-1 the regulated supply of the output made all the difference.
I can't help notice but see that Croft Acoustics have gone down the same path so something must make sense.
Many years ago I had the good fortune to be able to have a listen to an OTL-1 in Germany and I must say I'm still mightily impressed with it even though I wouldn't like to foot the electricity bill.
The amp (mine that is) still sounds better than most others I listened to over the past 15 years or so but you just don't know 'till you tried, right?
Anyway, opinions are welcomed.
Cheers,😉
Hans, there is no *whole* idea for amps, or OTL amps which, by their name, in any event, are *output* transformerless; perhaps the amp you're thinking of is O&ITL? An interstage transformer quite elegantly executes the required phase splitting, but is a transformer with its limitations. Reduces the driver stage by a tube, however, and rids the circuit of a number of capacitors and resistors.😎
- Status
- Not open for further replies.